Cog Ad Thesis

Investigating Cognitive Function in Clinical and
Non-clinical Samples using Electroencephalography

and Psychometric Assessment: A Comparative
Study
George Kalatzis
B Med Sci (Hons)
March 2021
Principal supervisor: Associate Professor Sara Lal (UTS)
Co-supervisors: Dr Najah Nassif (UTS)

Associate Professor Chris Zaslawski (UTS)
Submitted in partial fulfilment of the requirements for the degree Doctor of Philosophy
(Science) at the University of Technology Sydney.
I. Declaration
I, George Kalatzis, declare that this thesis is submitted in fulfilment of the requirements for the
award of Doctor of Philosophy (Science), in the School of Life Sciences at the University of
Technology Sydney. This thesis is wholly my own work unless otherwise referenced or
acknowledged.
In addition, I certify that all information sources and literature used are indicated in the thesis.
This document has not been submitted for qualifications at any other academic institution.
This research is supported by the Australian Government Research Training Program.
Production Note:
Signature: Signature removed prior to publication.
Date: 31/3/2021
ii
II. Acknowledgements
This thesis has been the most significant and exacting undertaking of my professional career,
requiring sustained self-discipline, determination, and resilience. The individuals mentioned
below merit appreciable praise; they contributed to maintaining not only my sanity, but also the
final polished version of this thesis. Words alone are insufficient to reflect my appreciation of
them.
Foremost, I must thank my principal supervisor, Associate Professor Sara Lal, for affording me
with the opportunity to undertake this doctoral research. Your unshakeable support, wisdom,
incisiveness, and scientific know-how empowered me to persevere during several stages of this
zigzagging journey despite repeated setbacks. These same qualities also motivated me to
continually advance myself intellectually and to always strive for high standards. In addition,
your critical, constructive feedback during the preparation of this thesis, as well as seminar
presentations, was also invaluable and helped develop a more robust piece of work. I must also
thank my co-supervisors, Dr Najah Nassif and Associate Professor Chris Zaslawski. Your
viewpoints, insights, and feedback on thesis chapters and seminar presentations helped
strengthen the quality of this research. I also extend my thanks to the Faculty of Science for
providing me with the necessary facilities and resources to undertake my doctoral candidature.
I am immeasurably grateful and indebted to Dr Elizabeth Louise May, who proofread the thesis
and provided substantial editing input. Her uncompromising standards for writing emboldened
me to become a stickler for proper English, using every word and comma with precision and
economy, and to consult the acknowledged authority (Fowler’s Modern English Usage) for all
matters concerning English usage whenever uncertain. It is my hope that the examiners enjoy
reading this dissertation half as much as they did reading hers.
iii
I would also like to extend sincere appreciation to the professional organisations, Diabetes
Australia, Diabetes New South Wales (NSW), and Alzheimer’s Australia, for supporting and
promoting the research and facilitating recruitment of the clinical groups. Although these
groups were difficult to recruit, I take pride in knowing this research may, infinitesimally and
indirectly, someday benefit people living with diabetes mellitus (Type 1 and Type 2) and
hypertension. Importantly, I hope it spotlights the need for continued research into these
challenging and prevalent chronic diseases.
Considerable thanks are also due for my colleagues in the Neuroscience Research Unit (NRU).
They inspired me to persevere in light of the highs and lows of academia, listened to countless
seminar practice sessions, and helped maintained my marbles over numerous lunch outings. I
also owe tremendous thanks to all study volunteers who participated in the study – whether
immediate family, friends, or strangers. Without their involvement (and their enthusiasm and
interest), this research would not have been possible or have advanced.
Finally, I express my utmost gratitude and deepest heartfelt thanks to my treasured family and
friends and dedicate this thesis to those who supported me every step of the way – the late nights
and the countless merlins-beard-what-have-I-gotten-myself-into moments – throughout this
doctoral candidature. You all know who you are. Your encouraging words, limitless patience
and understanding, innumerable pep talks, and unfailing support (psychological and emotional),
empowered me to persist in the face of adversity and the unknown long after I had given up
during this marathon. Just. Thank you.
iv
III. Thesis Format
This thesis is formatted as a conventional thesis and hence is structured as a series of chapters.
v
IV. Publications and Presentations
Published Abstracts/Conference Presentations:
1. Lees, T., Maharaj, S., Kalatzis, G., Nassif, N., Newton, P., & Lal, S. The neurocognitive
relationship between stress and anxiety, memory and decision-making performance of
Australian Nurses. Poster presentation: 58th Annual meeting of the Society for
Psychophysiological Research 2018, Quebec City, Canada.
2. Kalatzis, G., Lees, T., Nassif, N., Zaslawski, C., & Lal. S. Exploring cognitive function
in diabetes and non-diabetes samples using electroencephalography (EEG) and
psychometric assessment: A comparative study. Oral presentation: The 37th Annual
Scientific Meeting of the Australasian Neuroscience Society 2017, Sydney, 3rd – 6th
December.
3. Kalatzis, G., Lees, T., Nassif, N., Zaslawski, C., & Lal, S. Investigating cognitive
function in diabetes and healthy samples using electroencephalography (EEG) and
psychometric assessment: a comparative study. Oral presentation: The Inter-University
Neuroscience & Mental Health Conference 2016, Sydney, 20th – 21st September.
4. Kalatzis, G., Lees, T., Nassif, N., Zaslawski, C., & Lal, S. Investigating cognitive
function in diseased states: electroencephalography (EEG) and psychometric
assessment. Poster presentation: The New Horizons Conference 2015, Sydney, 23rd –
25th November.
5. Lees, T., Kalatzis, G., & Lal, S. (2015). Examining negative mental states and their
association to psychometric and electroencephalographic measures of cognitive
performance in Australian Nurses. Psychophysiology, 52 (S24), doi:
10.1111psyp.12495.
6. Lees, T., Kalatzis, G., & Lal, S. Examining negative mental states and their association
to psychometric and electroencephalographic measures of cognitive performance in
Australian Nurses. Poster presentation: 55th Annual meeting of the Society for
Psychophysiological Research 2015, Seattle, USA.
vi
Other (manuscripts currently submitted or pending outcome):
1. Lees, T., Maharaj, S., Kalatzis, G., Nassif, N., Newton, P., & Lal, S. (2020).
Electroencephalographic prediction of global and domain-specific cognitive
performance of clinically-active Australian Nurses. Physiological Measurement
(Accepted – 20/08/2020).
2. Kalatzis, G., Lees, T., Nassif, N., Zaslawski, C., & Lal, S. (2020). Changes in EEG
activity as an indicator of early cognitive dysfunction in diabetes mellitus: a review.
Journal of Diabetes and Its Complications (Due for resubmission).
vii
V. Table of Contents
I. Declaration .............................................................................................................................. ii
II. Acknowledgements ............................................................................................................... iii
III. Thesis Format ....................................................................................................................... v
IV. Publications and Presentations ............................................................................................ vi
V. Table of Contents ................................................................................................................ viii
VI. List of Figures ................................................................................................................... xiii
VII. List of Tables ................................................................................................................. xviii
VIII. Abbreviations ................................................................................................................ xxiii
IX. Abstract .......................................................................................................................... xxvii
Chapter 1 – Introduction ......................................................................................................... 1
1.1 Ageing and Health....................................................................................................... 1
1.2 Cognition ..................................................................................................................... 4
1.2.1 Mild Cognitive Impairment (MCI) ................................................................. 7
1.2.2 Major Neurocognitive Disorder (NCD) (Dementia) .................................... 11
1.2.3 Alzheimer’s Disease (AD) ............................................................................ 15
1.2.4 Vascular Dementia (VaD) ............................................................................ 21
1.3 Risk Factors for Cognitive Impairment..................................................................... 22
1.3.1 Diabetes Mellitus .......................................................................................... 24
1.3.2 Type 1 Diabetes Mellitus (T1DM) ............................................................... 26
1.3.3 Type 2 Diabetes Mellitus (T2DM) ............................................................... 27
1.4 Diabetes mellitus: Complications ............................................................................. 32
1.5 Diabetes Mellitus and Cognitive Function................................................................ 34
1.5.1 Type 1 Diabetes Mellitus (T1DM) and Cognitive Function ........................ 37
1.5.2 Type 2 Diabetes Mellitus (T2DM) and Cognitive Function ........................ 40
1.6 Mechanistic Contributors to Cognitive Dysfunction in Diabetes ............................. 41
1.6.1 Hyperglycaemia ............................................................................................ 41
1.6.2 Recurrent Hypoglycaemia ............................................................................ 43
1.6.3. Altered Insulin Signalling ............................................................................ 45
1.6.4. Blood-brain Barrier Dysfunction ................................................................. 48
1.7 High Blood Pressure ................................................................................................. 51
1.7.1 Hypertension: Complications ....................................................................... 53
viii
1.7.2 Hypertension and Cognitive Function .......................................................... 55
1.7.3 High Blood Pressure and Cognition: Evidence from Cross-sectional Studies
............................................................................................................................... 56
1.7.4 High Blood Pressure and Cognition: Evidence from Longitudinal Studies
............................................................................................................................... 57
1.8 Mechanistic Contributors to Cognitive Dysfunction in Hypertension ...................... 59
1.8.1 Blood-brain Barrier Dysfunction .................................................................. 59
1.8.2 Impaired Neurovascular Coupling ................................................................ 60
1.8.3 Small Vessel Disease (SVD) ........................................................................ 62
1.9 Electroencephalography (EEG) ................................................................................ 64
1.9.1 Delta Waves .................................................................................................. 67
1.9.2 Theta Waves ................................................................................................. 67
1.9.3 Alpha Waves ................................................................................................. 67
1.9.4 Beta Waves ................................................................................................... 67
1.9.5 Gamma Waves .............................................................................................. 68
1.10 Electroencephalography Changes in Diabetes Mellitus .......................................... 69
1.10.1 Changes in Slow-wave EEG activity in Diabetes Mellitus ........................ 70
1.10.2 Changes in Fast-wave EEG activity in Diabetes Mellitus .......................... 72
1.11 Electroencephalography Changes in Hypertension .............................................. 78
1.12 Effect of glucose lowering and anti-hypertensive medication on EEG ................ 80
1.13 Basis and Study Significance................................................................................ 81
1.13.1 Implications of the Present Study ............................................................... 81
1.14 Hypotheses ............................................................................................................ 84
1.15 General Aims ........................................................................................................ 84
1.16 Specific Aims (Aim 1) .......................................................................................... 85
1.17 Specific Aims (Aim 2) .......................................................................................... 85
Chapter 2 – Methodology....................................................................................................... 86
2.1 Methodology Summary .......................................................................................... 86
2.2 Ethics Approval and Consent ................................................................................. 86
2.3 Recruitment of Study Participants .......................................................................... 87
2.4 Study Inclusion/Exclusion Criteria ......................................................................... 87
2.5 Blood Pressure Measurement ................................................................................. 88
2.5.1 BP Inclusion/Exclusion Criteria ................................................................. 90
2.6 Blood Glucose Level (BGL) Determination .......................................................... 93
ix
2.7 Demographic Data Acquisition .............................................................................. 95
2.7.1 Lifestyle Appraisal Questionnaire (LAQ) .................................................. 95
2.8 Electroencephalography Data Acquisition ............................................................. 97
2.8.1 Stroop Colour Word Test ......................................................................... 101
2.9 Cognitive Assessment........................................................................................... 102
2.9.1 Mini-Mental State Examination (MMSE) ................................................ 103
2.9.2 Cognistat ................................................................................................... 106
2.10 Data Processing and Analysis ............................................................................. 113
2.10.1 Electroencephalography Data Pre-Processing ......................................... 113
2.11. Statistical Analysis ............................................................................................ 115
2.11.1 Power Analysis ........................................................................................ 115
2.11.2 Dependent Sample T-test ......................................................................... 115
2.11.3 Wilcoxon Signed Rank Test .................................................................... 115
2.11.4 Mann Whitney U Test.............................................................................. 116
2.11.5 Partial Pearson’s Correlation ................................................................... 116
2.11.6 Spearman’s Rank Order Correlation........................................................ 116
2.11.7 Multiple Analysis of Covariance (MANCOVA) ..................................... 117
Chapter 3 – Results: Demographic Characteristics (Non-clinical and clinical) ............. 118
3.1 Participant Summary ............................................................................................ 119
3.2 Demographic Variables ........................................................................................ 119
3.2.1 Cardiovascular Variables (SBP, DBP, and HR) ......................................... 122
3.2.2 Blood Glucose Level (BGL) ....................................................................... 125
3.2.3 Disease-specific Variables .......................................................................... 127
Chapter 4 – Associations between Blood Pressure, Blood Glucose Level, and Cognitive
Performance (Non-clinical) .................................................................................................. 129
4.1 Cognitive Performance ......................................................................................... 129
4.1.1 Global Cognitive Performance (MMSE) .................................................... 129
4.1.2 Domain-specific Cognitive Performance ................................................... 130
4.1.3 Stroop Colour Word Test............................................................................ 133
4.2 Associations between BP and Cognition for the Non-clinical Group .................. 137
4.3 Associations between BGL and Cognition for the Non-clinical Group ............... 137
4.4 Associations between BP, BGL, and Cognition for the T1DM Group ................ 139
4.5 Associations between BP, BGL, and Cognition for the T2DM Group ................ 141
4.6 Associations between BP, BGL, and Cognition for the HTN Group ................... 145
x
4.7 Discussion: Cognitive Performance of Non-clinical and Clinical Groups ........... 147
4.7.1 Cognitive Performance: Clinical Groups (T1DM, T2DM, and HTN) ....... 147
4.8 Associations: Non-clinical Group......................................................................... 155
4.9 Associations: Clinical Groups .............................................................................. 157
4.10 Conclusions: Cognitive Performance ................................................................. 159
Chapter 5 – Associations between Blood Pressure, Blood Glucose Level, and
Electroencephalography (Non-clinical) .............................................................................. 160
5.1. Associations between pre-study SBP and EEG activity ...................................... 160
5.2 Associations between post-study SBP and EEG activity ..................................... 168
5.3 Associations between pre-study DBP and EEG activity ...................................... 171
5.4 Associations between post-study DBP and EEG activity..................................... 173
5.5 Associations between pre-study BGL and EEG activity ...................................... 176
5.6 Associations between post-study BGL and EEG activity .................................... 180
5.7 Discussion: Associations between BP and BGL and EEG ................................... 182
5.7.1 Associations between BP and EEG ............................................................ 182
5.7.2 Associations between BGL and EEG ......................................................... 185
5.8 Conclusions........................................................................................................... 189
Chapter 6 – Associations between Blood Pressure, Blood Glucose Level, and
Electroencephalography (Clinical) ..................................................................................... 190
6.1 Type 1 Diabetes Mellitus ...................................................................................... 190
6.1.1 Associations between pre-study SBP and EEG activity ........................... 190
6.1.2 Associations between pre-study DBP and EEG activity .......................... 195
6.1.3 Associations between pre-study BGL and EEG activity .......................... 197
6.1.4 Associations between post-study SBP and EEG activity ......................... 202
6.1.5 Associations between post-study DBP and EEG activity ........................ 203
6.1.6 Associations between post-study BGL and EEG activity ........................ 206
6.1.7 Associations between disease-specific variables and EEG activity ......... 211
6.2 Type 2 Diabetes Mellitus ...................................................................................... 214
6.2.3 Associations between DBP (pre and post) and EEG activity ................... 221
6.2.4 Associations between BGL (pre and post) and EEG activity ................... 221
6.2.5 Associations between HbA1C and EEG activity ...................................... 224
6.2.6 Associations between disease duration and EEG activity ........................ 226
xi
6.3 Hypertension ......................................................................................................... 229
6.3.3 Associations between pre-study DBP and EEG activity .......................... 233
6.3.4 Associations between post-study DBP and EEG activity ........................ 236
6.3.5 Associations between pre-study BGL and EEG activity .......................... 239
6.3.6 Associations between post-study BGL and EEG activity ........................ 239
6.4 Discussion ............................................................................................................. 241
6.4.1 Associations between BP and EEG activity (T1DM and T2DM) ............ 241
6.4.2 Associations between BP and EEG activity (HTN) ................................. 242
6.4.3 Associations between BGL and EEG activity (T1DM and T2DM) ......... 246
6.5 Conclusions .......................................................................................................... 251
Chapter 7 – Limitations, Future Directions, and Conclusions ......................................... 253
7.1 Limitations and Future Directions ........................................................................ 253
7.2 Conclusions........................................................................................................... 261
Chapter 8 – Appendices ....................................................................................................... 265
8.1 Consent Form (Non-clinical and Clinical) ........................................................... 265
8.2 Emergency Protocol.............................................................................................. 266
8.3 Chronic Disease Questionnaire (Diabetes Mellitus)............................................. 268
8.4 Chronic Disease Questionnaire (Hypertension) ................................................... 270
8.5 Cognitive Profile ................................................................................................... 272
8.6 Study Summary Sheet........................................................................................... 273
8.7 Recruitment Poster................................................................................................ 274
8.8 Participant Remuneration Form (Clinical only) ................................................... 275
8.9 Breakdown of glucose-lowering medication ........................................................ 276
8.10 Breakdown of anti-hypertensive medication ...................................................... 277
Chapter 9 – References ........................................................................................................ 278
xii
VI. List of Figures
Figure 1.1 – Life expectancy by birth and sex (1886-2016) ................................................... 1
Figure 1.2 – Estimated number of Australians categorised by age group in 2018.................. 3
Figure 1.3 – Predicted number of Australians over the age of 65 years categorised by age
group (2016-2096) .............................................................................................. 3
Figure 1.4 – Lateral view of the cerebral cortex ..................................................................... 6
Figure 1.5 – Diagnostic criteria used to diagnose mild cognitive impairment........................ 8
Figure 1.6 – Hypothetical trajectories in normal brain ageing .............................................. 10
Figure 1.7 – Projected prevalence of dementia in Australians (2016-2056) ......................... 12
Figure 1.8 – Diagnostic criteria used to diagnose major neurocognitive disorder/dementia..14
Figure 1.9 – Key cognitive domains detrimentally affected by cognitive disorders ............. 14
Figure 1.10 – Healthy brain tissue compared to Alzheimer’s disease brain tissue ................. 17
Figure 1.11 – Brain areas affected in the early stages of Alzheimer’s disease ....................... 18
Figure 1.12 – Patterns in the accumulation of neuropathologies in Alzheimer’s disease ....... 19
Figure 1.13 – Leading causes of death in Australians (males and females) in 2016 .............. 20
Figure 1.14 – Risk factors for cognitive impairment across the lifespan ................................ 23
Figure 1.15 – Homeostatic regulation of blood glucose concentration ................................... 25
Figure 1.16 – Impaired regulation of blood glucose concentration in Type 2 diabetes .......... 28
Figure 1.17 – Predicted prevalence of diabetes (20-79 years) in 2019, 2035, and 2045 ........ 31
Figure 1.18 – Trends in leading cause of disease in Australia ................................................ 32
Figure 1.19 – Possible pathophysiological pathways linking Type 2 diabetes to dementia ... 37
Figure 1.20 – Disturbances at the cellular level associated with glucose neurotoxicity ......... 43
Figure 1.21 – Mechanisms linking impaired insulin signalling to Alzheimer’s disease ......... 47
Figure 1.22 – The blood-brain barrier ..................................................................................... 49
Figure 1.23 – Risk factors linked to the development of hypertension ................................... 52
xiii
Figure 1.24 – Disturbances associated with blood-brain barrier breakdown .......................... 60
Figure 1.25 – A neurovascular unit ......................................................................................... 61
Figure 1.26 – Cerebral arteries damaged by hypertension ...................................................... 63
Figure 1.27 – The International 10-20 system of electroencephalography ............................. 65
Figure 2.1 – Non-invasive digital blood pressure monitor .................................................... 89
Figure 2.2 – Posture and seating position for recording blood pressure ............................... 90
Figure 2.3 – Blood glucometer and sterile, single-use lancing device .................................. 94
Figure 2.4 – Normal postprandial blood glucose level range after a 2-hour fast .................. 94
Figure 2.5 – 32-channel, non-invasive elastic EEG cap........................................................ 97
Figure 2.6 – Top view of 10-20 system with electrode positions highlighted ...................... 98
Figure 2.7 – Electrode gel, sterile syringe, and blunted needle............................................. 99
Figure 2.8 – Unprocessed EEG tracing ............................................................................... 100
Figure 2.9 – Screenshot of the Stroop Colour Word Test Program .................................... 101
Figure 2.10 – The Mini-Mental State Examination (MMSE) ............................................... 104
Figure 2.11 – Stimulus sheet (Part 2) of the MMSE ............................................................. 105
Figure 2.12 – Participant completing Construction sub-test of the Cognistat....................... 107
Figure 2.13 – The Cognistat cognitive status profile ............................................................ 108
Figure 2.14 – Protocol for the present study ......................................................................... 112
Figure 3.1 – Breakdown of groups comprising the total study cohort ................................ 118
Figure 4.1 – Positive correlation between age of disease onset and average response time for
matched stimuli in the Stroop Test for Type 2 diabetes mellitus ................... 136
Figure 4.2 – Negative correlation between post-study blood glucose level and total
Mini Mental State Examination score for the non-clinical group................... 138
Figure 4.3 – Positive correlation between post-study diastolic blood pressure and judgement
performance in the Cognistat for Type 1 diabetes mellitus ............................ 140
xiv
Figure 4.4 – Positive correlation between post-study systolic blood pressure and construction
Figure 4.5 – Positive correlation between post-study diastolic blood pressure and judgement
Figure 4.6 – Positive correlation between post-study diastolic blood pressure and total
Cognistat score for Type 2 diabetes mellitus .................................................. 144
Figure 4.7 – Negative correlation between pre-study blood glucose level and the attention
domain of the Cognistat for Type 2 diabetes mellitus .................................... 145
Figure 4.8 – Negative correlation between pre-study blood glucose level and the similarities
domain of the Cognistat for Type 2 diabetes mellitus .................................... 146
Figure 5.1 – Positive correlation between pre-study systolic blood pressure and alpha power
at T7 during the baseline phase for the non-clinical group ............................. 162
Figure 5.2 – Positive correlation between pre-study systolic blood pressure and beta power at
FT8 during the baseline phase for the non-clinical group ............................... 163
Figure 5.3 – Positive correlation between pre-study systolic blood pressure and alpha power
at FC3 during the active phase for the non-clinical cohort.............................. 166
Figure 5.4 – Positive correlation between pre-study systolic blood pressure and gamma
power at FC3 during the active phase for the non-clinical group ................... 167
Figure 5.5 – Positive correlation between post-study systolic blood pressure and gamma
power at FP2 during the active phase for the non-clinical group .................... 170
Figure 5.6 – Positive correlation between pre-study diastolic blood pressure and alpha power
at PZ during the baseline phase for the non-clinical group ............................. 172
Figure 5.7 – Positive correlation between post-study diastolic blood pressure and gamma
power at FP2 during the active phase for the non-clinical group .................... 175
xv
Figure 5.8 – Positive correlation between pre-study blood glucose level and beta power at F3
during the baseline phase for the non-clinical group ...................................... 177
Figure 5.9 – Negative correlation between pre-study blood glucose level and theta power at
FZ during the active phase for the non-clinical group ..................................... 179
Figure 5.10 – Negative correlation between post-study blood glucose level and theta power at
CZ during the active phase for the non-clinical cohort ................................... 181
Figure 5.11 – The association between cerebral blood flow and EEG activity .................... 185
Figure 5.12 – Example of an EEG tracing recorded during euglycaemia and hypoglycaemia in
the same individual ......................................................................................... 187
Figure 6.1 – Negative correlation between pre-study systolic blood pressure and theta power
at FT7 during the active phase for Type 1 diabetes mellitus ........................... 193
Figure 6.2 – Negative correlation between pre-study systolic blood pressure and delta power
at FT7 during the active phase for Type 1 diabetes mellitus ........................... 194
Figure 6.3 – Negative correlation between pre-study diastolic blood pressure and theta power
at FT7 during the baseline phase for Type 1 diabetes mellitus ....................... 196
Figure 6.4 – Positive correlation between pre-study blood glucose level and theta power at
FP1 during the baseline phase for Type 1 diabetes mellitus ........................... 199
Figure 6.5 – Positive correlation between pre-study blood glucose level and theta power at
OZ during the active phase for Type 1 diabetes mellitus ................................ 201
Figure 6.6 – Positive correlation between post-study diastolic blood pressure and gamma
power at FC3 during the active phase for Type 1 diabetes mellitus ................ 205
TP7 during the baseline phase for Type 1 diabetes mellitus ........................... 208
P8 during the active phase for Type 1 diabetes mellitus ................................. 210
xvi
Figure 6.9 – Positive correlation between glycosylated haemoglobin and delta power at T7
during the active phase for Type 1 diabetes mellitus ...................................... 213
Figure 6.10 – Negative correlation between pre-study systolic blood pressure and delta power
at O2 during the baseline phase for Type 2 diabetes mellitus ......................... 215
Figure 6.11 – Positive correlation between pre-study SBP and gamma power at P3 during the
active phase for Type 2 diabetes mellitus ....................................................... 217
Figure 6.12 – Negative correlation between post-study systolic blood pressure and beta power
at C4 during the baseline phase for Type 2 diabetes mellitus ......................... 219
Figure 6.13 – Positive correlation between pre-study blood glucose level and delta power at
TP8 during the active phase for Type 2 diabetes mellitus ............................... 223
Figure 6.14 – Positive correlation between disease duration and gamma power at FT7 during
the active phase for Type 2 diabetes mellitus ................................................. 228
Figure 6.15 – Positive correlation between pre-study systolic blood pressure and theta power
at PZ during the baseline phase for hypertension ............................................ 230
Figure 6.16 – Positive correlation between pre-study diastolic blood pressure and theta power
at TP7 during the baseline phase for hypertension .......................................... 235
Figure 6.17 – Negative correlation between post-study diastolic blood pressure and gamma
power at FT7 during the active phase for hypertension .................................. 238
Figure 7.1 – Framework for diagnosing and evaluating cognitive dysfunction in Type 2
diabetes mellitus.............................................................................................. 262
xvii
VII. List of Tables
Table 1.1 – Key disease characteristics of type 1 and type 2 diabetes mellitus................... 30
Table 1.2 – Brain waves commonly observed in an electroencephalogram and their
corresponding psychophysiological state(s) ..................................................... 68
Table 1.3 – Summary of main findings from studies exploring EEG activity in diabetes
mellitus .............................................................................................................. 75
Table 2.1 – Blood pressure exclusion and inclusion limit thresholds .................................. 92
Table 2.2 – Maximum achievable scores and domain-specific impairment threshold scores
or each cognitive domain of the Cognistat ..................................................... 111
Table 3.1 – Key demographic characteristics for all groups.............................................. 121
Table 3.2 – Pre-study and post-study cardiovascular variables as well as the change that
occurred for each group .................................................................................. 124
Table 3.3 – Pre-study and post-study BGL as well as the change that occurred for each group
......................................................................................................................... 126
Table 3.4 – Disease-specific variables solicited from the clinical groups ......................... 127
Table 4.1 – Mean scores obtained by each group in the Mini-Mental State Examination ..130
Table 4.2 – Mean scores obtained by each group for individual domains of the Cognistat as
well as total Cognistat score ........................................................................... 131
Table 4.3 – Mean average response times obtained by each group for matched and
mismatched stimuli of the Stroop Colour Word Test ..................................... 134
Table 4.4 – Associations between disease-specific variables and matched aspects of the
Colour Word Test ........................................................................................... 135
Table 5.1 – Associations between pre-study SBP and EEG activity during the baseline phase
for the non-clinical group................................................................................ 161
xviii
Table 5.2 – Associations between pre-study SBP and EEG activity during the active phase
Table 5.3 – Associations between post-study SBP and EEG activity during the active phase
Table 5.4 – Associations between pre-study DBP and EEG activity during the baseline phase
Table 5.5 – Associations between post-study DBP and EEG activity during the baseline
phase for the non-clinical group ..................................................................... 173
Table 5.6 – Associations between post-study DBP and EEG activity during the active phase
Table 5.7 – Associations between pre-study BGL and EEG activity during the baseline phase
Table 5.8 – Associations between pre-study BGL and EEG activity during the active phase
Table 5.9 – Associations between post-study BGL and EEG activity during the active phase
for the T1DM group ........................................................................................ 191
in the T1DM group ......................................................................................... 192
for the T1DM group ........................................................................................ 195
Table 6.4 – Associations between pre-study DBP and EEG activity during the active phase
for the T1DM group ........................................................................................ 197
xix
Table 6.5 – Associations between pre-BGL and EEG activity during the baseline phase for
the T1DM group ............................................................................................. 198
Table 6.6 – Associations between pre-BGL and EEG activity during the active phase for the
T1DM group ................................................................................................... 200
Table 6.7 – Associations between post-study SBP and EEG activity during the baseline
phase for the T1DM group .............................................................................. 202
for the T1DM group ........................................................................................ 203
for the T1DM group ........................................................................................ 204
Table 6.11 – Associations between post-study BGL and EEG activity during the baseline
for the T1DM group ........................................................................................ 209
Table 6.13 – Associations between disease-specific variables (HbA1c, disease duration) and
EEG activity during the baseline phase for the T1DM group ........................ 211
Table 6.14 – Associations between disease-specific variables (HbA1C, disease duration) and
EEG activity during the active phase for the T1DM group ............................ 212
for the T2DM group ........................................................................................ 214
for the T2DM group ........................................................................................ 216
xx
for the T2DM group ........................................................................................ 220
for the T2DM group ........................................................................................ 221
Table 6.20 – Associations between pre-study BGL and EEG activity during the baseline phase
for the T2DM group ........................................................................................ 222
for the T2DM group ........................................................................................ 222
Table 6.22 – Associations between glycosylated haemoglobin (HbA1C) and EEG activity
during the baseline phase for the T2DM group .............................................. 225
Table 6.23 – Associations between glycosylated haemoglobin (HbA1C) and EEG activity the
active phase for the T2DM group ................................................................... 226
Table 6.24 – Associations between disease duration and EEG activity during the active phase
for the T2DM group ........................................................................................ 227
for the HTN group .......................................................................................... 229
for the HTN group .......................................................................................... 231
phase for the HTN group ................................................................................ 232
for the HTN group .......................................................................................... 233
xxi
for the HTN group .......................................................................................... 234
phase for the HTN group ................................................................................ 236
for the HTN group .......................................................................................... 237
for the HTN group .......................................................................................... 239
for the HTN group .......................................................................................... 240
xxii
VIII. Abbreviations
! – Alpha
Ab – Amyloid-beta
AChEI – Acetylcholinesterase Inhibitor
AD – Alzheimer’s Disease
ADA – American Diabetes Association
AgCl – Silver Chloride
AGE – Advanced Glycation End Product
AIHW – Australian Institute of Health & Welfare
APOE – Apolipoprotein E gene
ARC – Arcuate Nucleus
b – Beta
BBB – Blood-Brain Barrier
BGL – Blood Glucose Level
BMI – Body Mass Index
BP – Blood Pressure
BPM – Beats per Minute
CBF – Cerebral Blood Flow
CGM – Continuous Glucose Monitoring
CKD – Chronic Kidney Disease
CNS – Central Nervous System
CPP – Cerebral Perfusion Pressure
CSF – Cerebrospinal Fluid
CVD – Cardiovascular Disease
xxiii
DBP – Diastolic Blood Pressure
DC – Direct Current
DCCT – Diabetes Control and Complications Trial
DL – Dorsolateral
DM – Diabetes Mellitus
DPP4i – Dipeptidyl Peptidase-4 Inhibitor
DSM – Diagnostic & Statistical Manual of Mental Disorders
ECoG – Electro-Corticography
EEG – Electroencephalography
EOAD – Early Onset Alzheimer’s Disease
EOG – Electro-oculogram
ERP – Event Related Potential
fAD – Familial Alzheimer’s Disease
FFT – Fast Fourier Transform
FPG – Fasting Plasma Glucose
fMRI – Functional Magnetic Resonance Imaging
GLUT-1 – Glucose Transporter 1
HbA1C – Glycosylated Haemoglobin
HR – Heart Rate
HR – Hazard Ratio
HREC – Human Research Ethics Committee
HTN – Hypertension
Hz – Hertz
IDE – Insulin Degrading Enzyme
IDF – International Diabetes Federation
xxiv
IR – Insulin Resistance
KW – Kilo-ohms
LAQ – Lifestyle Appraisal Questionnaire
LTP – Long-Term Potentiation
MANCOVA – Multiple Analysis of Covariance
MCI – Mild Cognitive Impairment
mm Hg – Millimetres of mercury
mmol/L – Millimoles per litre
MMSE – Mini-Mental State Examination
MoCA – Montreal Cognitive Assessment
MRI – Magnetic Resonance Imaging
Ms – Milliseconds
µV/s2 – Microvolts per second squared
NFT – Neurofibrillary Tangles
NRU – Neuroscience Research Unit
NSW – New South Wales
PAD – Peripheral Arterial Disease
PNS – Peripheral Nervous System
PSEN1 – Presenilin 1
PSEN2 – Presenilin 2
PVN – Paraventricular Nucleus
RCT – Randomised Controlled Trial
ROS – Reactive Oxygen Species
RR – Relative Risk
r – rho value
xxv
sAD – Sporadic Alzheimer’s Disease
SBP – Systolic Blood Pressure
SD – Standard Deviation
SGLT2i – Sodium Glucose Co-Transporter-2 Inhibitor
SNR – Signal-to-Noise Ratio
SVD – Small Vessel Disease
T1DM – Type 1 Diabetes Mellitus
T2DM – Type 2 Diabetes Mellitus
UKPDS – United Kingdom Prospective Diabetes Study
UTS – University of Technology Sydney
VaD – Vascular Dementia
VCI – Vascular Cognitive Impairment
VMPFC – Ventromedial Prefrontal Cortex
VMH – Ventromedial Hypothalamus
WAIS-R – Weschler Adult Intelligence Scale Revised
WHR – Waist-Hip Ratio
WMH – White Matter Hyperintensities
> – greater than
≥ – greater than or equal to
< – less than
≤ – less than or equal to
xxvi
IX. Abstract
Diabetes mellitus (DM) (Type 1 (T1DM) and Type 2 (T2DM)) and hypertension (HTN) are
associated with subtle cognitive dysfunction; however, few studies have explored the cognitive
and electroencephalography (EEG) changes that occur in these conditions. The present cross-
sectional study assessed cognitive performance (global and domain-specific) in clinical
(T1DM, T2DM, and HTN) and non-clinical samples using established cognitive assessments
and EEG, and investigated their associations with blood pressure (systolic (SBP) and diastolic
(DBP)) and blood glucose level (BGL).
Results were obtained from 94 study participants divided into four groups: non-clinical (n =
49), T1DM (n = 13), T2DM (n = 17), and HTN (n = 15). The experimental protocol was
commenced by obtaining pre-study BP measurements and a BGL measurement. Participant
lifestyle factors and disease-specific variables (e.g. HbA1c, age of disease onset, etc.) were
obtained using the Lifestyle Appraisal Questionnaire (LAQ) and disease-specific
questionnaires, respectively. Brain activity was then measured using a 32-channel EEG over
two five-minute study phases (baseline (quiet sitting) and active (Stroop Test)). Subsequently,
two reliable and validated cognitive screening tools were administered, the Mini-Mental State
Examination (MMSE) and the Cognistat. The study was concluded with post-study BP
measurements and a BGL measurement.
No significant difference was found in global or domain-specific cognitive performance

between the groups. In the non-clinical group, post-study BGL was inversely associated with
total MMSE score (p < 0.05; r = - 0.32). In the T1DM and T2DM groups, higher BGL was
significantly associated (p < 0.05) with theta activity in anterior brain regions, while
glycosylated haemoglobin (HbA1C) and disease duration were found to be significantly
associated (p < 0.05) with slow-wave oscillations. In the HTN group, higher SBP and DBP was
significantly associated (p < 0.05) with slow-wave activities over central and parietal brain
areas.
xxvii
These findings provide novel insight into the associations between blood pressure (SBP and
DBP) and BGL and EEG activity in non-clinical and clinical groups. The data obtained suggest
that the EEG can consistently detect changes in oscillatory brain activity linked to small changes
in BP and BGL, identifying the EEG as a potential neurophysiological instrument for early
screening for the subtle changes in cognition linked to both DM and HTN. Future use of EEG
as a screening tool could avert adverse cognitive outcomes linked to these chronic diseases,
such as Alzheimer’s disease (AD) and dementia, and help reduce the substantial socioeconomic
and emotional burden associated with them.
xxviii
Chapter 1.
1. Introduction
1.1 Ageing and Health

The health of our ageing population is a major and pressing socioeconomic issue.
Significant advances in medicine, including but not limited to disease diagnosis and
prevention, developments in medical technologies, improved accessibility to treatment,
novel pharmacological interventions, and improved understanding of risk factors, have
resulted in individuals worldwide, on average, living longer (Figure 1.1). In Australia,
life expectancy at birth has increased considerably since the late 1800s, with Australians
now living approximately 30 years longer (males: 33.2 years; females: 33.8 years)
compared to a century ago (Australian Institute of Health & Welfare, 2018).
Life expectancy at birth (years)

90 Females
80
70 Males
60
50
40
30
20
10
0
1886 1896 1906 1916 1926 1936 1946 1956 1966 1976 1986 1996 2006 2016
Year
0.;8, &@=<<OG<:K8E:P9P9@IK?P<8I8E;J<O=IFD [ &@=<<OG<:K8E:P
for both sexes in Australia has increased gradually over the last century.
Adapted and modified from Australian Institute of Health & Welfare
(AIHW), (2018, p 10).
1
Chapter 1.
As longevity has increased worldwide, this has simultaneously increased demands and
strain on health care systems. The natural human ageing process is accompanied by
progressive, irreversible changes in physiological function (AIHW, 2018), including
gradual hearing and visual loss, reduced mobility, and increased frailty (AIHW, 2018).
One other irreversible physiological change associated with increasing age is age-related
cognitive decline. This often manifests as Alzheimer’s disease (AD), the most common
form of cognitive impairment (defined as performance in neuropsychological assessment
at 1.5 - 2 SDs (standard deviations) below the normative mean), although various other
forms of age-related cognitive decline exist (Citron, 2010; Nordberg, 2015; Biessels &
Despa, 2018). Currently, 3.9 million people are classified as older Australians (> 65 years
old) (AIHW, 2018) (Figure 1.2) and, by 2056, projections suggest that Australia’s ageing
population will increase twofold, rising to approximately 8.7 million (AIHW, 2016). By
2096, estimates predict the number of older Australians will reach 12.8 million (Figure
1.3). As neurodegenerative diseases and pre-symptomatic stages of cognitive impairment
typically begin manifesting in older Australians (~65 years old) (although they may
manifest earlier), this trend in population ageing poses a challenging socioeconomic issue
(Hampel & Lista, 2016; Elahi & Miller, 2017; Biessels & Despa, 2018).
2
��

-
-
-
-
-
-
-
-
-

"#)%
Figure 1.2. Estimated number of Australians, categorised by age group, in 2018.

��
Adapted from AIHW, (2018, p 7).
Figure 1.3. Predicted number of Australian individuals aged 65 years and over,
categorised by age group, between the period 2016 - 2096. The number of
individuals aged 65 years and over is expected to double by 2056. Adapted
from Australian Institute of Health & Welfare (AIHW), (2016, p 17).
3
Chapter 1.
1.2 Cognition
The adult human brain – an integral component of the central nervous system (CNS) – is
a complex organ comprised of billions of metabolically-active and nutrient-dependent
brain cells (neurons) that exert precise control over all behavioural and physiological
responses (Sweeney et al., 2018). It is commonly referred to as the “control-centre”, and
plays crucial roles in the precise regulation of physiological parameters, including blood
pressure (BP), blood glucose level (BGL), heart rate (HR), and body temperature
(Herculano-Houzel, 2009; Tau & Peterson, 2010; Ando et al., 2011; Grayson, Seeley, &
Sandoval, 2013). Glucose- and oxygen-dependent neurons interspersed throughout the
brain communicate with neighbouring cells to perform high-order functions, such as
decision-making and reasoning. Collectively, these mental processes are referred to as
cognitive function (Tau & Peterson, 2010). Interestingly, although the brain accounts for
approximately 2% of total body weight, it is the most metabolically-demanding organ,
consuming roughly 20% of the body’s glucose needs (Kisler et al., 2017; Sweeney et al.,
2018). It also relies on a continuous, uninterrupted supply of blood, consuming
approximately one-fifth of total blood supply (Kisler et al., 2017; Dementia Australia,
2020).
One broad, descriptive catch-all term commonly used to describe functions associated
with the brain is cognition: mental processes of attention, perception, thinking, learning,
and memory, which support various aspects of everyday behaviour (Holden, 2011;
Spyridaki et al., 2016; Biessels & Despa, 2018). Cognition is primarily controlled by
prefrontal cortical regions (the dorsolateral (DL) and ventromedial prefrontal cortex
(VMPFC)) and influences mental alertness, workplace productivity, conscious decision-
making ability, and personal well-being (Wood & Grafman, 2003; Frederick, 2005; Ando
et al., 2011). It also plays an integral role in supporting disease self-management tasks,
such as, for example, ongoing monitoring of blood glucose concentrations in patients with
diabetes mellitus (Wood & Grafman, 2003; Frederick, 2005; Ando et al., 2011; Biessels
et al., 2020). Therefore, preservation of optimal cognitive function for as long as possible,
without pharmacotherapy or alternative cognitive strategies, is essential. Cognition may
also be divided into several different domains, including attention (filtering of specific
stimuli), memory (retention and recall of information), language (understanding,
repeating, and vocalising individual words and sentences), visuospatial ability
(processing visual information and reproducing drawings), and executive function (high-
4
Chapter 1.
order cognitive processes that orchestrate goal-directed behaviours) (Biessels & Despa,
2018; Viggiano et al., 2020). Prior to describing the functions linked to specific brain
areas, a basic understanding of gross brain anatomy is warranted.
The adult human brain comprises three major divisions: the cerebrum, cerebellum, and
the brainstem. The cerebrum, which is highly folded, consists of two cerebral
hemispheres – left and right – interconnected by the corpus callosum, a thick band of
white matter connective tissue. It plays general roles in basic sensory and motor
information processing and the regulation of skeletal muscle contractions (Martini et al.,
2011). An area central to understanding and controlling speech, the Broca’s area, resides
in the left temporal area (Martini et al., 2011). Posterior to the cerebrum lies the
cerebellum, colloquially referred to as the second brain, a structure crucial for motor
command modulation and balance (Buckner, 2013). The cerebellum is divided into right
and left hemispheres, termed left and right cerebellar hemispheres, respectively.
Anchored to the cerebellum is the brain stem, which contains relay centres and nuclei
critical for regulating autonomic functions, such as BP and HR, as well as tracts and
nuclei that assist in the maintenance of consciousness (Martini et al., 2011). Damage or
lesions to this area of the brain can result in unconsciousness (Martini et al., 2011).
Blanketing the cerebrum is a thin, superficial layer of grey matter tissue known as the
cerebral cortex. The cerebral cortex consists of four major lobes: frontal, parietal,
occipital, and temporal (Figure 1.4). Each lobe controls and performs general functions
essential to everyday activities, although some functions are solely isolated to specific
lobes (Martini et al., 2011). For example, the temporal lobe contains auditory processing
centres (Herschel’s area) and brain structures involved in long-term memory storage,
formation and retrieval, and stress modulation (e.g. the amygdala and the hippocampus)
(Martini et al., 2011). Visual processing and colour perception is processed
predominantly by the occipital lobe. Increased activity in the parietal lobe has been linked
to visuospatial functioning and somatosensory processing (e.g. touch and pain), whereas
high-order cognitive functions, such as decision-making and executive function, are
chiefly subserved by the frontal lobe (Martini et al., 2011). Although specific functions
are performed by discrete brain areas, it is important to acknowledge that cognitive
function results from continuous communication between various brain regions.
5
Chapter 1.
Figure 1.4. Lateral diagrammatic view of the cerebral cortex highlighting the different
lobes of the brain as well as other landmark anatomical features. Each lobe
(colour-coded) controls and performs specific functions essential for
everyday activities. Adapted from Martini et al., (2011, p 444).
Various pathologies and modifiable (e.g. lifestyle) and non-modifiable (e.g. age, genetics,
race) risk factors, including but not limited to ageing, prolonged stress, head trauma,
genetic predisposition, infection, diabetes mellitus, hypertension, chronic alcoholism,
tobacco smoking, and high cholesterol, have been reported to directly or indirectly
contribute to the onset and progression of cognitive dysfunction, leading to early
development of cognitive impairment (Arnsten, 2009; Obisesan, 2009; Novak & Hajjar,
2010; Sims-Robinson et al., 2013; Kivipelto et al., 2018). Current estimates suggest
cognitive impairment affects approximately 447, 115 individuals in Australia and, by
2025, is predicted to cost Australia upwards of $18.7 billion dollars (National Centre for
Social and Economic Modelling, 2017). Alarmingly, the symptoms of cognitive
impairment develop perniciously, often beginning decades prior to symptomatic
presentation and, when overt, manifest as subtle disturbances in performing daily tasks
(e.g. financial planning); consequently, accurate and timely diagnosis is challenging
(Gauthier et al., 2006; Hampel & Lista, 2016). Currently, early identification of mild
cognitive impairment (MCI), the earliest detectable stage of cognitive decline, is the most
reliable method for identifying populations at high-risk of converting to dementia
(Hampel & Lista, 2016). In clinical practice, this is typically detected using formal
neurocognitive assessments and analysing biomarkers in cerebrospinal fluid (CSF)
(Hampel & Lista, 2016).
6
Chapter 1.
1.2.1 Mild Cognitive Impairment (MCI) (Mild Neurocognitive Disorder)

The concept of mild cognitive impairment (MCI) (mild neurocognitive disorder) was
coined by Reisberg et al. (1982) to classify older individuals exhibiting impaired
cognitive function, but not meeting criteria for dementia. It describes a transitional stage
between normal cognition and cognitive decline that identifies individuals at high-risk of
developing dementia (DeCarli, 2003; Reitz et al., 2011; Hampel & Lista, 2016; Biessels
& Despa, 2018; Viggiano et al., 2020) and is defined as objective cognitive impairment
(usually 1.5 SDs below the normative mean) in one or more cognitive domains. The main
cognitive domain affected is memory (amnesic MCI), but others can also be impacted
(Biessels & Despa, 2018; Viggiano et al., 2020). While MCI typically does not interfere
with performing complex daily tasks, which remain largely intact (Novak & Hajjar, 2010;
Sachdev et al., 2014; Hampel & Lista, 2016), engagement in everyday tasks and activities
becomes increasingly effortful (Sachdev et al., 2014; Hampel & Lista, 2016).
Although MCI is widely-viewed as a precursor to dementia, and investigators agree it is

useful in identifying populations at an increased risk of developing AD (Reitz et al.,
2011), some researchers question the reliability of using MCI as a predictor of progression
to dementia. This has been ascribed to the high number of individuals (~90–95%) who
do not develop dementia (Viggiano et al., 2020) and the mixed nature of MCI, which
complicates accurate prediction of the likelihood of developing dementia (Richard &
Brayne, 2014). In agreement with this, Roberts et al. (2014), in a recent large population-
based study (n = 1939, age: 70-89 years at baseline), found reversion rates to normal
cognition are relatively high. Following a 5.1 year median follow-up period, of the 534
cases of MCI identified at baseline, approximately 36% did not convert to dementia.
Thirty-eight percent (38%) also reverted to normal cognition; however, after a later
follow-up, it was reported that over half of these patients demonstrated symptoms
fulfilling diagnostic criteria for MCI (Figure 1.5). In Australia, it is estimated that
approximately 15% of patients with MCI convert to dementia per year (Dementia
Australia, 2020). Current literature also indicates that patients with MCI have a three to
five times increased risk of developing dementia compared to age-matched controls
(Dementia Australia, 2020). Given no effective intervention strategies exist to delay the
progression of MCI to AD or dementia, and the mechanisms linking MCI to these
7
� � � ��
c o g niti v e dis e as es r e m ai n i n c o m pl et e, e arl y d et e cti o n of M CI usi n g n o n-i n v asi v e

m e as ur es b ef or e irr e v ersi bl e c o g niti v e d efi cits h a v e m a nif est e d is criti c al.
B o x 2 | Di a g n o sti c crit eri a f or mil d n e ur o c o g niti v e di s or d er

A. E vi d e n c e of m o d e st c o g niti v e d e cli n e fr o m a pr e vi o u s l e v el of p erf or m a n c e i n
o n e or m or e c o g niti v e d o m ai n s ( c o m pl e x att e nti o n, e x e c uti v e f u n cti o n, l e ar ni n g
a n d m e m or y, l a n g u a g e, p er c e pt u al – m ot or, or s o ci al c o g niti o n) b a s e d o n:
1. C o n c er n of t h e i n di vi d u al, a k n o wl e d g e a bl e i nf or m a nt, or t h e cli ni ci a n t h at
t h er e h a s b e e n a mil d d e cli n e i n c o g niti v e f u n cti o n; a n d
2. A m o d e st i m p air m e nt i n c o g niti v e p erf or m a n c e, pr ef er a bl y d o c u m e nt e d
b y st a n d ar di z e d n e ur o p s y c h ol o gi c al t e sti n g or, i n it s a b s e n c e, a n ot h er
q u a ntifi e d cli ni c al a s s e s s m e nt.
B. T h e c o g niti v e d efi cit s d o n ot i nt erf er e wit h c a p a cit y f or i n d e p e n d e n c e i n
e v er y d ay a cti viti e s (t h at i s, c o m pl e x i n str u m e nt al a cti viti e s of d ail y li vi n g s u c h
a s p ayi n g bill s or m a n a gi n g m e di c ati o n s ar e pr e s er v e d, b ut gr e at er eff ort,
c o m p e n s at or y str at e gi e s, or a c c o m m o d ati o n m ay b e r e q uir e d).
C. T h e c o g niti v e d efi cit s d o n ot o c c ur e x cl u si v el y i n t h e c o nt e xt of a d eliri u m.
D. T h e c o g niti v e d efi cit s ar e n ot b ett er e x pl ai n e d b y a n ot h er m e nt al di s or d er
(f or e x a m pl e, m aj or d e pr e s si v e di s or d er or s c hi z o p hr e ni a).
Fi g u r e 1. 5. Di a g n osti c crit eri a us e d wi d el y b y m e di c al pr a ctiti o n ers a n d cli ni ci a ns t o

di a g n os e i n di vi d u als wit h mil d n e ur o c o g niti v e dis or d er, als o k n o w n as
mil d c o g niti v e i m p air m e nt ( M CI). A d a pt e d fr o m S a c h d e v �� ( 2 0 1 4, p
6 3 7).
A c c ur at e esti m at es of t h e pr e v al e n c e of M CI, pr o g n osis, a n d d et er mi n ati o n of i n di vi d u als

at i n cr e as e d ris k of pr o gr ess i o n t o d e m e nti a is als o c o m pl e x. T his h as b e e n attri b ut e d t o
i n c o nsist e nt d efi niti o ns of M CI, v ar yi n g pr e v al e n c e r at es of M CI r e p ort e d i n dif f er nte
p o p ul ati o ns, a n d diffi c ulti es i n diff er e nti ati n g a n d d efi ni n g b o u n d ari es b et w e e n t h e
v ari o us st a g es of c o g niti v e i m p air m e nt ( �� mil d c o g niti v e i m p air m e nt, A D, a n d
d e m e nti a) ( D e C arli, 2 0 0 3; H a m p el & List a, 2 0 1 6). Oft e n, t h er e is si g nifi c a nt o v erl a p. T h e
p u blis h e d pr e v al e n c e r at es f or M CI ar e als o hi g hl y v ari a bl e. A r e c e nt p o p ul ati o n- b as e d
e pi d e mi ol o gi c al m et a- a n al ysis esti m at e d t h e gl o b al pr e v al e n c e of M CI i n a d ults o v er 6 5
y e ars t o r a n g e b et w e e n 3- 3 7 % ( S a c h d e v �� , 2 0 1 5). I n t h e s a m e m et a- a n al ysis, w hi c h
c o m bi n e d d at a fr o m 1 1 l o n git u di n al, p o p ul ati o n- b as e d, cr oss-s e cti o n al st u di es
i n c or p or ati n g t h e r e vis e d d efi niti o n f or M CI ( Fi g ur e 1. 5), t h e esti m at e d gl o b al pr e v al e n c e
r at e f or M CI w as c o nsi d er a bl y l o w er, r a n gi n g b et w e e n 6- 1 2 % ( S a c h d e v �� , 2 0 1 5).
T h er ef or e, it is i m p er ati v e t h at f ut ur e st u di es esti m ati n g t h e pr e v al e n c e of M CI i n
p o p ul ati o ns c o m pl y wit h t h e r e vis e d crit eri a pr o p os e d r e c e ntl y f or M CI t o r e d u c e
mis cl assifi c ati o n of i n di vi d u als.
8
Chapter 1.
Once diagnosed with MCI, it is hypothesised that individuals will follow one of three
hypothetical trajectories: (i) reversion to normal healthy cognitive function, (ii)
maintenance of stable MCI, or (iii) progression to dementia (impaired cognition) (Hampel
& Lista, 2016) (Figure 1.6). Given the pathophysiological processes in MCI initiate
decades prior to symptomatic presentation, and no effective disease-modifying
pharmacological intervention strategies to delay progression to AD currently exist, early
detection of cognitive decline using reliable and accurate screening tools is critical. Early
identification of individuals at increased risk of developing dementia, particularly
asymptomatic populations, would be conducive to reducing the substantial
socioeconomic costs associated with AD. Addressing prevalent modifiable lifestyle risk
factors known to exacerbate cognitive decline, such as mid-life obesity, dyslipidaemia,
hypertension, and diabetes mellitus, would also be paramount to curbing the rising global
MCI and dementia burden.
9
Chapter 1.
Figure 1.6. Proposed hypothetical trajectories in cognition that individuals follow

during normal brain ageing and in the proposed major stages of cognitive
impairment. Adapted from Hampel & Lista, (2016, p 2).
Key: AD – Alzheimer’s Disease MCI – Mild Cognitive Impairment

b – beta
10
Chapter 1.
1.2.2 Major Neurocognitive Disorder (NCD) (Dementia)

Dementia, a crippling irreversible neurocognitive disorder commonly mistaken for
Alzheimer’s disease (AD), is an ‘umbrella’ term describing a syndrome characterised by
pronounced deterioration in multiple cognitive domains (Reitz et al., 2011; Ninomiya,
2014; Hampel & Lista, 2016). It is the most severe expression of cognitive impairment,
causing a progressive loss of underlying cortical neurons, neuroinflammation and gliosis,
profoundly disrupting vulnerable cortical networks responsible for normal social and
occupational functioning (Sachdev et al., 2014; Elahi & Miller, 2017; Kivipelto et al.,
2018). This complex neurodegenerative process results in irreversible impairments in
cognition, including significant memory loss, disorientation, personality changes, and
measurable decay in cognition from a previous grade of performance (AIHW, 2018). The
strongest known risk factor for dementia is age, with approximately 90% of dementias
manifesting after the age of 65 (Elahi & Miller, 2017; ). The risk of dementia also nearly
doubles every five years between the ages of 65 to 90 (Kivipelto et al., 2018). Despite
extensive research efforts dedicated to untangling the complex aetiology of the disease,
the precise underlying pathophysiology of dementia remains elusive. Current literature
suggests the pathophysiology is multi-factorial, involving complex interplay between
lifestyle, vascular, inflammatory, and psychosocial processes (Kivipelto et al., 2018;
Sweeney et al., 2018).
Similar to the growing unsettling trend in the prevalence of MCI, the prevalence of
dementia is also rising: according to the World Alzheimer Report (2015), dementia
affected approximately 46.8 million individuals worldwide. By 2050, primarily due to
the ageing population and the increasing mean age of the population, estimates predict
dementia will affect an estimated 131.5 million individuals globally (World Alzheimer
Report, 2015; Elahi & Miller, 2017; Kivipelto et al., 2018). Alarmingly, a similar pattern
in prevalence has also been reported in Australia, with current estimates suggesting
459,000 Australians suffer from dementia (Dementia Australia, 2018) (Figure 1.7). This
translates to approximately 250 individuals developing the syndrome every day
(Dementia Australia, 2018). Unless effective pharmacotherapies are developed, or a
significant medical breakthrough occurs, by 2028, the number of individuals affected by
dementia will rise to 598,000 (Dementia Australia, 2018). Further estimates suggest this
number could rise to 1,076,000 by 2058 (Dementia Australia, 2018). It is clear that
dementia is a major, pressing socioeconomic burden requiring urgent scientific attention.
11
Chapter 1.
Figure 1.7. Current and projected number of Australians (male and female) estimated
to be affected by dementia between 2016-2056. Adapted from Alzheimer’s
Australia, (2017, p 11).
The socioeconomic burden linked to treating and managing the various forms of dementia
(AD, vascular dementia, dementia with Lewy bodies, and frontotemporal dementia), is
also substantial. According to a report published in 2003, it was estimated that dementia
cost Australia $6.6 billion (Access Economics, 2003). In 2016, costs associated with
treating dementia exceeded $14 billion (direct costs: $8.8 billion; indirect costs: $5.5
billion), translating into an astounding average of $35,550 per individual (Alzheimer’s
Australia, 2017). By 2050, it is estimated that costs associated with treating dementia
globally will exceed $1.1 trillion (Prince et al., 2013). Given the rising global prevalence
of dementia, and emerging data indicating diabetes mellitus and hypertension are
associated with an increased risk of developing cognitive impairment, research aimed at
identifying other modifiable risk factors associated with an increased risk of dementia
and incipient signs and symptoms of early cognitive decline is urgently required.
Exploration of alternative non-invasive detection methods that may facilitate reliable and
early identification of cognitive deterioration, before irreversible cognitive deficits have
occurred, would also be of equal clinical importance.
12
Chapter 1.
Dementia is hypothesised to result from a broad range of aetiologies, including

modifiable lifestyle risk factors (e.g. diabetes mellitus, hypertension, substance abuse,
and chronic alcohol intoxication), but is classified on the basis of underlying
neuropathologies (Elahi & Miller, 2017). These include abnormal misfolded protein
aggregates accumulating in vulnerable neurons and glia (Elahi & Miller, 2017). However,
accurate and timely classification of dementia and differentiation of dementia from AD
and other underlying mental disorders is difficult, requiring fulfilment of the following:
(a) a comprehensive clinical examination of the patient’s history by a medical
professional supported by anecdotal observations from a dependable informant, (b)
objective cognitive and neuropsychological assessment, and (c) fulfilment of specific
diagnostic criteria published in the American Psychiatric Association’s Diagnostic and
Statistical Manual of Mental Disorders (5th edition) (DSM-5) (Figure 1.8) (Elahi &
Miller, 2017). The DSM-5 manual, revised in 2013 to assist clinicians in diagnosing
dementia, also identifies six key cognitive domains detrimentally affected by
neurocognitive disorders (Figure 1.9).
13
Chapter 1.
Figure 1.8. Diagnostic criteria used widely by clinicians to diagnose patients with
major neurocognitive disorder/dementia. Adapted from Sachdev et al.,
(2014, p 638).
Figure 1.9. Key neurocognitive domains that are commonly affected by

neurocognitive disorders, as recognised by the DSM-5 criteria. Adapted
from Sachdev et al., (2014, p 636).
14
Chapter 1.
1.2.3 Alzheimer’s Disease (AD)

Alzheimer’s disease (AD) is a serious and irreversible neurodegenerative disease
characterised by notable memory loss, significant decay in cognitive function, and
disturbances in personality and behaviour (Baugmart et al., 2015; Elahi & Miller, 2017).
It is the most common crippling form of dementia that primarily affects older populations
(~65 years old), accounting for approximately 70-80% of dementia cases in Australia and
an estimated 50-75% globally (Citron, 2010; Nordberg, 2015; AIHW, 2018; Dementia
Australia, 2020). Advanced age is the most significant risk factor in AD development,
with the risk of AD doubling every five years after age 65 (Citron, 2010; Nordberg, 2015;
Dementia Australia, 2020). Other notable risk factors for AD include a family history of
the condition, neuroinflammation, cerebrovascular disease, traumatic brain injury,
vascular pathology, and low education (Baugmart et al., 2015; Elahi & Miller, 2017).
Worldwide, AD affects approximately 46.8 million individuals, and projections suggest
this number will increase two-fold every twenty years, rising to 131.5 million by 2050
(World Alzheimer Report, 2015). Alarmingly, no effective disease-modifying treatment
options or objective diagnostic tools to reliably and accurately detect the disease presently
exist.
Two major clinical subtypes of AD are recognised: familial AD (fAD) (early onset) and
sporadic AD (sAD) (late onset) (Baglietto-Vargas et al., 2016; Elahi & Miller, 2017;
Dementia Australia, 2020). Most AD cases are sporadic AD (~98%), but some patients
develop familial AD (<1%), an uncommon form (Elahi & Miller, 2017). While both
subtypes share similar neuropathological hallmarks (e.g. amyloid-beta (Ab) plaques,
neurofibrillary tangles (NFTs), and notable neuronal and synaptic loss), the determinants
responsible for triggering neurodegeneration differ (Baglietto-Vargas et al., 2016).
Mutations in three specific genes – amyloid precursor protein (APP), presenilin-1
(PSEN1), and presenilin-2 (PSEN2) – have been shown to promote Ab protein deposition
and have been implicated in fAD pathogenesis and early-onset AD (EOAD) (1-5% of
cases); however, the pathogenesis of sAD remains largely unknown (Citron, 2010; Van
Cauwenberghe et al., 2016; Baglietto-Vargas et al., 2016; Kandimalla et al., 2017).
Current literature suggests the aetiology is multifactorial, resulting from complex
interactions between a combination of genetic risk alleles and lifestyle factors (both
modifiable and non-modifiable) (Van Cauwenberghe et al., 2016; Baglietto-Vargas et al.,
15
Chapter 1.
2016; Kandimila et al., 2017). At the genetic level, the ε4 (epsilon) allele of the
apolipoprotein E gene (APOE) is strongly associated with an increased risk of sAD (15-
fold higher in homozygotes; three-fold higher in heterozygotes) (Farrer et al., 1997; Van
Cauwenberghe et al., 2016).
Unlike MCI, which develops with no detectable changes in brain tissue, AD is

characterised clinically by underlying neuropathological hallmarks (Figure 1.10). The
two main pathologies include: (1) the accumulation of insoluble, neurotoxic amyloid-beta
(Ab) aggregates (neuritic plaques), and (2) intracellular neurofibrillary tangles (NFTs),
caused by hyper-phosphorylated tau protein (Reitz et al., 2011; Nordberg, 2015;
Baglietto-Vargas et al., 2016; Kandimila et al., 2017). Substantial literature indicates
these pathologies contribute synergistically to and underlie the cognitive deficits
commonly reported in AD (Nordberg, 2015; Baglietto-Vargas et al., 2016; Kandimila et
al., 2017). However, convincing data obtained by Braak et al. (2011), who found NFTs
in the absence of amyloid in the brains of 2,332 patients with AD in the early stages of
disease, have challenged the established view that Ab predominantly underlies the
progressive cognitive decline in AD. Such a finding has also sparked discussion between
investigators to reconsider AD as a tauopathy, although AD is still considered by many
as a dual proteinopathy (Elahi & Miller, 2017).
16
Chapter 1.
Figure 1.10. Cross-sectional representation comparing healthy brain tissue to

Alzheimer’s disease brain tissue. Adapted from U.S Department of Health
& Human Services, (2008).
Alzheimer's disease is associated with global cortical shrinkage and perturbations in

underlying cellular and molecular processes, including neuronal loss, oxidative stress,
neuro-inflammation, disrupted cholinergic neurotransmission, gliosis, and synaptic
degeneration (Baglietto-Vargas et al., 2016; Kandimila et al., 2017; Elahi & Miller,
2017). Such disturbances and cortical atrophy, purportedly triggered by the
proteinopathies described above (Ab and NFTs), cause neurodegeneration in underlying
affected brain tissue, resulting in cognitive deficits such as memory loss and personality
changes. These deficits, which often manifest perniciously and increase gradually in
severity, correlate with the degree of neurodegeneration in underlying brain tissue
(Mormino et al., 2014).
17
Chapter 1.
Alzheimer’s disease affects several brain regions and, once diagnosed, patients on
average survive 9 years (Masters et al., 2015; Elahi & Miller, 2017) (Figure 1.11). The
neurodegenerative process is progressive: initially, the disease preferentially disrupts
vulnerable neurons and glia in brain regions responsible for episodic memory, including
the transentorhinal cortex of the medial temporal lobes, hippocampus, noradrenergic
neurons of the locus coeruleus, and basal forebrain (Elahi & Miller, 2017). Degeneration
in these brain areas causes deficits in attention and short-term episodic memory
(Apostolova & Thompson, 2007; Elahi & Miller, 2017). As AD progresses, characterised
by the gradual accumulation of abnormal neurotoxic protein aggregates (Ab) and NFTs
in the temporo-parietal cortices (Figure 1.12), it causes pronounced personality changes,
visuospatial dysfunction, disturbances in attention, acalculia, and memory loss
(Apostolova & Thompson, 2007; Elahi & Miller, 2017). By end-stage AD, patients
demonstrate significant memory loss and lack the ability to recognise familiar faces
(prosopagnosia) (Apostolova & Thompson, 2007; Elahi & Miller, 2017). They also
struggle to perform simple daily tasks, developing complete dependence on caregivers
and family members. This imposes appreciable strain on society, health care systems, and
caregivers (Citron, 2010; Nordberg, 2015).
Figure 1.11. Neuroimaging scan highlighting brain areas affected by sporadic

Alzheimer’s disease (AD) in the early stages. Atrophy in the medial
temporal lobes is a signature structural abnormality of early-stage AD.
Affected regions are highlighted in yellow. Adapted from Elahi & Miller,
(2017, p 464).
18
��
Figure 1.12. Patterns in the accumulation of neuropathologies (amyloid beta (A�)

plaques and neurofibrillary tangles (NFTs)) in Alzheimer’s disease and their
relation to disease severity. Adapted from Masters �� (2015, p 2).
One prominent issue with current diagnostic criteria for cognitive impairment criticised
repeatedly by clinicians, is that the aetiology is determined based on the nature of the
symptoms; therefore, clinicians recommend using biomarkers for determining accurately
the underlying aetiology. Although promising biomarkers of AD pathology and disease
progression have been identified in cerebrospinal fluid (CSF) (A�42: A�40; highest
diagnostic accuracy for AD), minimal progress has been made in recent years in
developing effective disease-modifying therapies against AD (Citron, 2010). To date, no
approved pharmacotherapy or disease-modifying interventions can impede the onset or
progression of AD, so future cognitive-based research is urgently required. Current
clinical treatment for AD includes acetylcholinesterase inhibitors (AChEI) together with
psychosocial support; however, these treatment options do not confer neuroprotective
benefits, providing only symptomatic improvement and delaying the progressive
cognitive decline. In 2016, dementia and AD were the second leading cause of death in
Australians, trailing closely behind coronary heart disease (AIHW, 2018) (Figure 1.13).
Unless curative or effective pharmacotherapies are developed, the devastating
socioeconomic burden of AD will continue to rise, impacting sufferers, caregivers, and
society worldwide, costing Australia an estimated $18.7 billion dollars by 2025
(Dementia Australia, 2020).
19
��

Figure 1.13. Leading causes of death in Australian males and females in 2016.
Dementia and Alzheimer’s disease were the second leading causes of
death, accounting for an estimated 11% of deaths in females. Adapted
from Australian Institute of Health & Welfare (AIHW), (2018, p 89).
In view of increased prevalence rates predicted for AD, future research must identify
biomarkers of early cognitive impairment (MCI) or the various stages of cognitive
impairment (pre-clinical, prodromal, and syndromal). This may enable early detection of
incipient cognitive decline and identification of populations at increased risk of
developing cognitive impairment long before irreversible cognitive deficits have
manifested, thus allowing early therapeutic intervention. Elahi & Miller (2017) argue
these biomarkers should reflect early-stage pathological processes, as their predictive
utility deteriorates with age. Concurrent determination of prevalent modifiable lifestyle
risk factors associated with accelerating cognitive decline, such as high cholesterol, mid-
life obesity, sedentary lifestyle, diabetes mellitus, and hypertension, would also be
advantageous. Such risk factors have been consistently linked to AD, dementia, and
another increasingly-common form of dementia, vascular dementia.
20
Chapter 1.
1.2.4 Vascular Dementia (VaD)

Vascular dementia (VaD) (also vascular cognitive impairment (VCI)) is the second most
common form of cognitive impairment, encompassing all vascular pathologies that
contribute to cognitive decline (van der Flier, et al., 2018; Iadecola et al., 2019; Dementia
Australia, 2020). It is caused by cerebrovascular events that disrupt cerebral blood flow
or damage brain tissue directly (e.g. stroke) and accounts for ~15-20% and ~15-30% of
total dementia cases in Australia and worldwide, respectively (van der Flier et al., 2018).
The resulting vascular pathologies – ischaemia-induced infarcts or white matter
hypersensitivities (WMHs) – damage the highly-vascularised network of blood vessels
of the brain, interrupting the supply of crucial oxygen, energy metabolites, and nutrients
to glucose-dependent neurons (Sweeney et al., 2018; Iadecola et al., 2019; Dementia
Australia, 2020). This causes noticeable impairment in cognition (van der Flier et al.,
2018).
Several risk factors (modifiable and non-modifiable) have been implicated in VaD
development, notably advanced age and those outlined in section 1.2. Substantial
epidemiological evidence has also revealed poorly-controlled diabetes mellitus (DM) and
hypertension significantly elevate VaD risk, with recent estimates attributing 50% of VaD
cases to hypertension (Dementia Australia, 2020). Most cases of VaD (>90%) are
classified as mixed aetiology, as pure VaD (i.e. dementia directly attributable to
cerebrovascular events) is uncommon (<10% dementia cases) (van der Flier et al., 2018).
Interestingly, emerging data indicate that patients with Type 2 diabetes mellitus (T2DM)
who have dementia demonstrate abnormalities in vasculature similar to those observed in
VaD. This finding has prompted investigators to refer to dementia in diabetes as “diabetic
dementia”, a unique form of dementia that differs from typical dementia (Morley, 2017).
VaD affects several brain regions (temporal and parietal cortices, thalamus, and basal
ganglia), with the cognitive deficits typically correlating with the location of the
underlying cerebrovascular pathology and degree of damage (van der Flier et al., 2018).
While memory and behaviour may be affected, VaD preferentially disrupts high-order
cognitive processes controlled by the frontal lobe, such as executive functioning
(planning, organising) and information processing (Garrett et al., 2004; Iadecola et al.,
2016; van der Flier et al., 2018). Similar to the neurocognitive diseases described earlier,
21
Chapter 1.
a diagnosis of VaD is also complex, requiring a comprehensive neurological examination

using established, objective neuroimaging modalities (magnetic resonance imaging –
MRI) and psychometric assessments to validate the observed cognitive disturbances
(Elahi & Miller, 2017; van der Flier et al., 2018). However, due to the limited time
available in everyday clinical practice and the complexity in performing neuroimaging,
VCI is commonly diagnosed by clinicians when evidence of vascular pathology is evident
and other potential causes of cognitive impairment have been excluded (van der Flier et
al., 2018).
Unlike the progressive degenerative nature of AD, VaD causes abrupt deterioration in
cognitive function that proceeds in a stepwise pattern following adverse cardiovascular
events such as strokes. These are the most common cause of VaD. However, impairments
in cognition may also result from subclinical vascular brain injury (Elahi & Miller, 2017).
These sharp declines in cognition cause noticeable disruptions to day-to-day social and
occupational functioning, resulting in reduced survival (approximately 3-5 years after
diagnosis) (Iadecola et al., 2016; van der Flier et al., 2018; Dementia Australia, 2020).
Although improvement in cognitive function may occur after cardiovascular events, it
typically worsens following each successive deleterious cardiovascular event (Dementia
Australia, 2020).
1.3 Risk factors for cognitive impairment

Several modifiable and non-modifiable lifestyle risk factors (Section 1.2) have been
associated, to varying degrees, with directly and indirectly triggering or contributing to
the development and progression of cognitive dysfunction. Such widely-recognised risk
factors include advanced age, high cholesterol (hypercholesterolaemia), physical
inactivity, poor diet, sedentary behaviour, mid-life obesity, low educational attainment,
and tobacco smoking (Biessels & Reagan, 2015) (Figure 1.14); however, the mechanisms
linking these specific lifestyle risk factors and cognitive dysfunction remain controversial
and poorly elucidated. Prolonged wakefulness and disrupted sleep have recently also been
suggested as a risk factor for cognitive impairment, although further evidence is required
to confirm this relationship, which falls outside the scope of this thesis (Bubu et al., 2017).
22
Chapter 1.
Concepts proposed to explain mechanisms associated with protection

against dementia:
Protective factors and

r Brain reserve
r Cognitive reserve
mechanisms
Physical, cognitive and social activity
Education
Childhood Adulthood Midlife Late-life
0 20 Age (years) 60 75
Hypertension, obesity and dyslipidaemia

Risk factors and pathogenic mechanisms
APOE, other genetic factors and familial aggregation
Unhealthy diet, alcohol misuse, smoking, diabetes mellitus and depression
Risk factor interactions and clusters:

r APOE*ε4 can magnify effects of other risk factors, including lack of physical
activity, poor diet, smoking and alcohol drinking
r People with a greater number of risk factors have an increased risk (assessed by
CAIDE score)
Mechanisms associated with dementia progression:

r Neuronal damage
r Vascular insults
r+PȱCOOCVKQP
Factors commonly associated with dementia onset in late life (>75 years of age):
r Decline in blood pressure levels
r Decline in body weight
r Decline in blood levels of lipids
r Memory complaints
Figure 1.14. Modifiable and non-modifiable risk factors for cognitive impairment
across the entire lifespan. Some lifestyle risk factors affect the risk of
developing cognitive impairment between specific time periods (e.g.
obesity, dyslipidaemia), whereas others affect dementia risk at any stage of
life (e.g. diabetes mellitus, unhealthy diet, depression). Adapted from
Kivipelto et al., (2018, p 3).
Although increasing age is universally accepted as the strongest risk factor for the
development of the various forms of cognitive impairment (Kivipelto et al., 2018),
substantial epidemiological evidence indicates that prevalent modifiable risk factors
increase and accelerate the likelihood of developing cognitive impairment. Two prevalent
modifiable lifestyle risk factors that have been consistently linked to exacerbating
cognitive decline and are associated with an increased risk of developing
neurodegenerative diseases, such as those described earlier, are diabetes mellitus and
hypertension, and these conditions will comprise a major focus of this thesis.
23
Chapter 1.
1.3.1 Diabetes Mellitus

Diabetes mellitus (DM) is a modern-day global epidemic and its debilitating nature
worldwide has been described as a “serious threat to global health that respects neither
socioeconomic status nor national boundaries” (International Diabetes Federation
Diabetes Atlas, 9th Edition, 2019, p2). It is a highly-complex, non-communicable
metabolic disorder characterised by dysregulation of glucose homeostasis (Ashcroft &
Rorsman, 2012; McCrimmon, Ryan, & Frier, 2012; Grayson, Reely, & Sandoval, 2013;
Koekkoek et al., 2015; International Diabetes Federation, 2019) (Figure 1.15). Various
modifiable (e.g. physical inactivity, overweight/obesity, poor diet, high caloric intake,
tobacco smoking, and alcoholism) and non-modifiable risk factors (e.g. age, sex, family
history, and history of gestational diabetes) have been identified, but to date the precise
underlying cause of diabetes remains elusive (Back & Kaufman, 2012; Rajagopalan, &
Brook, 2012; Szendroedi et al., 2012; Atkinson et al., 2014). Diabetes mellitus causes
transitory and sustained elevations in circulating blood glucose concentrations, a medical
sequela known clinically as hyperglycaemia, which is associated with numerous
deleterious and costly long-term complications (Section 1.4). Although pharmacotherapy
(glucose-lowering therapies) and lifestyle modification (increased physical activity,
nutritionally-balanced diet, cessation of smoking and alcohol consumption) may
minimise the severity, progression, and onset of the disease, diabetes currently remains
incurable (IDF, 2019).
24
��
Figure 1.15. Diagrammatic representation of normal homeostatic regulation of blood

glucose concentration. In both forms of diabetes mellitus, blood glucose
homeostasis is dysregulated, leading to hyperglycaemia. Adapted from
International Diabetes Federation (IDF), (2018, p 25).
25
Chapter 1.
Two major forms of DM are broadly recognised: type 1 diabetes mellitus (T1DM) and
type 2 diabetes mellitus (T2DM). Both forms are characterised by raised blood glucose
concentration (hyperglycaemia) (fasting plasma glucose (FPG) ≥ 7.0 mmol/L (millimole
per litre); two-hour plasma glucose ≥ 11.1 mmol/L)), a strong risk factor for the
development of microvascular complications (see section 1.4) (Sims-Robinson et al.,
2010; Baumgart et al., 2015; Biessels & Reagan, 2015; Koekkoek et al., 2015; DeFronzo
et al., 2017). While newer subtypes are being identified, this thesis will focus on the
established forms, T1DM and T2DM.
1.3.2 Type 1 Diabetes Mellitus (T1DM)

Type 1 diabetes mellitus (formerly insulin-dependent diabetes mellitus) most commonly
manifests in childhood (but may develop at any age) and accounts for ~5-10% of all
diabetes cases (Koekkoek et al., 2015; IDF, 2019). It is caused by the irreversible
autoimmune-mediated destruction of insulin-secreting pancreatic beta (b) cells (Sims-
Robinson et al., 2010; Biessels & Reagan, 2015; Koekkoek et al., 2015). This results in
complete insulin deficiency, leading to hypergylcaemia (Sims-Robinson et al., 2010;
Biessels & Reagan, 2015; Koekkoek et al., 2015). The irreversible destruction of
pancreatic b-cells is posited to eventuate via complex interactions between genetic risk
alleles and environmental factors; however, the precise underlying pathophysiology is
currently unclear (Harcourt et al., 2013; IDF, 2019). As insulin-secreting pancreatic b-
cells are destroyed by an autoimmune reaction, patients with T1DM require lifelong
dependency on exogenous insulin therapy and stringent ongoing monitoring of blood
glucose concentrations to maintain acceptable glycaemic control (Harcourt et al., 2013;
IDF, 2019). This imposes a considerable burden for both the patient and the numerous
allied health care professionals directly involved in the management of the disease.
Symptoms of T1DM include excessive thirst, unexplained weight loss, frequent urination,
lethargy, and constant hunger (IDF, 2019).
26
Chapter 1.
1.3.3 Type 2 Diabetes Mellitus (T2DM)

In contrast, T2DM is characterised by insulin resistance (IR), purportedly caused by a
combination of progressive pancreatic b-cell dysfunction, diminished sensitivity to
insulin, and impaired insulin secretion (Figure 1.16) (Harcourt et al., 2013; Grayson et
al., 2013; Biessels & Reagan, 2015; Koekkoek et al., 2015). Target cell (skeletal muscle,
adipose tissue, and liver) unresponsiveness to insulin leads to reduced glucose uptake and
stimulates pancreatic alpha (a) cells to secrete glucagon, stimulating hepatic glucose
production (HPG), leading to hyperglycaemia (Zheng et al., 2018). Similar to T1DM, the
pathophysiology of T2DM is currently unknown, although several well-established
lifestyle risk factors, including obesity, tobacco smoking, insufficient physical activity,
genetic susceptibility, and poor diet have been implicated in IR and exacerbating
pancreatic b-cell dysfunction (Sims-Robinson et al., 2010; Koekkoek et al., 2015). The
strongest predictor of T2DM is overweight/obesity, with weight gain in early adulthood
(25-40 years) being strongly associated with a higher risk and earlier onset of T2DM
(Riboli et al., 2002).
27
Chapter 1.
Figure 1.16. Dysregulation of blood glucose homeostasis in Type 2 diabetes mellitus

(T2DM). In individuals with T2DM, the incretin response, which is
responsible for ~70% of insulin secretion following an oral glucose load, is
diminished. This results in impaired glucose-stimulated insulin secretion,
leading to hyperglycaemia. Adapted from Grayson, Seeley, & Randoval,
(2013, p 3).
28
Chapter 1.
Most risk factors for T2DM are modifiable; therefore, T2DM is considered a largely
preventable and reversible disease (IDF, 2019). Unlike T1DM, T2DM chiefly manifests
in adulthood and accounts for approximately 85-90% of all diabetes cases worldwide
(IDF, 2019); however, due to sedentary behaviour, increased longevity, poor diet, and
trends in the ageing of the population, T2DM is increasingly being observed in younger
populations (Biessels & Despa, 2018; IDF, 2019). While T1DM and T2DM share similar
core symptoms (e.g. hyperglycaemia), recurrent fungal infections, delayed wound
healing, and numbness/tingling in extremities distinguish T2DM from T1DM (IDF,
2019).
Although T1DM and T2DM share similar pathophysiological hallmarks (impaired insulin
secretion, pancreatic b-cell dysfunction, and hyperglycaemia), their respective
epidemiology, accompanying comorbidities, and pathophysiology differ (McCrimmon et
al., 2012; Atkinson et al., 2014). Key disease characteristics, including approximate
prevalence rates and treatment options available for both forms, are summarised in Table
1.1.
29
Chapter 1.
Table 1.1. Key disease characteristics of type 1 and type 2 diabetes mellitus. Adapted
and modified from Koekkoek et al., (2015).
T1DM T2DM
Characteristic
(insulin-dependent) (non-insulin dependent)
Autoimmune disorder;
Progressive pancreatic b-cell
irreversible destruction of
Pathophysiology dysfunction, insulin
insulin-secreting pancreatic
resistance
beta (b) cells
Peaks in adulthood;
Incidence peaks during
however, increasingly
childhood (5 – 7 years) or
Period of Onset observed in younger
early adulthood but can
populations due to sedentary
occur at any age
behaviour
Prevalence (%) ~ 5-10% ~ 90-95%
Lifestyle modification
(physical activity,
nutritionally-balanced diet,
Constant glucose patient education, cessation
monitoring and lifelong of alcohol and smoking,
Treatment exogenous insulin weight loss) in combination
administration via injections with pharmacotherapy
or insulin pump therapy (metformin, DPP-4is, and
SGLT2is are the most
commonly prescribed
glucose-lowering therapies)
Key:
T1DM - Type 1 Diabetes Mellitus T2DM - Type 2 Diabetes Mellitus
DPP-4i - Dipeptidyl-peptidase 4 inhibitor
SGLT2i - Sodium Glucose Cotransporter 2 inhibitor
30
Chapter 1.
Over the last two decades, the global prevalence of diabetes mellitus has risen steadily.
Current estimates suggest diabetes mellitus affects roughly 463 million individuals
worldwide and, if current trends continue, attributable to increased life expectancy and
sedentary lifestyles, the number of affected individuals is predicted to rise to 700 million
(~20%) by 2045, with low- and middle-income populations primarily driving this trend
(Zheng et al., 2018; IDF, 2019) (Figure 1.17). An estimated 1.7 million Australians are
affected by diabetes mellitus and a further 6.1% of adults self-report having the metabolic
disorder (Diabetes Australia, 2017; Australia’s Health, 2018). Similarly, projections
indicate this number will also increase (Diabetes Australia, 2019). Alarmingly, estimates
predict that by 2023 diabetes will supersede dementia as the fastest growing chronic
disease in Australia (McCrimmon et al., 2012; Atkinson et al., 2014) (Figure 1.18).
25
20
15 2019
% 2030
10 2045
0
20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 65–69 70–74 75–79
Age groups (years)
Figure 1.17. Predicted prevalence of diabetes mellitus, categorised by age group (20 -
79 years), in 2019, 2030, and 2045. The prevalence of diabetes mellitus is
predicted to increase with increasing age. Adapted from IDF, (2019, p 37).
31
Chapter 1.
Figure 1.18. Trends in leading causes of disease and burden in Australia. Diabetes is
predicted to supersede dementia as the fastest growing chronic disease in
Australia by 2023. Adapted from AIHW, (2010, p 59).
Given the increasing prevalence rates predicted for DM globally and in Australia,
research aimed at identifying the salient risk factors that increase the likelihood of
developing diabetes is critical to reducing or preventing the deleterious complications
associated with uncontrolled or untreated DM, which typically manifest insidiously.
1.4 Diabetes Mellitus: Complications

Diabetes mellitus is a multi-factorial systemic disease, which, when poorly-managed or
uncontrolled, is associated with numerous life-threatening and debilitating complications
(Harcourt et al., 2013; IDF, 2019). The pathophysiology involves various organs
(pancreas, gastrointestinal tract, heart, liver, kidney, and skeletal muscle) and the
resulting complications – microvascular and macrovascular – have been consistently
linked to hyperglycaemia, particularly the microvascular complications (Nathan, 1993;
Vithian & Hurel, 2010; Forbes & Cooper, 2013). The microvascular complications
(retinopathy, nephropathy, and neuropathy) chiefly affect highly-vascularised organs,
whereas the macrovascular complications (myocardial infarction, stroke, peripheral
arterial disease (PAD)) impair normal blood vessel functioning (Vithian & Hurel, 2010;
Forbes & Cooper, 2013; Harcourt et al., 2013; American Diabetes Association, 2020;
32
Chapter 1.
IDF, 2019). However, intensification of glycaemic control early in the disease has been
shown to reduce and delay the development of these adverse long-term complications
(The Diabetes Control and Complications Trial, 1993; The United Kingdom Prospective
Diabetes Study, 1998), with a 1% reduction in glycosylated haemoglobin (HbA1C)
leading to a substantial decrease (37%) in all-cause mortality from diabetes-related
complications (UKPDS, 1998). While the micro- and macrovascular complications of
diabetes are well documented, it is important to recognise that diabetes is also associated
with other complications, including sexual dysfunction, autonomic neuropathy, and
depression (Mezuk et al., 2008; Kuehl & Stevens, 2012).
The substantial socioeconomic burden of diabetes can also be illustrated by considering

the rate of diabetes-related hospitalisations and deaths. Diabetes mellitus accounted for
approximately 929,000 hospitalisations between 2013 and 2014, a disturbing 9% of total
hospitalisations in Australia (Australian Institute of Health & Welfare, 2016). It was
responsible for approximately 10% of all deaths in Australia (15,100) in 2013 and
approximately 16,450 deaths in 2016, highlighting the devastating and wide-reaching
burden of diabetes (Australia’s Health, 2018). Diabetes is also the leading cause of kidney
disease in Australia, with approximately 30% and 50% of patients with T1DM and T2DM
developing kidney disease, respectively (Thomas et al., 2015).
Although the peripheral complications of DM are well established, one other under-
recognised complication of poorly-managed DM is cognitive dysfunction, manifesting as
either diabetes-associated cognitive decrements, MCI, or mixed dementia (Koekkoek et
al., 2015; Biessels & Despa 2018; Biessels et al., 2020). While documented for almost a
century, this decline in cognition is frequently overlooked and under-recognised in
standard diagnostic/medical practice. It is also commonly undetected by
neuropsychological assessment, as it progresses insidiously. However, in light of
emerging evidence suggesting diabetes exacerbates cognitive decline, the neurological
complications of diabetes are being increasingly recognised as a significant co-morbidity
(Biessels & Despa, 2018; Biessels et al., 2020) and hence will comprise a major focus of
this thesis.
33
Chapter 1.
1.5 Diabetes Mellitus and Cognitive Function

The relationship between diabetes mellitus and cognitive function has attracted
significant debate, but to date remains highly controversial. Despite rigorous studies
(cross-sectional and longitudinal) establishing a clear link between diabetes and CNS
dysfunction, significant variability and contention exists in the literature concerning the
precise cognitive modalities affected. The variability between studies has been ascribed
to various factors: methodological limitations, the diverse cognitive assessments
deployed, variation in cognitive assessment administration techniques, and investigators
failing to account for the various diabetes-related metabolic factors (e.g. diabetes
duration, glycosylated haemoglobin (HbA1c), fasting plasma glucose (FPG)). Unclear
among investigators are also the underlying pathophysiological mechanisms linking
diabetes to cognitive impairment, a rapidly-growing research area attracting considerable
exploration with potentially broader societal implications for patient management,
particularly in elderly age groups (>65 years of age) (Biessels & Despa, 2018; Biessels et
al., 2020; Srikanth et al., 2020). Most forms of cognitive impairment begin manifesting
in this age group, although it may occur earlier (e.g. early-onset Alzheimer’s disease
(EOAD)).
Several studies have assessed cognitive function in DM since Miles & Root (1922) first
observed that patients with diabetes perform worse in assessments examining memory
and attention compared to subjects without diabetes. However, the exact nature and
pattern of cognitive dysfunction in DM (T1DM and T2DM) has, to date, confounded
researchers, specifically in the case of T2DM. It is now increasingly recognised that
cognitive dysfunction is an important complication of DM (both T1DM and T2DM)
warranting urgent attention, due to the established association between diabetes and an
increased risk of cognitive impairment and the increased co-occurrence of diabetes and
cognitive impairment (Koekkoek et al., 2015; Biessels & Despa, 2018; Biessels &
Whitmer, 2019; ADA, 2020; Biessels et al., 2020; Srikanth et al., 2020). Although
guidelines have been developed to assist general practitioners in clinical practice to
address cognitive dysfunction in diabetes, some patients report that their healthcare
professionals occasionally struggle to address diabetes-related cognitive dysfunction
(Biessels & Whitmer, 2019; Srikanth et al., 2020). This has been partly attributed to a
lack of awareness of cognitive dysfunction in DM, which still reportedly lags behind that
34
Chapter 1.
of the other established peripheral complications (Biessels & Whitmer, 2019). Srikanth
et al. (2020) suggest this results in delayed identification of cognitive dysfunction.
Various terms have been suggested to describe the subtle decline in cognition triggered
by diabetes. Such terms include ‘diabetic encephalopathy’, ‘diabetes-related cognitive
dysfunction’ and, more recently, ‘diabetes-associated cognitive decline’ (Koekkoek et
al., 2015; Biessels & Despa, 2018). Currently, the modest changes in cognition linked to
diabetes are known as diabetes-associated cognitive decrements (Koekkoek et al., 2015;
Biessels & Despa, 2018; Biessels & Whitmer, 2019; Biessels et al., 2020; Srikanth et al.,
2020). These are subtle decrements in cognitive functioning in one or more cognitive
domains that often progress perniciously and affect all age groups (young adults to oldest
age (>85 years of age) (Biessels & Despa, 2018; Biessels & Whitmer, 2019; Biessels et
al., 2020; Srikanth et al., 2020). In T1DM, the decrements manifest early during the
disease and remain relatively stable over time, whereas in T2DM they are hypothesised
to develop during the pre-diabetes stage and progress insidiously as glycaemic control
worsens (Biessels & Despa, 2018; Biessels & Whitmer, 2019). Alarmingly, it has been
estimated that the decrements in cognition associated with T2DM develop approximately
50% faster than the normal ageing process (Biessels et al., 2014; Biessels & Despa, 2018).
This accelerated cognitive deterioration in T2DM is commonly referred to as “accelerated
brain ageing” and is posited to be mediated by various pathophysiological pathways
which, to date, also remain poorly-elucidated (section 1.6). While the cognitive
decrements may cause cognitive complaints (often disclosed by the patient), they are, by
definition, subtle; thus, they generally do not interfere with diabetes self-management or
daily occupational and social functioning until advanced stages (Biessels & Despa, 2018;
Biessels & Whitmer, 2019; Biessels et al., 2020). They are also typically undetected by
formal neuropsychological assessment, due to their slowly progressive nature (Koekkoek
et al., 2015; Biessels & Despa, 2018; Biessels et al., 2020). Consequently, this
complicates the ability of clinicians to establish whether individuals are affected by
diabetes-associated cognitive dysfunction (Koekkoek et al., 2015; Biessels & Despa,
2018).
35
Chapter 1.
Substantial epidemiological evidence indicates that DM is associated with an increased

risk of all types of cognitive impairment (MCI, AD, VaD, dementia). The relative risk
(RR) for all types of dementia has been documented as 1.5 to 2.5 times greater for T2DM
(Biessels, 2006; Cheng et al., 2014), whereas patients with T1DM have been estimated
to have a 65% increased risk of dementia (Smolina et al., 2015). Systematic reviews and
meta-analyses of data from more than two million participants have estimated the relative
risk (RR) for AD and VaD in diabetes as 1.53 and 2.27 greater than in individuals without
diabetes, respectively (Gudala et al., 2013; Zhang et al., 2017). Alarmingly, data from a
recent large cohort study in 2015 revealed that newly-diagnosed diabetes is associated
with an elevated risk of dementia (hazard ratio - HR 1.16). Similarly, an increased risk of
MCI (both amnesic and non-amnesic) and poorer prognosis of MCI has also been
documented in subjects with DM compared to people without DM: one study reported a
HR of 1.5 for amnesic MCI and 1.2 for non-amnesic MCI (Luchsinger et al., 2007),
whereas another documented a HR of 1.6 and 1.4 for amnesic and non-amnesic MCI,
respectively (Robert et al., 2014). In one meta-analysis examining the prognosis of MCI
in patients with diabetes, the RR of conversion to dementia was determined to be 1.7
compared to subjects with MCI, but not diabetes. Despite both DM and cognitive
impairment co-occurring frequently, and several studies having explored the association
between diabetes and dementia, the precise relation between these conditions remains
unclear.
Emerging data have recently also challenged the widely-accepted view that diabetes
causes AD-like brain changes, suggesting that patients with T2DM demonstrate vascular
abnormalities similar to those reported in VaD (Secnik et al., 2017). This novel finding
has prompted investigators to refer to dementia in diabetes as “diabetic dementia”, a
unique form of dementia that differs from typical dementia and arises from different
underlying mechanisms (Morley, 2017; Biessels & Despa, 2018) (Figure 1.19). Although
most patients with T2DM develop dementia after the age of 65 years, evidence suggests
that diabetes increases the risk of early-onset dementia (before the age of 65 years)
(Biessels & Despa, 2018; Biessels et al., 2020). However, compared to the persistent
year-by-year decline in cognition reported in dementia, diabetes-associated cognitive
decrements progress perniciously. Thus, due to dissimilar trajectories in cognitive
decline, investigators recommend considering diabetes-associated cognitive dysfunction
and dementia as distinct entities (Morley, 2017; Biessels & Despa, 2018; Biessels et al.,
36
Chapter 1.
2020). Interestingly, although both forms of diabetes share core similarities

(hyperglycaemia), the affected cognitive domains and brain areas vulnerable to insult
differ between both T1DM and T2DM. One derangement in cognition common to both
forms is reduced information processing speed (Messier, 2005; Kodl & Seaquist, 2008;
Koekkoek et al., 2015).
Figure 1.19. Possible pathophysiological pathways and contributors linking diabetes to

diabetic dementia in Type 2 diabetes mellitus, a unique form of dementia
similar to vascular dementia. Several key determinants are hypothesised to
contribute directly to diabetes-associated cognitive dysfunction, including
glycaemic events (hypo- and hyperglycaemia), advanced glycation end
products (AGEs), and insulin resistance. Adapted from Morley, (2017, p
1).
1.5.1 Type 1 Diabetes Mellitus (T1DM) and Cognitive Function

Numerous cross-sectional and longitudinal studies have consistently shown that subjects
with T1DM (children and adults) demonstrate modest yet detectable decrements (on
average half a SD [0.3-0.7 SDs]) in cognitive function (compared to age-matched
controls) across several cognitive domains, as measured by neuropsychological
assessment (Brands et al., 2005). The major cognitive modalities commonly reported to
37
Chapter 1.
be affected by T1DM include psychomotor speed, verbal fluency, general intelligence,

cognitive flexibility, and attentional performance, with decrements in these domains
having been ascribed primarily to poor glycaemic control, disease duration, and early
disease onset (before 7 years of age) (Brands et al., 2005; Brands et al., 2006; Wessels et
al., 2007; Gaudieri et al., 2008; Weinger et al., 2008; Ryan et al., 2016). While
impairments in problem-solving, vocabulary, memory, and construction have also been
documented, data supporting these observations are scarce. Occasionally, investigators
have also found no differences in cognition between T1DM patients and those without
diabetes (Lawson et al., 1984). Thus, the relationship between T1DM and cognitive
dysfunction, despite intensive exploration, remains unclear. It is also complicated by the
administration of different cognitive assessments. Geijselaers et al. (2017) argue this
results in inconsistent findings obtained between investigators and recommend deploying
similar established neuro-psychometric batteries.
Ryan et al. (2003) investigated cognitive function in adults with T1DM and age- and
education-matched healthy controls. The relationship between diabetes-related
complications and cognitive dysfunction over a 7-year period was also examined (n =
160: 103 patients with diabetes [43 males and 60 females, mean age: 40.4 ± 6.2 years],
and 57 healthy subjects [22 males and 35 females, mean age: 41.8 ± 7.1 years]). Cognitive
function was divided into three major domains: (i) learning and memory, (ii) problem-
solving and spatial ability, and (iii) psychomotor efficiency; and assessed using two
established neuropsychological assessments: the revised Wechsler Adult Intelligence
Scale (WAIS-R) (Wechsler, 1955) and Digit Vigilance Test (Lewis & Rennick, 1979).
These tests assess psychomotor efficiency and sustained attention in children and adults,
respectively. The investigators found that subjects with T1DM demonstrated significantly
worse psychomotor speed compared to non-diabetes subjects (p-value < 0.001). No
differences in performance were observed in other examined domains. The authors also
reported that microvascular complications (retinopathy and autonomic neuropathy) were
associated with exacerbating the decline in psychomotor function, a finding supported by
recent studies (Brands et al., 2005; Weinger et al., 2008).
38
Chapter 1.
While the study of Ryan et al. (2003) was a well-designed longitudinal investigation,
experimental limitations were evident. The study initially examined a large cohort, but
several subjects with diabetes were not reassessed at the 7-year follow-up period (for
undisclosed reasons), reducing the study’s statistical power. The authors also did not
account for other diabetes-related variables, such as disease duration and glycaemic
control, and this may have moderated the relationship and contributed to the accelerated
deterioration in psychomotor speed observed. The confounding effects of diabetes-related
variables are well established and several investigators argue these should be accounted
for in future studies if precise associations are to be determined (Munshi et al., 2006;
Roberts et al., 2008; Roriz-Filho et al., 2009).
Though often dismissed in cognitive investigations, the time at which cognitive function
is assessed can influence experimental data, and the time of testing was not reported by
Ryan et al. (2003). Circadian rhythm, controlled by hypothalamic suprachiasmatic nuclei,
causes changes in alertness over time and alertness typically falls between 2-4pm
(Valdez, 2019). Therefore, testing conducted during these hours poses the risk of
obtaining inaccurate data as this may yield an imprecise representation of peak cognitive
performance and a patient’s cognitive profile. The reliability coefficients/psychometric
properties for each assessment administered were also not reported. Future studies
examining cognitive function in subjects with diabetes using standardised neuro-
psychometric batteries should report the psychometric properties of all cognitive
assessments administered. Investigators should also administer cognitive screening tools
recommended by emerging clinical guidelines for screening diabetes-associated
cognitive decrements (e.g. the Mini-Mental State Examination (MMSE) (Folstein,
McHugh, & Folstein, 1975; ADA, 2020; Srikanth et al., 2020). This will improve the
likelihood of detecting potential subtle cognitive deficits and could potentially reduce the
inconsistency in reports of the cognitive domains affected by diabetes.
One of the most seminal studies published exploring the relationship between T1DM and
cognitive function is that of Brands et al. (2005), who conducted a large meta-analysis of
the impact of T1DM on cognitive function. The meta-analysis included 33 studies and
the sample population consisted of adults with diagnosed T1DM (aged >18, mean age not
reported). The association between diabetes-related metabolic variables such as disease
duration and glycaemic control, as well as the presence of complications, was also
39
Chapter 1.
assessed. Brands et al. (2005) concluded that subjects with T1DM demonstrate
significantly worse performance compared to people without DM in several cognitive
domains. Significantly worse performance was reported in seven cognitive domains:
intelligence, speed of information processing, psychomotor efficiency, sustained
attention, cognitive flexibility, and visual perception. Poor cognitive function was also
found to be strongly linked to microvascular complications rather than severe
hypoglycaemic episodes, disease duration, or poor metabolic control, a finding that both
supports and contradicts current literature (Ryan et al., 2016). Although the magnitude of
deterioration in each affected cognitive domain was modest, as determined using effect
sizes (Cohen’s d), the authors suggested that the subtle cognitive decrements could
potentially interfere with daily tasks central to diabetes self-management (e.g. monitoring
of blood glucose concentrations).
1.5.2 Type 2 Diabetes Mellitus (T2DM) and Cognitive Function

Similar to the relationship observed in T1DM, neurocognitive assessments reveal patients
with T2DM also exhibit subtle yet measurable decrements in cognition across various
cognitive domains compared to non-diabetes samples (Cohen’s d effect size [0.2 – 0.5])
(Kodl & Seaquist, 2008; Reijmer et al., 2010; Palta et al., 2014). However, unlike T1DM,
which preferentially damages brain areas involved in cognitive flexibility and
psychomotor speed, T2DM detrimentally affects brain centres associated with memory,
learning, information processing speed, and executive functioning. These cognitive
processes, excluding information processing, are typically preserved in T1DM. Modest
decrements in visuoconstructional ability, language and perception have also been
reported, although data supporting these findings are scarce (Brands et al., 2007; Ruis, et
al., 2009). The stages and severity of cognitive dysfunction in adults with T2DM may
also be divided into three approximate stages: (i) diabetes-associated cognitive
decrements, (ii) MCI, and (iii) dementia (Biessels & Despa, 2018).
While vascular complications are common to both forms of poorly-managed DM, T2DM
is frequently accompanied by various comorbidities, including obesity, depression,
dyslipidaemia, and hypertension, which have been associated with exacerbating the
cognitive decline in T2DM (Sims-Robinson et al., 2010; Reijmer et al., 2010;
McCrimmon et al., 2012). These comorbidities also often moderate strongly the
relationship between T2DM and cognition, complicating determination of the exact
40
Chapter 1.
relation between T2DM and cognition (Messier, 2005; Ferrannini & Cushman, 2012;
McCrimmon et al., 2012). The confounding effects of these comorbidities have also been
repeatedly argued to account for the diverse cognitive modalities commonly affected by
T2DM. Current literature suggests obtaining information about the various diabetes-
related moderating variables, such as glycosylated haemoglobin (HbA1C) and any
presenting comorbidities, as thoroughly as possible to address this limitation (Munshi et
al., 2006; Roberts et al., 2008; Roriz-Filho et al., 2009). In light of the increasing global
prevalence of T2DM and the progressive nature of cognitive decrements in diabetes,
preventive treatments for, and determination of the critical determinants contributing to
the progression of the subtle cognitive dysfunction in DM, are urgently needed.
Understanding the pathophysiological pathways that link DM to cognitive dysfunction
would also be crucial.
1.6 Mechanistic Contributors to Cognitive Dysfunction in

Diabetes
Several pathophysiological mechanisms have been proposed to account for the subtle
cognitive dysfunction commonly observed in patients with diabetes (T1DM and T2DM);
however, our understanding to date remains incomplete. The literature suggests the
contribution of each risk factor to cognitive dysfunction is also small (Biessels & Despa,
2018). The key causative determinants suggested to contribute to the development and
progression of diabetes-associated cognitive dysfunction are described below.
1.6.1 Hyperglycaemia
Neurons require a continuous, uninterrupted supply of glucose for optimum cognitive
functioning (McNay & Cotero, 2010; Frier, 2014). Acute disturbances in BGL cause
immediate and possibly permanent decrements in cognitive function (McNay & Cotero,
2010; Frier, 2014). Emerging evidence also suggests that fluctuations in blood glucose
concentrations in T2DM may be linked to aggravating the cognitive decrements in
diabetes and increasing the risk of dementia in late life (Rawlings et al., 2017). When
glucose concentrations remain persistently elevated (as in hyperglycaemia), glucose
neurotoxicity may ensue (Tomlinson & Gardiner, 2008). This can lead to irreversible
cellular damage and microvascular abnormalities, which accelerate the cognitive decline
and result in cognitive dysfunction. (Biessels et al., 2006). Therefore, hyperglycaemia is
41
Chapter 1.
widely considered a critical pathological determinant central to the aggravation of

diabetes-associated cognitive dysfunction, particularly chronic hyperglycaemia.
The neurotoxic consequences of hyperglycaemia are well recognised and hypothesised to

be mediated through various pathophysiological processes. Although hyperglycaemia has
been associated with interrupting cerebral blood flow (CBF) and inducing brain hypoxia
(Morley, 2017), depriving glucose and oxygen-dependent neurons of crucial nutrients,
several lines of evidence suggest hyperglycaemia exacerbates cognitive decline in
diabetes via three principal mechanisms: (i) increasing glucose flux via the polyol and
hexosamine pathways, (ii) disrupting intracellular neuronal and second messenger
pathways, and (iii) most commonly reported, triggering the formation of advanced
glycation end products (AGEs) (Biessels et al., 2006; Tomlinson & Gardner, 2008; Roriz-
Filho et al., 2009; Sims-Robinson et al., 2010; Harcourt et al., 2013; Baglietto-Vargas et
al., 2016; Morley, 2017). These complex pathophysiological processes are posited to
damage underlying cerebral tissue directly and have been associated with inducing both
microvascular and macrovascular abnormalities (Gispen & Biessels, 2000; Brownlee,
2001).
Advanced glycation end products (AGEs) are reactive substances formed by irreversible,
non-enzymatic fusions of sugars with amino groups of proteins and lipids and have been
associated with stimulating production of reactive oxygen species (ROS) (Sims-Robinson
et al., 2010; Morley, 2017; Biessels et al., 2020). Elevated quantities of AGEs and ROS
have been reported to trigger oxidative brain damage, damaging the integrity and function
of critical biomolecules such as proteins and lipids (Brownlee, 2001; Cobb & Cole, 2015).
Oxidative brain stress has been linked to activating inflammatory pathways and
increasing inflammatory cytokine production, which has been associated with stimulating
amyloid precursor protein and accelerating deposition of amyloid beta in
neurodegenerative diseases such as AD and dementia. Taken together, these processes
could potentially trigger early AD pathology or accelerate conversion to AD (Sims-
Robinson et al., 2010; Morley, 2017).
42
Chapter 1.
Further to adversely affecting the described processes, hyperglycaemia has also been
linked to perturbations in neuronal function at the cellular level. Such disturbances
include altered axonal transport, demyelination, and impaired neurotrophic support
(Tomlinson & Gardiner, 2008) (Figure 1.20). It is through these putative molecular
mechanisms that hyperglycaemia is hypothesised to contribute to the accelerated
cognitive decline and progression of cognitive dysfunction observed in diabetes patients.
Figure 1.20. Adverse molecular effects associated with glucose neurotoxicity in

neurons. At the molecular level, glucose neurotoxicity can disrupt normal
neuronal function, leading to altered axonal transport, ion channel
dysfunction, and demyelination. Adapted from Tomlinson & Gardiner,
(2008, p 42).
1.6.2 Recurrent Hypoglycaemia (RH)

Abnormally low blood glucose (plasma glucose concentration < 3.0 mmol/l) is a
common, reversible adverse effect of intensive insulin therapy commonly reported in
both T1DM and T2DM patients (Frier, 2014). It is classified clinically according to
whether an individual can self-treat (mild) or not self-treat (severe) and has also been
associated with deficits in cognitive function (Frier, 2014). It is estimated that patients
with T1DM experience one to two episodes of mild hypoglycaemia per week, whereas
those with T2DM experience 0.3-0.7 episodes per week (Ostenson et al., 2014). However,
accurate retrospective recall of hypoglycaemic episodes is poor; severe episodes can be
accurately remembered up to one year, whereas mild episodes can only be reliably
recalled for up to one week (Frier, 2014). This complicates ascertainment of the precise
43
Chapter 1.
relationship between hypoglycaemia and cognitive function, but it is clear that

hypoglycaemia is an important risk factor for diabetes-associated cognitive dysfunction.
The cognitive domains affected by moderate and severe hypoglycaemic episodes are
predominantly those associated with frontal lobe function, such as short-term memory
and reaction time (Frier, 2014). Alarmingly, it has been reported that complete cognitive
recovery in these domains following the return to normoglycaemia may not occur for
approximately 60 minutes (Zammitt et al., 2008). Although the short-term neurological
sequelae of acute hypoglycaemia are well understood, the literature is unclear as to
whether recurrent hypoglycaemia causes long-term, irreversible cognitive deterioration.
Mixed findings have been obtained: some studies have reported permanent cognitive
disturbances, substantiated using neuroimaging modalities (Chalmers et al. 1991),
whereas others have observed no association (Bruce et al., 2009). These conflicting
results have been ascribed to difficulty in determining and controlling for patient
glycaemic history as well as the many diabetes-specific variables such as disease
duration, hypoglycaemic episodes, underlying micro- or macrovascular complications,
and pre-existing comorbidities (McNay & Cotero, 2010). Frier (2014) argues the
relationship is age-dependent.
The literature is not particularly helpful in elucidating the mechanisms through which
hypoglycaemia, and particularly recurrent hypoglycaemia, may induce cognitive
dysfunction in diabetes. While disturbances in cognition caused by acute hypoglycaemia
are understood to result directly from glucose deprivation, impairing glucose-sensitive
hippocampal and cortical brain areas, the pathophysiological processes mediated by
recurrent hypoglycaemia remain elusive (Languren et al., 2013). Emerging data from
animal studies indicate recurrent hypoglycaemia exacerbates brain oxidative damage,
causing irreversible neuronal death and leading to cognitive dysfunction (Languren et al.,
2017). Support for this view is provided by Languren et al., (2017), who observed that
moderate recurrent hypoglycaemia, after an episode of severe hypoglycaemia, over seven
days, aggravated brain oxidative damage in three-month-old male Wistar rats (280-300g).
Similarly, Won et al. (2012b) also observed brain oxidative damage, indicated by
lipoperoxidation of 4-hydroxynonenal, in the rat hippocampal CA1 dendritic layer. Taken
together, these data suggest that recurrent hypoglycaemia may contribute to cognitive
dysfunction by inducing brain oxidative damage, specifically in the hippocampus, instead
of directly causing neuronal death.
44
Chapter 1.
Counterregulatory failure, a maladaptive homeostatic response that prevents timely

detection of falling blood glucose concentrations and coordination of appropriate counter-
regulatory responses, has also been proposed (Sprague & Arbelaez, 2011). The brain
contains specialised neuronal populations (glucose-excitatory and glucose-inhibitory)
that play crucial roles in maintaining glucose homeostasis and coordinating the
counterregulatory response (Roh et al., 2016). Evidence from animal studies indicates
that sustained hypoglycaemia diminishes the sensitivity of these specialised glucose-
sensing and glucose-inhibiting neurons of the ventromedial hypothalamus (VMH),
arcuate nucleus (ARC), and paraventricular nucleus (PVN) responsible for initiating the
counterregulatory response (Song & Routh, 2006). This finding has led researchers to
suggest that recurrent hypoglycaemia may cause defects in the counterregulatory
response, such as alteration of the threshold for the onset of the counterregulatory
response and blunting of the counterregulatory response to subsequent hypoglycaemic
episodes (Cryer, 2006; Beall et al., 2012; Languren et al., 2017). Conversely, other
researchers argue that moderate recurrent hypoglycaemia, paradoxically, provides a
beneficial adaptive response, shielding the brain against the detrimental effects induced
by severe hypoglycaemia (Puente et al., 2010). Therefore, it can be seen that the
mechanisms underlying hypoglycaemia-associated cognitive dysfunction in diabetes
remain controversial and warrant further exploration to clarify the exact
pathophysiological mechanisms induced by recurrent hypoglycaemia.
1.6.3 Altered Insulin Signalling

Accumulating evidence indicates that defective insulin signalling may also contribute to
the development and progression of diabetes-associated cognitive dysfunction and may
accelerate AD pathology (Sims-Robinson et al., 2010; Cholerton et al., 2013; Baglietto-
Vargas et al., 2016). In fact, AD has been referred to as Type 3 diabetes or an insulin-
resistant brain state (de la Monte & Wands, 2008; Sims-Robinson et al., 2010), suggesting
that both diabetes and AD share similar pathophysiological pathways.
45
Chapter 1.
The brain contains numerous insulin receptors interspersed throughout several key CNS
areas, notably hippocampal and cortical regions (Biessels et al., 2006; Cholerton et al.,
2013; Biessels & Reagan, 2015). Although the brain was once considered an insulin-
independent organ, it is now understood that insulin in the brain plays crucial roles in
influencing memory and learning (Cholerton et al., 2013; Biessels & Despa, 2018).
Insulin plays central roles in the maintenance of synaptogenesis and long-term
potentiation (LTP), the latter an important process for memory formation. Insulin in the
brain has also been linked to influencing the activity of major excitatory neurotransmitters
implicated in cognition, specifically acetylcholine and norepinephrine (Kopf & Baratti,
1999).
Impaired insulin signalling in both forms of diabetes has been associated with
disturbances in cognition (Sims-Robinson et al., 2010). In T1DM, chronic brain insulin
deficiency blunts long-term potentiation (LTP), disrupting hippocampal and spatial
functioning (Sims-Robinson et al., 2010). In contrast, insulin resistance (IR) – a
pathophysiological hallmark of T2DM – has been associated with compensatory
hyperinsulinaemia, particularly in early stage T2DM (Biessels et al., 2006) (Figure 1.21).
Evidence exists in the literature that hyperinsulinaemia is a modifiable risk factor for
cognitive decline, attributable to the vasoactive effects of insulin (Kalmijn et al., 1995).
Support for this view is provided by Kalmijn et al., (1995), who assessed global cognitive
function using the MMSE and found that subjects without DM, but with
hyperinsulinaemia, performed worse than those with DM. Prolonged compensatory
hyperinsulinaemia has also been hypothesised to promote hyperphosphorylation of tau
and Ab deposition, causing irreversible neuronal death (Biessels et al., 2006; Sims-
Robinson et al., 2010). Interestingly, investigators have also reported hyperinsulinaemia
in patients with sAD. This raises the possibility that (i) diabetes may contribute to AD
pathophysiology, and (ii) that both diabetes and AD share similar pathophysiological
pathways.
46
Chapter 1.
Figure 1.21. Proposed mechanisms linking impaired insulin signalling in diabetes

(Type 1 and Type 2) to Alzheimer's disease. In Type 1 diabetes mellitus,
chronic insulin deficiency disrupts long-term potentiation (LTP),
impairing hippocampal function, leading to AD. In Type 2 diabetes
mellitus, insulin resistance and corresponding compensatory
hyperinsulinaemia causes hyperphosphorylation of tau and neurofibrillary
tangles, leading to irreversible neuronal death. Adapted from Sims-
Robinson et al., (2010, p 553).
Disrupted cerebral insulin signalling has also been implicated in influencing the activity
of insulin-degrading enzyme (IDE) (Sims-Robinson et al., 2010), which is primarily
responsible for the degradation of insulin in neurons and microglia. However, evidence
indicates that IDE also plays important roles in the intracellular degradation and clearance
of amyloidogenic proteins involved in the pathogenesis of AD pathology, notably
amyloid beta (Kurauti et al., 2017). Excessive insulin has been linked to stimulating
amyloid beta secretion and obstructing the extracellular proteolytic degradation of
amyloid beta by directly competing with IDE. This results in decreased clearance of
amyloid beta. Together, elevated insulin levels and reduced clearance of amyloid beta
due to altered insulin signalling are hypothesised to contribute synergistically to amyloid
beta aggregation and plaque formation. Such a pathophysiological synergistic interaction
could potentially account for and contribute to diabetes-associated cognitive dysfunction
(Sims-Robinson et al., 2010).
47
Chapter 1.
1.6.4 Blood-Brain Barrier (BBB) Dysfunction

The blood-brain barrier (BBB) is a highly-selective and protective physiological barrier
that isolates the central nervous system (CNS) from all non-neural tissue (Tomlinson &
Gardiner, 2008; Sweeney et al., 2018) (Figure 1.22). Its integrity is maintained by sealed
tight junctions, formed by continuous capillary brain endothelial cells and perivascular
astrocytic end-feet that line cerebral microvessels (Tomlinson & Gardiner, 2008;
Sweeney et al., 2018). This specialised barrier performs various critical functions,
including the stringent maintenance of the highly-regulated internal milieu of the CNS
(ionic composition, neurotransmitters, water), and the shielding of vulnerable neuronal
populations and glia from insult from potentially neurotoxic substances circulating in the
blood (Abbott et al., 2006; Tomlinson & Gardiner, 2008; Sweeney et al., 2018). Growing
evidence from clinical and experimental studies indicates that uncontrolled DM is
associated with BBB breakdown (Horani & Mooradian, 2003; Prasad et al., 2014;
Takechi et al., 2017). Breakdown of the BBB putatively permits the entry of potentially
neurotoxic substances, disturbing the highly-regulated microenvironment of the CNS.
The unrestricted influx of toxic blood-derived substances into the brain has been linked
to activating inflammatory chemicals and triggering the onset and progression of the
neurodegenerative processes (Prasad et al., 2014; Takechi et al., 2017; Sweeney et al.,
2018).
48
��

"
$

"

2GTKXCUEWNCT+5(ȯQY
2CTCXCUEWNCTȯQY "
$

"
" !

Figure 1.22. The� blood-brain barrier (BBB). It is comprised of endothelium, pericytes,

a basement membrane, and astrocytes. Breakdown of the BBB putatively
permits the unrestricted entry of plasma-derived substances into the
sensitive micro-environment of the CNS, disrupting normal neuronal
functioning. Adapted from Sweeney �� (2018, p 135).
Disruption of the BBB in diabetes has been hypothesised to result from several possible
pathophysiological mechanisms, but the literature generally suggests that poor glycaemic
control and glycaemic events (hypo- and hyperglycaemia) are the key causative
determinants (Takechi �� 2017; Sweeney �� 2018). Glucose, which rapidly
traverses the BBB �� the insulin-dependent, facilitated glucose transport member 1
(GLUT1), is a key energy substrate for the brain; however, in dangerously low
(hypoglycaemia) and abnormally high concentrations (hyperglycaemia) it becomes
especially neurotoxic (Tomlinson & Gardner, 2008). In animal models of diabetes,
studies have reported that hyperglycaemia (acute and chronic) is associated with a down-
regulation of essential BBB glucose transporters, notably GLUT-1, decreasing glucose
uptake into the brain (Lorenzi �� 1986). Some, however, have not observed this effect.
In human studies, which are few (likely attributable to the confounding effects of common
diabetes-related metabolic variables), researchers have observed increased BBB
permeability using magnetic resonance imaging (MRI) in subjects with well-controlled
T2DM (Starr �� 2003).
49
Chapter 1.
While the molecular mechanisms underpinning how glycaemic events (hypo- and
hyperglycaemia) disrupt BBB integrity are currently unclear, recent literature suggests
they cause BBB leakiness by disrupting the continuous, end-to-end sealed tight junction
complex (Tomlinson & Gardner, 2008; Sweeney et al., 2018). This perforation of the
BBB putatively permits the entry of neurotoxic chemicals and excessive glucose, leading
to glucose neurotoxicity and neuroinflammation (Tomlinson & Gardner, 2008; Sweeney
et al., 2018). This downstream process triggers a cascade of pathophysiological
processes, including the production of ROS and oxidative stress, which have been
associated with directly inducing both microvascular and macrovascular abnormalities
and activating inflammatory molecules linked to triggering the neurodegenerative process
(Tomlinson & Gardner, 2008; Sweeney et al., 2018).
Given the pathophysiological mechanisms underlying diabetes-associated cognitive

dysfunction and pathways linking diabetes to various forms of cognitive impairment
remain unclear, there is an urgent need for objective neurophysiological measures that
can reliably and accurately detect the subtle cognitive decrements associated with
diabetes.
50
Chapter 1.
1.7 High Blood Pressure

Hypertension (HTN) (also high blood pressure, raised blood pressure, chronic arterial
hypertension, or systemic arterial hypertension)) is a highly-prevalent chronic condition
characterised by abnormally elevated blood pressure (BP) in the systemic arteries
(systolic blood pressure ³ 140mm/Hg and diastolic blood pressure ³ 90mm/Hg)
following repeated examination (Sörös et al., 2013; Iadecola et al., 2016; Oparil et al.,
2018; Unger et al., 2020). Current estimates suggest hypertension affects 1.3 billion
individuals worldwide (~31.1%) and projections indicate this number will increase in a
similar way to the prevalence patterns predicted for diabetes (Iadecola et al., 2016;
Drummond et al., 2019; Mills et al., 2020; Unger et al., 2020). It is also estimated that
3.5 billion adults globally have sub-optimal systolic BP (³ 110 – 115mm/Hg), driven
largely by non-adherence to anti-hypertensive therapy, with hypertension preferentially
affecting individuals in undeveloped geographical areas with weak healthcare systems
(Oparil et al., 2018; Unger et al., 2020). This translates to an unsettling almost one in four
adults suffering from hypertension globally (Forouzanfar et al., 2017). While raised blood
pressure (BP) predominantly affects older populations (> 65 years of age), it may also
affect young and middle-aged populations and is increasingly being reported in these
groups (Lyngdoh et al., 2013; Unger et al., 2020). Investigators ascribe this latter pattern
to sedentary behaviour, physical inactivity, and poor diet (Mills et al., 2020).
Several prevalent modifiable (e.g. tobacco smoking, alcohol intake, obesity, poor diet,
physical inactivity, and sedentary behaviour) and non-modifiable (e.g. age, ethnicity, and
gender) lifestyle risk factors have been implicated in the development of hypertension
(Figure 1.22); however, as with diabetes, the precise underlying cause of hypertension
currently remains elusive (Iadecola et al., 2016; Oparil et al., 2018; Unger et al., 2020).
The aetiology appears to be multi-factorial, resulting from a complex interplay between
genetic and environmental factors (Oparil et al., 2018; Unger et al., 2020) (Figure 1.23).
Interestingly, other researchers have suggested that hypertension results from a disruption
in normally tightly-regulated physiological processes, chiefly cardiovascular, renal, and
vascular function (Drummond et al., 2019). While various forms of hypertension have
been described (monogenic forms, treatment-resistant, Liddle syndrome), this thesis will
concentrate on the well-documented heterogeneous form, commonly referred to as
51
Chapter 1.
‘essential’ or primary hypertension. This form is generally asymptomatic and affects up

to 90% of individuals with hypertension (Bartoloni et al., 2018; Oparil et al., 2018).

1+!,.!)-%*)

Figure 1.23. Common modifiable lifestyle risk factors associated with triggering and
contributing to the development of hypertension. Adapted and modified
from Oparil et al., (2018).
Hypertension is commonly referred to as the “silent killer” and is associated with

significant premature morbidity worldwide (Mills et al., 2020). The prevalence of HTN
(both in Australia and worldwide) has also increased steadily: according to the World
Health Organisation (2018), in 2008, hypertension affected approximately 40% of
individuals aged 25 years and more worldwide. In Australia, during the period 2012 -
2013, approximately six million individuals (34%) aged 18 years and over had
hypertension and, in the same period, it was recognised as the most common chronic
disorder treated and managed by Australian general practitioners (Australian Institute of
52
Chapter 1.
Health & Welfare, 2014). Alarmingly, recent published data revealed an estimated 68%
(4.1 million) of Australians did not control or treat their hypertension (AIHW, 2016).
Although the global prevalence of hypertension decreased marginally between 1980 -
2008, estimates suggest the prevalence of hypertension will continue to increase, largely
due to changes in population trends in ageing, increased life expectancy, and modifiable
lifestyle risk factors characteristic of sedentary lifestyles described above (Weber et al.,
2014).
The significant socioeconomic burden attributable to hypertension can also be illustrated

by considering the substantial contribution of hypertension to the burden of
cardiovascular disease (CVD) and death both globally and in Australia. Worldwide,
hypertension is the leading preventable risk factor for CVD and all-cause mortality,
accounting for approximately 10.4 million deaths per year (Global Burden of Disease
Study, 2018; Mills et al., 2020). It also accounts for a staggering 12.8% of all
cardiovascular-related deaths, with vascular events directly linked to HTN responsible
for approximately 9.4 million deaths (Oparil et al., 2018). In Australia, during the period
2007 - 2008, CVD was the highest-costing disease group, costing the Australian
healthcare system approximately $7.7 billion (Oparil et al., 2018). Unless a significant
medical breakthrough occurs, or patients begin controlling their blood pressure using
effective blood pressure-lowering medications or alternative therapies, the substantial
CVD burden associated with hypertension will continue to rise, imposing considerable
strain on healthcare systems worldwide. Therefore, given hypertension is a well-
established risk factor for both CVD and chronic kidney disease (CKD), research aimed
at identifying the major determinants responsible for accelerating hypertension
development is urgently needed. The concurrent development and implementation of
appropriate risk-reduction countermeasures for known risk factors (e.g. dietary salt
reduction, etc.) to delay or prevent the development of hypertension would also be of
clinical significance.
1.7.1 Hypertension: Complications

The adverse peripheral vascular complications associated with uncontrolled hypertension
have been extensively reported (Oparil et al., 2018). Such complications include
myocardial infarction, atrial fibrillation, and stroke, with a study estimating 54% of
strokes being attributable to HTN (Lawes et al., 2008). Poorly-controlled or untreated
53
Chapter 1.
hypertension may also be both a direct cause or consequence of CKD, with HTN strongly
aggravating progression of renal failure (Bartoloni et al., 2018). One other deleterious
complication of raised BP is cognitive dysfunction, manifesting as deficits in cognitive
domains (Iadecola et al., 2016). This received little attention until observational studies
in 1960, when diminished cognitive function was observed in air traffic controllers and
pilots with high blood pressure. Despite the deleterious relationship between high blood
pressure and cognition having been documented for over half a century, as is also the case
with DM, the relation between hypertension and cognition remains highly controversial,
evidenced by conflicting data obtained from numerous studies (cross-sectional and
longitudinal) (section 1.7.2).
Similar to the decrements in cognitive function reported in diabetes, the decline in

cognition triggered by hypertension progresses insidiously, complicating timely
detection. They also often remain undetected by neurocognitive assessment (Iadecola et
al., 2016). Iadecola et al. (2016) emphasise that hypertension-associated cognitive
decline represents a significant public health challenge and argue that research elucidating
the link between BP and cognition is urgently required. Given the predicted increases in
hypertension prevalence rates, and the substantial epidemiological evidence revealing
that hypertension exacerbates cognitive decline and contributes to dementia, research
geared towards understanding the mechanisms linking hypertension to cognitive
impairment is critical. The identification of non-invasive neurological measures that can
reliably and accurately detect early hypertension-associated cognitive dysfunction would
also be of great advantage.
54
Chapter 1.
1.7.2 Hypertension and Cognitive Function

The relationship between blood pressure and cognitive function has been extensively
examined since first documented by Elias (1969), who observed reduced psychomotor
speed performance in hypertensive air traffic controllers and pilots compared to
normotensive subjects. However, despite numerous investigations (cross-sectional and
longitudinal) exploring the relationship between BP and cognition, the precise association
remains highly controversial (Birns & Kalra, 2009). The pathophysiological mechanisms
underlying hypertension-associated cognitive dysfunction (Section 1.8) and the
electroencephalography (EEG) changes that occur in hypertension also remain unclear
and largely uninvestigated; hence, they will comprise a significant focus of this thesis
(Section 1.11).
One prominent issue raised commonly by investigators exploring the relationship

between blood pressure and cognition, is that the relation is often complicated by several
factors. This has led to inconsistent findings being obtained between studies. The mixed
results between studies have been ascribed to many possible factors, including:
- the highly dynamic and variable nature of blood pressure (Schulze et al. 2000;
Franklin et al., 2013);
- potential interference from white-coat hypertension (Franklin et al., 2013);
- differences in blood pressure measurement technique (Schulze et al., 2000);
- the number of blood pressure measurements recorded (Goldstein et al., 2013);
- use and duration of anti-hypertensive treatments (Obesisan, 2009);
- different classifications of hypertension (Birns & Kalra, 2009);
- irreversible organ damage resulting directly from uncontrolled or poorly-managed
hypertension (Harrington et al., 2000);
- co-morbidity with pre-existing chronic diseases, e.g. T2DM (Ferrannini &
Cushman, 2012);
- consideration of important covariates associated with hypertension (Obesisan,
2009);
- variable participant exclusion criteria between studies (Birns & Kalra, 2009,);
- dissimilar age groups examined;
- duration of follow-up in longitudinal studies (Iadecola et al., 2016); and
55
Chapter 1.
- the significant variability in cognitive instruments administered to assess cognitive

performance (Birns & Kalra, 2009; Iadecola et al., 2016).
Although researchers recommend administering neuropsychological assessments

sensitive to cognitive domains detrimentally affected by hypertension, Birns and Kalra
(2009) argue that randomised controlled trials (RCTs) are required for robust assessment
of the association between high blood pressure and cognitive function. Sörös et al. (2013)
also argue that different participant inclusion/exclusion criteria create a significant
methodological limitation, emphasising this complicates comparability between studies.
Therefore, future research exploring the relationship between high blood pressure and
cognitive function is urgently warranted and should attempt to address the limitations
listed above as much as possible for meaningful comparisons to be made between studies.
1.7.3 High Blood Pressure and Cognition: Evidence from Cross-Sectional

Studies
Mixed findings concerning raised blood pressure and cognitive function have been
obtained in cross-sectional investigations across all age groups: mid-life (age 40 - 64
years), late-life (age 65 - 84 years) and oldest age (³ 85 years). While most studies have
reported negative associations between raised blood pressure and cognition (Starr et al.,
1993; Kilander et al., 1998; Obisesan et al., 2008), others have observed U-shaped and J-
shaped associations (Waldstein et al., 2005). Some studies have reported no association
at all (Farmer et al., 1987). Interestingly, some investigators have reported an inverse
relationship: that is, elevated blood pressure confers improvements in cognitive
performance (Launer et al., 1995). Support for this view was provided from a study
exploring BP links to cognition in centenarian Australians, which found higher systolic
BP was associated with stronger global cognitive performance (Richmond et al., 2011).
56
Chapter 1.
1.7.4 High Blood Pressure and Cognition: Evidence from Longitudinal

Studies
Although various investigators suggest longitudinal studies provide the best assessment
of the temporal relation between BP and cognition, inconsistent findings have also been
obtained across all age groups (Birns & Kalra, 2009). Most longitudinal studies have
consistently observed strong negative associations between high blood pressure and
cognitive decline (Yaffe et al., 2014); however, some have observed J- and U-shaped
associations (as reported for cross-sectional studies) (Waldstein et al., 2005).
Interestingly, some studies examining BP and cognitive function have failed to replicate
the well-documented negative association, reporting improved cognitive function with
elevated BP in the oldest age groups. This is the same unexpected outcome observed in
cross-sectional studies described above (Guo et al., 1997; Kähönen-Väre et al., 2004).
On the opposite end of the age continuum, no association between raised BP and
cognitive function has been found in younger populations (adolescents) (Lyngdoh et al.,
2013).
The most well-documented association reported in longitudinal investigations has been

between mid-life BP and cognition. Several studies have consistently demonstrated a
strong negative association between mid-life hypertension (especially high systolic BP)
and late-life cognitive impairment and dementia (Elias et al., 1993; Kilander et al., 2000;
Elias et al., 2004), but some have observed no relationship (Kesse-Guyot et al., 2015).
One study found that 10 mmHg (millimetres of mercury) increases in systolic blood
pressure (SBP) and diastolic blood pressure (DSP) in stroke-free subjects during mid-life
were linked to reduced overall global cognitive performance and poor performance in
attention and memory domains (Elias et al., 1993). Similarly, using an adjusted cognitive
model, another large-scale study found subjects with high SBP (³160 mmHg, now
classified as Grade 2 Hypertension) in mid-life had a two-fold heightened risk of
performing poorly in global cognitive measures after 25 years (Launer et al., 1995).
Another study found that pre-hypertension in both middle-aged and older women was
associated with impaired information processing and verbal memory ten years later (Chen
et al., 2015). Therefore, the relation between mid-life raised blood pressure and cognitive
function is well established.
57
Chapter 1.
In addition to the controversial relationships reported above, numerous studies (cross-

sectional and longitudinal) assessing global and domain-specific cognitive performance
using validated neuro-psychometric batteries have also demonstrated that hypertensive
subjects perform worse in global cognition and specific domains of cognition compared
to non-hypertensive subjects (Harrington et al., 2000; Lande et al., 2003; Waldstein et
al., 2005). This has been shown predominantly in commonly-administered cognitive
screening tools of global cognitive performance, such as the Mini-Mental State
Examination (MMSE) (Folstein, McHugh, & Folstein, 1975), but subtle decrements in
specific cognitive domains (executive function) have also been reported. Interestingly,
diminished cognitive function has also been observed in hypertensive subjects at the time
of cognitive assessment, irrespective of a prior diagnosis of hypertension, suggesting
short-term changes in BP detrimentally influence cognitive function (Waldstein et al.,
2005). In patients presenting with comorbidity (e.g. T2DM), worse cognitive function has
been reported, potentially indicating that both conditions contribute synergistically to
deteriorating cognitive function (Ferrannini & Cushman, 2012).
Unlike DM, which affects a diverse range of cognitive domains, the cognitive modalities
detrimentally impacted by hypertension are those controlled by frontal lobe functioning,
such as information processing and executive functioning. The latter is a complex
cognitive domain vital for adequate everyday functioning and ongoing disease self-
management (Harrington et al., 2000; Lande et al., 2003; Waldstein et al., 2005; Novak
& Hajjar, 2010; Iadecola et al., 2016). Reduced memory performance has also been
documented, although deterioration in this domain is modest and data supporting these
findings are limited. Function in other cognitive domains non-dependent on frontal lobe
functioning, such as visuospatial function and calculation, typically remain preserved.
While it is known that the cognitive decrements in DM progress perniciously and are
irreversible, it is unknown whether the cognitive decline in hypertension is reversible.
Given hypertension is widely recognised as an established modifiable risk factor for
cognitive impairment, and no effective disease-modifying treatments to delay the onset
of cognitive decline exist, research aimed at understanding the mechanistic pathways
linking hypertension to cognitive impairment is crucial for the development of future
therapies.
58
Chapter 1.
1.8 Mechanisms Underlying Cognitive Dysfunction in

Hypertension
Several pathophysiological mechanisms have been suggested to underlie hypertension-
associated cognitive decline; however, similar to findings related to diabetes-associated
cognitive dysfunction, the exact pathophysiological mechanisms that link hypertension
to cognitive impairment still remain unclear. Current literature suggests the
pathophysiology is multi-factorial, resulting from the interplay between several
pathways. The major mechanisms proposed to contribute to the development and
progression of hypertension-associated cognitive dysfunction and link hypertension to
cognitive impairment are described briefly below.
1.8.1 Blood-Brain Barrier (BBB) dysfunction

Accumulating evidence from experimental and clinical studies indicates that persistently-
elevated blood pressure disrupts the integrity of the BBB, leading to BBB breakdown and
increased permeability (Iadecola et al., 2016) (Figure 1.24). Disruption of this critical
physiological barrier disturbs the highly-regulated internal CNS milieu, resulting in
impaired neuronal connectivity, synaptic function, and information processing (Sweeney
et al., 2018). (Figure 1.24). Although the precise mechanisms underlying hypertension-
associated BBB dysfunction are unclear and probably multi-factorial, current literature
suggests that hypertension causes BBB breakdown by inducing degeneration of pericytes
and the underlying endothelium. These are both critical structural components of the BBB
(Sweeney et al., 2018). Endothelial and pericyte degeneration promotes the
destabilisation of essential tight junction proteins and adherens junctions in the vessel
wall, leading to BBB leakiness (Sweeney et al., 2018). As the BBB is compromised,
neurotoxic blood-derived factors circulating in the blood (plasminogen, thrombin,
pathogens, and Fe2+ from the breakdown of iron-containing proteins) flow unregulated
into the sensitive brain parenchyma. The uncontrolled influx of neurotoxic chemicals into
the CNS has been linked to activating inflammatory mediators, such as microglia and
astrocytes, and inducing oxidative stress, releasing harmful reactive oxygen species that
impair normal brain function. These processes have been associated with triggering and
contributing to the complex neurodegenerative cascade that eventually leads to AD and
dementia (Sweeney et al., 2018).
59
Chapter 1.
Figure 1.24. Blood-brain barrier (BBB) breakdown and its accompanying adverse
molecular effects. Disruption of the BBB results in pericyte and
endothelium degeneration, triggering a cascade of physiological responses.
Such responses, which include impaired transport, erythrocyte
extravasation, and inflammatory responses, lead to impaired CNS function
(synaptic dysfunction, irreversible neuronal injury, and loss of neurons),
causing neurodegeneration. Adapted from Sweeney et al., (2018, p 144).
1.8.2 Impaired Neurovascular Coupling

The adult human brain is dependent on a continuous, uninterrupted supply of blood,
consuming approximately one-fifth of total blood supply (Iadecola, 2004; Novak &
Hajjar, 2010; Iadecola et al., 2016; Kisler et al., 2017; Dementia Australia, 2020). Acute
and sustained interruptions in cerebral blood flow (CBF) can severely impair underlying
vulnerable nutrient-dependent brain cells responsible for cognitive function, with
irreversible neuronal damage occurring within minutes (Iadecola, 2004; Novak & Hajjar,
2010; Iadecola et al., 2016). Thus, an appropriately regulated CBF is essential for optimal
brain homeostasis (Kisler et al., 2017).
60
Chapter 1.
Neurovascular coupling (also known as functional hyperaemia) refers to ongoing cell-

cell interactions between brain cells (astrocytes and other neuroglia) and adjacent smooth
muscle and endothelial cells that function together as a single functional unit. This is
known as a neurovascular unit (Figure 1.25) (Iadecola, 2004; Novak & Hajjar, 2010;
Iadecola et al., 2016). It is a pivotal cerebrovascular mechanism that enables selective
redistribution of cerebral blood flow (CBF) to metabolically-demanding brain areas and
the simultaneous elimination of toxic metabolic by-products (Iadecola, 2004; Novak &
Hajjar, 2010; Iadecola et al., 2016). Such a dynamic and adaptive physiological process
enables the brain to continuously receive adequate quantities of crucial nutrients (O2 and
glucose) during intensive neural activation, safeguarding the highly-regulated internal
milieu of the CNS and maintaining normal cognitive functioning (Iadecola, 2004; Novak
& Hajjar, 2010; Iadecola et al., 2016).
!"&
$+& %&$"+&
'$"!
% !&
$!
Figure 1.25. A simplified representation of a neurovascular unit (NVU), which consists

of brain cells (astrocytes and other neuroglia) and adjacent smooth muscle
and endothelial cells that function together as a single unit. Adapted from
Sweeney et al., (2016, p 772).
Changes in blood pressure (hypo- and hypertension) have been associated with
interruptions in CBF, resulting in disturbances in cerebral perfusion, oxygenation, and
vascular reserve capacity (Novak & Hajjar, 2010). Convincing evidence from
experimental studies suggest that hypertension detrimentally affects the dynamic
neurovascular coupling process, disrupting cerebral blood flow to metabolically-active
cortical areas and leading to reduced cerebral perfusion, oxygenation, and vascular
reserve capacity (Novak & Hajjar, 2010). As neurons in metabolically-active regions are
deprived of crucial nutrients (O2 and glucose), brain ischaemia, neuronal dysfunction and
irreversible cellular damage will occur, which manifests as cognitive decline. Sustained
deprivation of nutrients and blood to brain cells in activated brain regions due to impaired
cerebral blood flow is hypothesised to contribute to the progression of hypertension-
associated cognitive dysfunction and link hypertension to cognitive impairment.
61
Chapter 1.
1.8.3 Small-Vessel Disease (SVD)

The adult human brain contains an estimated 644 kilometres (km) of small and large
cerebral vessels, including arteries, arterioles, capillaries, and venules that serve dual
functions: (i) delivering oxygen and nutrient-rich blood (energy metabolites) to brain cells
to maintain optimal perfusion, and (ii) eliminating neurotoxic metabolic by-products
(CO2) from the brain parenchyma to the systemic circulation (Sörös et al., 2013; Iadecola
et al., 2016; Sweeney et al., 2018). It is also the most metabolically-demanding organ,
consuming approximately 20% of the body’s oxygen; therefore, healthy blood vessels
play a central role in maintaining optimal brain homeostasis and perfusion (Iadecola et
al., 2016; Kisler et al., 2017; Sweeney et al., 2018).
Findings from brain imaging studies (MRI) suggest that exposure to persistently-elevated
BP progressively disrupts the vasculature of vulnerable cerebral blood vessels (Sörös et
al., 2013). The major arteries susceptible to early damage from chronic arterial HTN
primarily include the middle cerebral artery and the lenticulostriate arteries (Figure 1.26).
This has been ascribed to their short extension from the base of the brain (Sörös et al.,
2013). These arteries play crucial roles in supplying oxygen, energy metabolites, and
nutrients to key brain centres, including the brainstem, basal ganglia, and thalamus (Sörös
et al., 2013). Damage to these vital blood vessels causes vascular remodelling and
fibrinoid degeneration, resulting in arterial stiffness and reduced lumen diameter. Micro-
aneurysms in the vessel wall lead to a gradual narrowing (lacunar infarct) and rupturing
(intracerebral haemorrhage) of arteries (Sörös et al., 2012; Iadecola et al., 2016). This is
known as small-vessel disease (SVD) and commonly manifests clinically as
arteriosclerosis which, in advanced stages, is associated with microbleeds and thickened
vessel walls. Taken together, alterations in blood vessel integrity and diameter result in
impaired cerebral blood flow (CBF), causing hypoperfusion and insult to white matter
brain areas. This increases the risk of stroke (ischaemic and haemorrhagic). Such
pathophysiology could contribute to cognitive dysfunction and link hypertension to the
early cognitive dysfunction commonly reported in patients with hypertension.
62
Chapter 1.
Lenticulostriate
Intracerebral arteries
ar teries
haemorrhage
haemorrhage
Lacunar infarct
Rupture
of arter
ar teryy
Middle cerebral Narrowing

Narrowing
arter
ar teryy of arter
ar teryy
Figure 1.26. The major cerebral arteries hypothesised to be disrupted by untreated

hypertension. Adapted from Sörös et al., (2013, p 2).
Given prevalence rates for both diabetes and hypertension are predicted to increase, and
both conditions have been consistently linked to exacerbating cognitive dysfunction via
mechanisms currently unknown, it is clear an objective indicator and cognitive screening
tools that can accurately and consistently detect the subtle changes in cognition associated
with these conditions is urgently required. Such cognitive measures could have broader
societal implications, including the early and accurate detection of incipient cognitive
decline. They could also alert clinicians of individuals at high risk of progressing to these
cognitive diseases, enabling the instigation of robust risk-reduction measures currently
recommended in emerging guidelines to avert adverse cognitive outcomes (e.g. adequate
cardiovascular risk factor management). One objective, non-invasive neurophysiological
measure that has shown promising potential in monitoring changes and trajectories in
cognition in progressive neurodegenerative diseases and early stages of cognitive
impairment (MCI), is electroencephalography (EEG).
63
Chapter 1.
1.9 Electroencephalography (EEG)

First recorded from the scalp of humans in 1929 by German neurophysiologist, Hans
Berger, electroencephalography (EEG) is a sensitive neurophysiological technique that
records ongoing electrical brain activity generated by underlying cortical pyramidal
neurons (Da Silva, 1991; Smith, 2005; Jia & Kohn, 2011; Kaiboriboon et al., 2012;
Michel & Murray, 2012; Huster et al., 2013; Khanna et al., 2015). This electrical activity
is recorded using electrodes attached non-invasively to the scalp according to the standard
international 10-20 system of electrode placement (Jasper, 1958), a universally-adopted
and standardised system that ensures uniform scalp coverage (Figure 1.27). While the
electroencephalogram is widely considered the ‘gold-standard’ in clinical practice for
diagnosing, screening, classifying, and monitoring changes in brain activity in
neurological disorders (e.g. epilepsy, sleep disorders), it has also shown promising
potential in monitoring trajectories in cognition in progressive, neurodegenerative
diseases, such as AD and dementia (Smith, 2005; Kaiboriboon et al., 2012; Babiloni et
al., 2013; Straaten et al., 2014, McBride et al., 2014; Khanna et al., 2015).
64
Chapter 1.
Figure 1.27. The standard international 10-20 system of electrode placement. Letters
correspond to underlying brain areas. Odd numbers represent left-
hemispheric postions; even numbers, right hemispheric positions. The
electrodes circled in red are the most common sites of the 10-20 system
used in clinical practice. Adapted and modified from Mert & Akan, (2018,
p 5).
65
Chapter 1.
Although the low spatial resolution of the EEG is often criticised (Fazli et al., 2012;
Michel & Murray, 2012; Burle et al., 2015), many researchers argue the EEG
demonstrates several advantageous properties (Srinivasan, 2007; Michel & Murray,
2012; Khanna et al., 2015, Modi & Sahin, 2017). First, the EEG complements existing
cognitive assessments and neuroimaging technology, validating abnormalities observed.
The EEG is non-invasive, cost-effective, and its application is straightforward
(Srinivasan, 2007; Michel & Murray, 2012; Khanna et al., 2015; Houmani et al., 2018;
Lord et al., 2020). It is readily available, unaffected by habituation or repetitive effects,
portable and, unlike other neuroimaging modalities, does not expose patients to radiation
(Lord et al., 2020). It also provides researchers robust temporal resolution (Srinivasan,
2007; Sauseng & Klimesch, 2008; Michel & Murray, 2012; Giacino et al., 2014; Khanna
et al., 2015; Modi & Sahin, 2017). This high temporal resolution affords novel insight
into cerebral function on a millisecond scale, allowing investigators to explore brain
electrical activity non-invasively in different brain areas during cognitive tasks in real-
time (Srinivasan, 2007; Sauseng & Klimesch, 2008; Khanna et al., 2015). For these
reasons, electroencephalography was selected to assess cognitive function in the present
study.
The electroencephalogram reflects brain activity from summated postsynaptic potentials

generated by millions of cortical neurons distributed in the cerebral cortex (Sauseng &
Klimesch, 2008; Modi & Sahin, 2017). Excitation of these cortical neurons results in the
generation of distinct electrical signatures, referred to as brain waves (Da Silva, 1991;
Sauseng & Klimesch, 2008; Martini et al., 2011; Campisi & La Rocca, 2014). Berger
(1929) first observed high-amplitude alpha oscillations while examining brain activity in
resting healthy subjects and since then, several other brain waves of varying frequencies
have been described (Sauseng & Klimesch, 2008; Campisi & La Rocca, 2014; Khanna et
al., 2015). Five major brain waves can be derived from EEG tracings (Sauseng &
Klimesch, 2008; Campisi & La Rocca, 2014; Khanna et al., 2015). These are classified
as delta, theta, alpha, beta or gamma, and are associated with various psychophysiological
states (Sauseng & Klimesch, 2008; Campisi & La Rocca, 2014; Straaten et al., 2014;
Khanna et al., 2015) (Table 1.2).
66
Chapter 1.
1.9.1 Delta Waves
Delta waves are low-frequency (< 4 Hz), large-amplitude (75-200 µV) brain waves
generated by thalamo-cortical circuits (Sauseng & Klimesch, 2008; Campisi & La Rocca,
2014). Delta waves predominate the electroencephalogram recording in deep sleep (Modi
& Sahin, 2017). In cognitive studies, they have been implicated in attention processess
(Schroeder & Lakatos, 2009).
1.9.2 Theta Waves

Hypothesised to originate from thalamic and hippocampal nuclei, theta rhythms are large-
amplitude waves with a frequency range of 4-8 Hz (Sauseng & Klimesch, 2008; Massar
et al., 2014). Together with delta waves, they are collectively referred to as ‘slow-wave
activity’. Theta waves are primarily associated with drowsiness and reduced mental
alertness, but have been linked with learning and memory as well as rapid eye movement
(REM), sleep, and hypnagogic imagery (Sauseng & Klimesch, 2008; Da Rosa, &
Rodrigues, 2011; Massar et al., 2014; Modi & Sahin, 2017). Increased theta activity has
also been observed in subjects with cognitive impairment (Jelic et al., 1996; Jelic et al.,
2000).
1.9.3 Alpha Waves

First observed by Berger (1929) in resting healthy subjects, alpha rhythms are high-
amplitude, high-frequency (8-13 Hz) oscillations evident in EEG recordings during
relaxed wakefulness (Sauseng & Klimesch, 2008; Bazanova & Vernon, 2013; Campisi
& La Rocca, 2014). Alpha waves are abundant over parieto-occipital and cortical brain
areas and vanish from EEG recordings during cognitively-demanding activities, such as
attention and executive functioning (Martini et al., 2011; Campisi & La Rocca, 2014).
1.9.4 Beta Waves

Known as “fast-wave oscillations”, beta oscillations are high-frequency (14-30 Hz) brain
waves associated with mental alertness, motor activity, and an activated cortex (Sauseng
& Klimesch, 2008; Campisi & La Rocca, 2014). They are pronounced over all cortical
areas, including frontal, parietal, somatosensory and motor regions of the brain (Sauseng
& Klimesch, 2008; Campisi & La Rocca, 2014; Modi & Sahin, 2017).
67
Chapter 1.
1.9.5 Gamma Waves

Gamma rhythms are cortically-generated high-frequency (30-80 Hz) brain waves. They
are posited to underlie high-order cognitive processes, including attention, perception,
consciousness, and sensory and memory processing (Fries, 2009; Sauseng & Klimesch,
2008; Campisi & La Rocca, 2014; Ray & Maunsell, 2015; Modi & Sahin, 2017). Reduced
gamma activity has been associated with abnormalities in cognitive function, such as
cognitive impairment.
Table 1.2. The main brain waves commonly observed in an electroencephalogram and
their corresponding psychophysiological state(s). Adapted and modified
from Modi & Sahin, (2017).
Brain Wave Frequency (Hz) Psychophysiological State(s)
Delta 1–4 Deep sleep, attentional processes,
Drowsiness, fatigue, rapid eye

Theta 4-8 movement, learning and memory
processing,
Relaxed wakefulness, attention,
Alpha 8 - 12
selective processing
Mental alertness, sustained
Beta 12 - 30
concentration, motor activity
Attention, short-term memory
Gamma 30 - 100 processing, consciousness, selective
inhibition
Key:
Hz - Hertz
68
Chapter 1.
1.10 Electroencephalography Changes in Diabetes Mellitus

Research investigating electroencephalographic changes associated with diabetes
mellitus (T1DM and T2DM) is sparse. There is a lack of recent data reporting and
validating EEG changes that occur in diabetes and most studies have examined EEG
activity from limited electrode sites; consequently, the precise underlying
neurophysiological changes associated with diabetes remain unclear. Research in this
area is also predominantly exploratory, with few longitudinal investigations having been
conducted. This scarcity limits understanding of the electroencephalography changes that
occur during disease progression and the clinical utility of the electreoncephalogram.
Early published reports classify abnormal EEG activity into broad arbitrary groups (e.g.
abnormal/normal activity), limiting understanding of the precise electrophysiological
abnormalities. Many studies also fail to report and account for the various diabetes-related
mediating factors in detail, which are known to moderate the relationship between
diabetes and cognition (Munshi et al., 2006; Roberts et al., 2008; Roriz-Filho et al.,
2009). This results in conflicting data being obtained.
Evidence from early EEG studies indicate that both children and adults with DM (T1DM
or T2DM) demonstrate electrophysiological abnormalities compared to subjects without
diabetes (Greenblatt, Murray, & Root, 1946; Izzo et al., 1953; Eeg-Olofsson & Petersen
1971; Brismar et al., 2002; Hyllienmark et al., 2005). The main changes in EEG activity
commonly reported in subjects with DM include: (i) sharp increases in slow-wave brain
activities (theta and delta) and (ii) declines in fast-wave brain activities (alpha, beta, and
gamma) (Eeg-Olofsson & Petersen 1971; Brismar et al., 2002; Hyllienmark et al., 2005).
These EEG abnormalities have been primarily observed over temporo-occipital and
frontal brain areas; however, global and focal changes in alpha, beta, and theta frequency
bands have also been reported (Izzo et al., 1953; Eeg-Olofsson, 1971; Brismar et al.,
2002; Hyllienmark et al., 2005; Cooray et al., 2011a). Less reported have been modest
reductions in fast-frequency gamma activity. Several investigators suggest the
electrophysiological abnormalities observed in subjects with diabetes result from
abnormal blood glucose concentrations (Soltèz & Acsádi, 1989). This view is supported
by recent studies, showing that abnormal blood glucose concentrations influence the
electrical activity of the brain (Graveling et al., 2013; Rachmiel et al., 2016). Therefore,
ascertainment of blood glucose concentration at the time of cognitive assessment is
pivotal to mitigate potential influence from hypo- or hyperglycaemic states.
69
Chapter 1.
1.10.1 Changes in slow-wave EEG activity in diabetes mellitus

Several early EEG studies investigating neurophysiological changes have found both
children and adults with diabetes demonstrate global and focal increases in slow-wave
activity compared to healthy subjects (Greenblatt, Murray, & Root, 1946; Izzo, Schuster,
& Engel, 1953; Herrlin et al., 1962; Eeg-Olofsson & Petersen 1977). Izzo et al. (1953)
found young adults and elderly subjects with diabetes (n = 81: mean age: 50.1 years, age
range: 15-76 years) exhibited increased slow-wave activity (delta and theta) and reduced
alpha activity in frontal brain regions using a Grass, six-channel, model III C
electroencephalogram (electrode positions not reported). Herrlin et al. (1962) recorded
EEG activity from twenty-one electrode positions in young children and adults (n = 80,
age range: 16-38 years) with long duration DM (diabetes duration: 15 years) over a two-
year period and reported pronounced increases in slow-wave activity in the left parietal
region. Interestingly, Herrlin et al. (1962) found no association between BGL and
abnormal EEG activity or common diabetes-mediating factors, including diabetes
duration, onset of diabetes, and degree of metabolic control. The precise EEG electrode
positions explored by Herrlin et al. (1962) were also not reported. Similarly, Eeg-
Olofsson and Petersen (1966) reported increased theta activity in children with diabetes;
however, unlike Herrlin et al. (1962), this pattern was strongly linked to hypoglycaemia.
Supporting Eeg-Olofsson and Petersen (1966), Tsalikian et al. (1981) also observed
similar abnormalities in low-frequency EEG activity in newly- and previously-diagnosed
children with uncontrolled T1DM (n = 39; 24 boys, 15 girls, age range: 11.5 months –
16.5 years) with ketosis, suggesting poor glycaemic control causes the observed
electrophysiological abnormalities. However, Hung et al. (2010) suggest that more
reliable associations may be observed bewteen HbA1C and EEG activity than with finger
prick blood glucose tests, as blood glucose concentrations vary continuously.
The available evidence generally suggests there is a generalised slowing of EEG activity
in patients with DM (T1DM and T2DM), particularly in temporal and occipital brain
regions (Mooradian et al., 1988; Pramming et al., 1988; Tallroth et al., 1990). Mooradian
et al. (1988) reported increased slow-wave activity over the central cortex (electrode
locations: FZ, CZ, and PZ) and reductions in alpha activity in the parietal region in elderly
subjects (n = 43, mean age: 66.3 ± 0.3 years, diabetes duration: 13.3 ± 1.8 years) with
T2DM compared to age-matched controls. No relationship between blood glucose
70
Chapter 1.
concentration and EEG activity was documented. In contrast, Pramming et al. (1988)
found marked increases in theta oscillations over temporal and parieto-occipital areas in
patients with T1DM (n = 13, mean age: 28 years, diabetes duration: 8 years) during
induced hypoglycaemia. Similar to Pramming et al. (1988), Tallroth et al. (1990)
observed sharp increases in slow-wave frequency activity in T1DM subjects (n = 8, mean
age: 28.0 ± 7.4 years, duration of diabetes: 15.5 ± 5.1 years) during insulin-induced
hypoglycaemia over anterior brain regions. Interestingly, the authors reported that EEG
activity normalised after blood glucose concentrations stabilised, reinforcing that
hypoglycaemia causes transitory alterations in brain electrical activity (Deary, & Frier,
2013; Rachmiel, et al., 2015).
Other researchers have reported similar findings in subjects with DM. Brismar et al.
(2002) investigated EEG activity in young adults with well-controlled T1DM without a
history of recurrent hypoglycaemia (n = 100: 49 patients with diabetes, 51 healthy
controls, age: 21-41 years) and found pronounced increases in slow-wave brain activity:
increased delta activity in frontal and temporo-parietal areas, and elevated theta activity
in frontal and left central brain regions. During controlled hypoglycaemia, Bjorgaas et al.
(1998) similarly showed that children with T1DM (n = 19, diabetes duration: >1.5 years)
exhibit global increases in theta activity in cortical areas compared to healthy subjects
using quantitative EEG.
Conversely, a recent study by Cooray et al. (2011) reported diminished global slow-wave
power in patients with T1DM (n = 119, age range: 22-56 years, diabetes duration: > 5
years). The dissimilar outcome obtained could be due to several factors: differences in
the degree of metabolic control of diabetes participants recruited, the shorter duration of
diabetes in participants recruited by Bjorgaas et al. (1998), and potential influence from
the mediating effects of diabetes-related factors. Bjorgaas et al. (1998) also predicted the
degree of metabolic control solely from glycosylated haemoglobin (HbA1C) data
provided. Although HbA1C is widely considered the ‘gold-standard’ marker of long-term
glycaemic control, it does not reflect minute-to-minute fluctuations in BGL (Kovatchev,
2017; ADA, 2020). It is also insensitive to hypoglycaemic episodes (Kovatchev, 2017;
ADA, 2020). Hence, this may have accounted for the different outcome observed.
Importantly, the presence of electrophysiological abnormalities in children suggests the
developing human brain is vulnerable to the early neurotoxic effects of diabetes (Ryan,
71
Chapter 1.
2006; Biessels, Deary, & Ryan, 2008). However, the ability of the EEG to consistently
detect changes in cortical activity non-invasively supports the use of EEG as a potential
suitable cognitive measure for rapidly monitoring ongoing changes in brain activity in
diabetes.
1.10.2 Changes in fast-wave EEG activity in diabetes mellitus

Neurophysiological studies have also revealed that subjects with diabetes demonstrate
measurable reductions in fast-wave brain activities compared to subjects without
diabetes. One consistently-reported finding in recent studies has been diminished beta
power. This has been primarily observed over temporal brain areas (Brismar et al., 2002;
Hyllienmark et al., 2005; Cooray, Hyllienmark, & Brismar, 2011). Howorka et al. (2000)
found T1DM patients (n = 13, mean age: 36.1 ± 10.2, mean diabetes duration: 16.7 ± 7.4
years) with a recurrent history of severe hypoglycaemia demonstrated global widespread
reductions in beta power. Slowing of central beta activity was also documented. In three
more recent studies, all of which investigated EEG activity in young adults with T1DM,
the investigators reported declines in beta power compared to age-matched healthy
controls (Brismar et al., 2002; Hyllienmark et al., 2005, Cooray et al., 2011).
Interestingly, no study found an association between beta power and a history of
hypoglycaemia. Diminished beta power was also reported over different brain regions:
Brismar et al. (2002) observed this diminution over posterior temporal and occipital areas,
Hyllienmark et al. (2005) over the posterior temporal region, while Cooray et al. (2011)
reported this pronounced decline in beta power in temporal regions.
Marked reductions in upper alpha activity have also been reported, although mixed
findings have been obtained (Eeg-Olofsson and Petersen, 1966). In an early study
investigating oscillatory activity in children and young adults with DM (n = 80, mean
age: 10 years, age range: 2-16 years, diabetes duration: 4.6 years), Eeg-Olofsson and
Petersen (1966) found these patients demonstrated reduced alpha activity. However, the
location of this diminished alpha activity was not reported. No relationship was found
also between age, metabolic control, age of diabetes onset, or diabetes duration, and the
observed abnormalities in alpha activity were determined to be significantly correlated to
the frequency of hypoglycaemic comas. Three decades later, Tribl et al. (1996) observed
similar abnormalities in alpha activity during induced hypoglycaemia in adults with
T1DM (n = 14; 8 males, 6 females, mean age: 33.1 ± 8.9 years, diabetes duration: 12.8 ±
72
Chapter 1.
6.0 years, mean HbA1C: 7.2 ± 1.1 %), suggesting that glycaemic events are associated
with detectable changes in cerebral electrical activity. In a study conducted two-years
later in children during controlled hypoglycaemia using quantitative EEG, Bjorgaas et al.
(1998) obtained contradictory results, observing modest increases in alpha activity over
fronto-central and temporal regions.
One EEG frequency band largely unexplored in neurophysiological studies in diabetes is

the gamma wave. Gamma waves are high-frequency brain waves associated with high-
order cognitive processes, including short-term information processing and attention
(Engel, Fries, & Singer, 2001; Sauseng & Klimesch, 2008; Blokland et al., 2015; Ray &
Maunsell, 2015). Available evidence indicates that subjects with diabetes exhibit reduced
gamma activity (Brismar et al., 2002; Hyllienmark et al., 2005, Cooray et al., 2011).
Brismar et al. (2002) and Hyllienmark et al. (2005) found young adults with T1DM
exhibited reduced gamma power in posterior temporal brain regions. This pattern of
reduced gamma activity was similarly reproduced more recently by Cooray et al. (2011),
who found pronounced reductions in gamma power over the mid-parietal region.
However, to date, only few studies have reported such patterns. No recent investigations
have replicated this pattern in neuronal gamma oscillations in patients with DM (T1DM
and T2DM). Therefore, it is critical that future studies continue investigating high-
frequency gamma band oscillations in subjects with diabetes (Engel, Fries, & Singer,
2002). Exploration of activity in this frequency band may reveal possible disturbances in
cortical regions responsible for generating these brain waves, indicating possible early
cognitive deterioration.
Whether the duration of diabetes correlates with the severity of EEG abnormalities
observed also remains unclear. While the duration of AD is understood to correlate
strongly with the severity of EEG abnormalities, the literature is not helpful in clarifying
whether such a relationship also exists in diabetes mellitus. Mixed results have been
obtained: some early studies have reported an association (Izzo et al., 1953), whereas
others have not (Haumont et al., 1979; Soltèz & Acsádi, 1989). Soltèz & Acsádi (1989)
acknowledge the lack of an association between duration of diabetes and EEG
abnormalities could have been ascribed to the short duration of diabetes (mean duration
of diabetes: 5 years) in participants recruited in their study. Other researchers attribute
conflicting findings to arbitrary classifications of metabolic control and limited
73
Chapter 1.
information obtained by early studies regarding the various diabetes-related moderating

factors. The complex and multi-factorial nature of diabetes also complicates this relation.
Thus, it is clear from the mixed findings obtained that this area of research requires further
exploration. Table 1.3 summarises the main changes in EEG activity reported in patients
with DM (T1DM and T2DM) from the studies described above.
74
Chapter 1.
Table 1.3. Summary of main findings from studies investigating electroencephalography activity in diabetes mellitus.
Study Sample Group Main EEG Findings

Young adults and elderly patients with DM (n
↑ delta, ↑ theta (global), ↓ alpha in frontal
Izzo et al. (1953) = 81: mean age: 50.1 years, age range: 15-76
regions
years, type of diabetes: not specified)
Young children and adults (n = 80, age range:
Herrlin et al. (1962) 16-38) with long duration T1DM (diabetes ↑ delta, ↑ theta in left parietal regions
duration: 15 years)
Young children with T1DM (n = 80, mean
age: 10 years, age range: 2-16 years, diabetes
Eeg-Olofsson and Petersen (1966) ↑ theta, ↓ alpha
duration: 4.6 years) with a history of
hypoglycaemia
Newly- and previously-diagnosed children
with uncontrolled T1DM (n = 39; 24 boys, 15
Tsalikian et al. (1981) ↑ delta, ↑ theta
girls, age range: 11.5 months – 16.5 years) and
ketosis
Elderly subjects with T2DM (n = 43, mean
age: 66.3 ± 0.3 years, diabetes duration: 13.3 ± ↑ delta, ↑ theta over central regions, ↓
Mooradian et al. (1988)
1.8 years) and controls (n = 41, mean age: 65.3 alpha in parietal regions
± 0.6 years)
75
Chapter 1.
Patients with T1DM (n = 13, mean age: 28

↑ theta over temporal and parieto-
Pramming et al. (1988) years, diabetes duration: 8 years) during
occipital areas
induced hypoglycaemia
Patients with T1DM (n = 8, mean age: 28.0 ±
7.4 years) with long duration diabetes (15.5 ±
↑ delta, ↑ theta over anterior brain
Tallroth et al. (1990) 5.1 years) during insulin-induced
regions
hypoglycaemia and control (n = 12, 26.4 ± 4.2
years)
Slight hypoglycaemia: ↑ delta, ↑ theta
Patients with T1DM (n = 14; 8 males, 6 over lateral frontal regions
females, mean age: 33.1 ± 8.9 years, diabetes Hypoglycaemia: ↑ delta ↑ theta, ↓ alpha
Tribl et al. (1996)
duration: 12.8 ± 6.0 years, mean HbA1C: 7.2 ± Severe hypoglycaemia: ↑ delta ↑ theta in
1.1 %) centro-temporal and parieto-occipital
regions
Children with T1DM (n = 19, diabetes ↑ theta over cortical areas, ↑ alpha over
Bjorgaas et al. (1998)
duration: >1.5 years) fronto-central and temporal regions.
T1DM patients (n = 13, mean age: 36.1 ± 10.2,
Howorka et al. (2000) diabetes duration: 16.7 ± 7.4 years) with a ↓ beta (global)
recurrent history of severe hypoglycaemia
76
Chapter 1.
Young adults with T1DM (n = 49, 21 – 41

↓ alpha, beta, gamma in posterior and
years of age, diabetes duration: 9.4 ± 3.5 years,
Brismar et al. (2002) temporal regions, ↓ beta in occipital
mean HbA1C: 6.9 ± 1.2 %) and controls (n =
regions
51)
Young adults with T1DM (n = 35, mean
age:17.1 ± 1.7 years, diabetes duration: 7.6 ± ↑ delta, theta in frontal regions, ↓ alpha,
Hyllienmark et al. (2005) 4.6 years, age of diabetes onset: 9.6 ± 4.6 beta, gamma in posterior temporal
years, glycaemic control: not reported) and regions
controls (n = 45, mean age:16.8 ± 1.6 years)
Patients with T1DM (n = 119, mean age: 43.3
↓ beta over temporal regions, ↓ gamma
Cooray et al. (2011) ± 7.6 years, mean HbA1C: 7.3 ± 1.2 %,
over mid-parietal region
diabetes duration: 27.1 ± 11.6 years)
Key:
T1DM – Type 1 Diabetes Mellitus T2DM – Type 2 Diabetes Mellitus n – sample size
HbA1C – Glycosylated haemoglobin ↑ – increase ↓ – decrease
77
Chapter 1.
1.11 Electroencephalography Changes in Hypertension

Studies exploring neurophysiological changes in hypertensive subjects are lacking. There
are no recent studies that have investigated EEG activity in subjects with hypertension.
The longstanding neglection of this area of research may be partly due to the dilemma
that hypertensive subjects are often medicated or reliant upon medication to maintain
acceptable blood pressure. Consequently, this complicates recruiting non-medicated
patients for EEG testing. It is understood that changes in cerebral blood flow (CBF) are
associated with changes in oscillatory activity (Hossmann, 1994; Jordan, 2004). Clinical
studies have also reported that blood pressure-lowering medication can enhance cerebral
perfusion, potentially preserving cognitive function (Obisesan, 2009). Current evidence
suggests that anti-hypertensive therapy can lower the risk of dementia and AD by 12 and
16%, respectively (Ding et al., 2020), but the optimum age to initiate therapy and the
duration of antihypertensive medication required to observe such a benefit remains
unclear (Iadecola & Gottesman, 2019). No convincing data also exist to indicate the most
effective blood-pressure lowering therapy for favourable cognitive outcomes (Iadecola &
Gottesman, 2019; Ding et al., 2020).
The only available study exploring EEG activity in hypertension was that conducted by
Mani and Townsend (1964), who investigated EEG activity in subjects with clinically-
diagnosed benign intracranial hypertension (BIH) (n = 14; 7 males and 7 females, mean
age not reported) and obstructive hydrocephalus (n = 31). No study participants had a
prior history of epilepsy or cerebrovascular disease. EEG activity was recorded from eight
electrode positions (not disclosed) and categorised into five arbitrary groups (A: well-
defined alpha rhythm, little other activity, B: good alpha rhythm, slight excess of other
frequencies, C: little alpha rhythm, excess of other activity, D: dominant fast activity, E:
dominant slow activity). Burst activity was also categorised into three different gradations
(Grade 1: minimal, Grade 2: definite, Grade 3: marked). The authors found that subjects
with BIH demonstrated mostly normal EEG activity compared to subjects with
obstructive hydrocephalus, with BIH subjects exhibiting mostly EEG activity fulfilling
category B criteria. Frequent bursts in EEG activity were also observed mostly in the BIH
group, which the investigators ascribed to rising intracranial pressure.
78
Chapter 1.
While the study of Mani and Townsend (1964) demonstrated that BIH is associated with
changes in brain oscillatory activity, experimental limitations weakened the study.
Predominantly normal EEG activity was observed in BIH sufferers; however, this could
have been linked to the small sample size examined (n = 14), which reduced the study’s
statistical power and the number of adjustments performed in the final analysis. Limited
electrode positions were also assessed (8-channel Ediswan EEG system). Such a limited
assessment of brain activity overlooks potential changes in brain activity occurring across
the entire cortex. Lal & Craig (2001) suggest research validity may be improved by
assessing more scalp locations.
Additionally, although the electrodes were distributed evenly over the cortex, they were
not attached in accordance with the standard international 10-20 system of EEG. Future
studies exploring changes in EEG activity in hypertension should utilise the universally-
accepted standardised international 10-20 system using a montage that ensures uniform
scalp coverage (e.g. the 19-channel 10-20 montage). This will improve the comparability
of the findings between subsequent studies. Future investigations should also report the
grade/classification of hypertension at the time of electrophysiological assessment.
Various grades of hypertension have been described (e.g. Grade 1, Grade 2, etc.) and
these could influence electroencephalography activity. Taken together, these factors may
account for the unusual findings obtained by Mani and Townsend (1964). However, it is
clear that there is a paucity of data examining neurophysiological changes in
hypertension.
79
Chapter 1.
1.12 Effect of glucose lowering and anti-hypertensive

medication on EEG
Whether anti-hypertensive and glucose-lowering medication influence EEG activity
and can reverse the electrophysiological abnormalities observed in these conditions
remains unclear. It has been suggested that blood pressure-lowering therapy enhances
cerebral perfusion (Obisesan, 2009). Current evidence also indicates that anti-
hypertensive medication lowers the risk of dementia and AD by 12% and 16%,
respectively (Ding et al., 2020); however, no convincing data exist to suggest the most
effective class for optimum cognitive outcomes (Iadecola & Gottesman, 2019; Ding et
al., 2020). Interestingly, no studies have examined the impact of anti-hyperglycaemic
medication on brain oscillatory activity in patients with T2DM. Observational studies
suggest that some anti-diabetic agents (e.g. sodium glucose cotransporter 2 inhibitors
(SGLT2is) and glucagon-like peptide 1 receptor agonists (GLP-1RAs)) improve
cognition by influencing critical brain processes (e.g. metabolism, inflammation,
regeneration, synaptogenesis), but no conclusive evidence indicates that these
pharmacotherapies modify the risk of cognitive dysfunction in T2DM (De Galan et al.,
2009; Areosa Sastre et al., 2017). No data also exist to suggest that glucose-lowering
therapy can reverse aberrant oscillatory activity. Investigators recommend that future
randomised controlled trials (RCTs) should explore cognitive outcomes as a secondary
endpoint (Biessels & Despa, 2018; Biessels & Whitmer, 2019). This may identify
associations between specific anti-hyperglycaemic agents and cognitive outcomes. It
is clear this area of research requires further exploration and that future studies
exploring EEG activity in patients with DM (T1DM and T2DM) and HTN should
report medications taken to establish possible associations.
80
Chapter 1.
1.13 Basis and Study Significance
1.13.1 Implications of the Present Study

Diabetes mellitus (T1DM and T2DM) and hypertension (HTN) are highly-prevalent
chronic diseases that are increasing in incidence and prevalence globally. Both are also
associated with numerous debilitating and life-threatening complications. One previously
under-recognised complication common to both conditions, now being increasingly
recognised as an important comorbidity in clinical practice, is cognitive dysfunction. This
often manifests as subtle, irreversible cognitive decrements in DM, and cognitive decline
in HTN. Although diabetes-associated cognitive decrements are known to progress
perniciously, it is unknown whether hypertension-associated cognitive dysfunction is
reversible or can be attenuated by pharmacotherapy. The mechanisms underlying the
accelerated cognitive decline in both conditions remain unclear, so further research is
urgently warranted to understand better the relationship between these conditions and
cognition.
Numerous studies (cross-sectional and longitudinal) have assessed cognitive functioning

in subjects with DM (T1DM and T2DM) and HTN using objective neuroimaging
modalities and psychometric assessments; however, the precise cognitive domains
affected remain unclear, as does the relationship between BGL and BP (SBP and DBP)
and EEG activity and performance in individual domains of cognition. Given the trends
in prevalence predicted for each condition, it is clear that future cognitive research is
urgently warranted. Administration of reliable and validated cognitive screening tools,
such as the Mini-Mental State Examination (MMSE) (Folstein, McHugh, & Folstein,
1975) and the Cognistat (Kiernan et al., 1987), could reveal the cognitive domains
detrimentally affected by these conditions and determine the suitability of these
assessments for screening for the subtle cognitive decrements linked to these conditions.
They could also identify individuals at high-risk of progressing to MCI, the earliest
detectable stage of cognitive impairment, before irreversible deterioration in cognition
has occurred. This would avert the adverse cognitive outcomes and the substantial
socioeconomic and emotional costs associated with both conditions.
81
Chapter 1.
Substantial epidemiological evidence has revealed an association between both DM

(T1DM and T2DM) and HTN and an increased risk of cognitive impairment (MCI, AD,
VaD, and dementia) (Sierra et al., 2012; Koekkoek et al., 2015). The relative risk for
dementia has been documented as a 1.5 to 2.5 times increased risk in patients with T2DM
(Strachan et al., 2011) compared to healthy subjects, whereas those with T1DM have a
60% increased risk of dementia (Smolina et al., 2015). In contrast, the relative risk for
dementia in hypertension remains unknown. It has also been estimated that one in three
AD cases worldwide indirectly arises from these conditions (Norton et al., 2014). Given
evidence consistently indicates that both conditions are associated with an increased risk
of cognitive impairment, it is paramount that further research is undertaken to understand
better the pathophysiological pathways that link these conditions to cognitive impairment,
particularly in the early stages. Such research could ultimately have broader societal
implications, including the identification of the prominent causative determinants
responsible for aggravating the cognitive decline commonly reported in these conditions.
This could enable early therapeutic intervention and delay the onset and progression to
neurodegerenative diseases such as AD and dementia. It could also contribute to reducing
the substantial socioeconomic and emotional costs linked to these cognitive diseases.
Several cognitive measures are available to assess cognition in DM and HTN; however,
there is currently no consensus among investigators concerning the most suitable
cognitive measures for screening the cognitive decrements associated with these
conditions. No objective neurological instruments or cognitive measures can also reliably
and accurately detect the subtle cognitive decrements triggered by these conditions, as
they manifest and progress insidiously (Biessels & Despa, 2018; Biessels & Whitmer,
2019). The pernicious nature of these decrements complicates timely and accurate
detection of incipient signs and symptoms, leading to delays in identification and
appropriate intervention (Srikanth et al., 2020). The EEG is an established
neurophysiological measure that has shown promising potential in monitoring changes in
brain activity in MCI and in both DM and HTN and trajectories in cognition. Previous
research indicates the EEG can consistently and reliably detect changes in oscillatory
activity associated with these conditions and early stages of cognitive impairment (MCI)
(Jelic et al., 2000; Brismar et al., 2002; Hyllienmark et al., 2005; Cooray, Hyllienmark,
& Brismar, 2011). The EEG is also cost-effective, non-invasive, does not emit radiation,
and demonstrates high temporal resolution. Such robust temporal resolution allows for
82
Chapter 1.
rapid and accurate detection of abnormalities in brain activity, which compares

favourably with costly neuroimaging technologies. The latter also often require formal
training for proper usage and repetition, due to head movements during scans (Lord et
al., 2020). Given the rising global tide of neurodegenerative diseases, there is an urgent
need for a non-invasive biomarker that can accurately and consistently detect incipient
cognitive decline. The present study could yield evidence to support the widespread
deployment of EEG in clinical practice to facilitate reliable and accurate identification of
the subtle and slowly-progressing cognitive dysfunction triggered by both DM and
hypertension. This is often undetected by formal neuropsychological screening
assessments.
There is a lack of research examining the neurophysiological changes associated with

DM and HTN, specifically in the case of the latter and T2DM. The research in this area
is also predominantly exploratory, limiting an in-depth understanding of the precise
electroencephalography changes associated with these disorders. Consequently, the
neurophysiological changes remain poorly understood. While studies have investigated
EEG activity in subjects with DM, no study has explored the electroencephalography
changes linked to hypertension. The deleterious relationship between raised blood
pressure and cognition has been documented since the 1960s yet the mechanisms
underlying hypertension-associated cognitive dysfunction remain elusive. Given the
prevalence of hypertension is increasing, exploring the EEG changes that occur in
hypertension is critical and relevant. Such research could reveal the brain areas
susceptible to early deterioration from hypertension and possible signature EEG-based
biomarkers of raised blood pressure. It could also determine the suitability of the EEG
for detecting the subtle cognitive dysfunction linked to hypertension.
The comparative nature of the study comparing cognitive functioning between DM and
HTN using objective neurophysiological measures and subjective psychometric tools is
also novel. Previous investigations have compared cognitive function between clinical
and non-clinical samples, but none have compared cognitive functioning (global and
domain-specific) in subjects with DM (T1DM and T2DM) and HTN using established
cognitive screening tools (the MMSE and the Cognistat). Both are reliable and validated
neurocognitive assessments widely administered in clinical contexts to screen for early
cognitive impairment (Folstein, McHugh, & Folstein, 1975; Tombaugh & McHugh,
83
Chapter 1.
1992; Pangman et al., 2000; Lancu & Olmer, 2006). Notably, the MMSE is currently
recommended in emerging guidelines for screening for subtle cognitive decrements and
cognitive impairment in elderly patients with DM (ADA, 2020; Srikanth et al., 2020).
Therefore, the present research could provide preliminary insight into the suitability of
these cognitive assessments for identifying the slowly-progressing and subtle changes in
cogntion associated with these conditions.
The present exploratory study, and subject of this thesis, is a novel, cross-sectional
investigation that aims to address the limitations described herein by exploring cognitive
function in four sample groups (n = 49, non-clinical; n = 30 diabetes subjects (n = 13,
T1DM; and n = 17, T2DM) and n = 15 HTN patients) using electroencephalography and
non-invasive cognitive measures (the Mini-Mental State Examination (Folstein, Folstein,
& McHugh, 1975) and the Cognistat (Kiernan et al., 1987)).
1.14 Hypotheses
In clinical (T1DM, T2DM, and HTN) and non-clinical samples using cognitive measures
(EEG and psychometric assessment), it is hypothesised that:
1. There will be differences in cognitive performance between clinical and non-clinical

samples, and
2. There will be correlations between cognitive measures and blood pressure (BP) and
blood glucose level (BGL)
1.15 General Aims

1. Investigate differences in cognitive performance between clinical and non-clinical
samples
2. Investigate associations between cognitive measures (MMSE, the Cognistat, and
EEG) and blood pressure (BP) and blood glucose level (BGL)
84
Chapter 1.
1.16 Specific Aims (Aim 1)

1. Investigate differences in global cognitive performance (Mini-Mental State
Examination) between clinical (T1DM, T2DM, and HTN) and non-clinical samples
2. Investigate differences in domain-specific cognitive performance (the Cognistat)
between clinical (T1DM, T2DM, and HTN) and non-clinical samples
3. Investigate differences in electroencephalography (EEG) activity between clinical
(T1DM, T2DM, and HTN) and non-clinical samples
1.17 Specific Aims (Aim 2)

1. Investigate associations between cognitive measures (MMSE, the Cognistat, and EEG)
and pre-study and post-study systolic and diastolic blood pressure (BP)
2. Investigate associations between cognitive measures (MMSE, the Cognistat, and EEG)
and pre-study and post-study blood glucose level (BGL)
85
Chapter 2.
2. Methodology
2.1 Methodology Summary

The methodology described in this chapter was developed to address the aims and
hypotheses introduced in Chapter 1. The cross-sectional study was conducted in the
Neuroscience Research Unit (NRU) at the University of Technology Sydney (UTS), in a
temperature- and lighting-controlled and sound-attenuated neurophysiology laboratory,
with minimal ambient interference. Testing involved one session, taking approximately
two hours for each participant. All instrumentation (objective and subjective measures)
used in this investigation was reliable and validated at the time of assessment. Given
circadian rhythms have been shown to influence cognitive performance (Valdez,
Ramírez, & García, 2012; Wright, Lowry, & LeBourgeois, 2012), testing was conducted
during peak wakefulness periods (9 am – 2 pm and 4 pm – 8 pm).
2.2 Ethics Approval and Consent

The present study was conducted under ethics approval (HREC: 201400010) obtained
from the UTS Human Research Ethics Committee (HREC). Prior to commencing
experimental testing, all test subjects were provided with a concise overview of the study,
methodology involved, and the study inclusion/exclusion criteria. Study volunteers were
additionally informed that participation was voluntary and that they could discontinue
involvement in the research at any time, without providing reasons for withdrawal. If the
inclusion criteria were fulfilled (Section 2.4), written informed consent was then obtained
from all test participants agreeing to participate in the study, prior to data collection. Both
the participant and the researcher then read and signed the consent form and, in
accordance with UTS ethics requirements, retained a copy of the consent form (Appendix
8.1).
86
Chapter 2.
2.3 Recruitment of Study Participants

49 healthy (hereafter referred to as non-clinical) volunteers, 30 participants with DM
(T1DM: n = 13; T2DM: n = 17) and 15 with HTN (systolic and/or diastolic blood pressure
≥ 140mmHg/90mmHg) aged between 18-80 years were recruited from the local Sydney
community for the present study. Study volunteers (both non-clinical and clinical) were
recruited using various strategies: advertisement via posters (electronic and physical)
(Appendix 8.7) in populated urban spaces and relevant medical clinics throughout
Sydney; advertisement via relevant and professional organisations, including Diabetes
Australia, Diabetes NSW, and Alzheimer’s Australia; presentations at relevant
community events; and by word-of-mouth.
2.4 Study Inclusion/Exclusion Criteria

Present study inclusion criteria required participants from the non-clinical cohort to be
between 18-80 years of age with no underlying chronic disease (such as diabetes,
hypertension, asthma, etc.), intellectual impairment, or psychosis (depression, substance
abuse) that could potentially limit compliance in the study or influence the data.
Participants reporting clinically-diagnosed DM (T1DM or T2DM) or HTN were eligible
for inclusion in the clinical group and this was determined by ascertaining the
participant’s response to Question 18 of the Lifestyle Appraisal Questionnaire (LAQ)
(Craig, Hancock, & Craig, 1996). Participants with DM (T1DM or T2DM) or HTN who
reported taking medications to control their condition or complications linked to their
respective chronic disease (e.g. microvascular or macrovascular complications), were
also eligible for inclusion.
If participants (from either the non-clinical or clinical sample) indicated one or more of
the following, as solicited by the Lifestyle Appraisal Questionnaire (LAQ) (Craig,
Hancock, & Craig, 1996), they were immediately excluded from further study
participation: illicit substance use/dependence, psychotropic medication, alcoholism (>16
standard alcoholic drinks per day), smoking (>10 cigarettes daily), severe intellectual
disorder or psychosis. The literature suggests these lifestyle risk factors can cause
irreversible changes in underlying brain structures, affecting normal cognitive function
(Le Berre et al., 2014; Karama et al., 2015); hence, this would have influenced the data
obtained in the present study.
87
Chapter 2.
2.5 Blood Pressure (BP) Measurement

In accordance with established and recommended blood pressure (BP) measurement
guidelines (Pickering et al., 2005; Unger et al., 2020), after a five-minute rest period,
participant brachial blood pressure was recorded three times, both before and after
cognitive testing. Blood pressure was measured using a reliable and validated automatic
non-invasive digital BP monitor (OMRON Healthcare Co., Ltd, IA1B (HEM-7000-C1L),
Kyoto, Japan)) (Figure 2.1) and recorded from the participant’s right upper arm
(positioned at the level of the heart) in the upright sitting position, with feet flat on the
floor and no talking before, between, or during measurements (Figure 2.2). Three
measurements were recorded, since the literature indicates this improves accuracy and
attenuates potential influence of the effects of ‘white-coat hypertension’ (office systolic
blood pressure ≥ 140/90 mm Hg at least three times), a syndrome affecting approximately
10-30% of individuals worldwide (Pickering et al., 2005; Unger et al., 2020). This
technique also negates auscultation-induced errors (Unger et al., 2020). To minimise
potential carry-over from previous measurements and permit the restoration of elastic
properties of blood vessels, participants were given a rest period (1-2 minutes) between
each BP measurement (Unger et al., 2020). Subsequently, each of the three BP readings
were averaged to determine mean BP before and after the study, for each participant.
88
Chapter 2.
Figure 2.1. The non-invasive automatic blood pressure device used to record
participant brachial blood pressure.
89
Chapter 2.
Figure 2.2. Appropriate posture and seating position for recording blood pressure
using a non-invasive automatic blood pressure monitor. Adapted from
Unger et al., (2020, p 4).
2.5.1 BP Inclusion/Exclusion Criteria

Blood pressure of subjects from the non-clinical sample determined their inclusion or
exclusion from the study. Subjects from the non-clinical group with average BP (for either
systolic BP or diastolic BP, or both) meeting Grade 2 HTN criteria (≥ 160/100 mm Hg)
were immediately excluded from further study involvement and, in accordance with the
UTS HREC approved emergency protocol, were offered to be escorted to a nearby
medical centre (Appendix 8.2). Blood pressure meeting this threshold is associated with
premature cardiovascular mortality and long-term adverse cardiovascular outcomes
(Weber et al., 2014; Unger et al., 2020).
90
Chapter 2.
If BP measurements (for systolic or diastolic alone, or both) for the non-clinical group
met Grade 1 HTN criteria (≥ 140/90 mm Hg but ≤ 160/100 mm Hg), they were also
excluded from the study. In this instance, the participant was notified of their elevated BP
and advised to consult their medical professional for further clinical evaluation. No
exclusion threshold for BP applied for participants with clinically-diagnosed T1DM or
T2DM or HTN, as BP is typically elevated in these chronic diseases (Deshpande, Harris-
Hayes, & Schootman, 2008; DeFronzo et al., 2017). However, as outlined in the UTS
HREC approved emergency protocol, participants from the clinical cohort (T1DM or
T2DM or HTN) were still advised to consult their medical practitioner (BP ≥ 140/90 mm
Hg) and were offered to be escorted to the nearest medical centre if their BP met Grade
1 or 2 hypertension criteria (Table 2.1).
91
Chapter 2.
Table 2.1. Blood pressure inclusion and exclusion limit thresholds.
Clinical Sample (T1DM,

Category SBP (mm Hg) DBP (mm Hg) Non-clinical Sample
T2DM, and HTN)
Normal BP < 130 < 85 Included Excluded
High – normal 130 – 139 85 – 89 Included Excluded

Excluded and offered an
Included and advised to
Grade 1 hypertension 140 – 159 90 – 99 escort to accessible
consult physician
medical centre
Excluded and offered an
Included and advised to
Grade 2 hypertension ≥ 160 ≥ 100 escort to accessible
consult physician
medical centre
Key:
BP – Blood Pressure ≥ – greater than or equal to < – less than
mm Hg – millimetres of mercury T1DM –Type 1 Diabetes Mellitus T2DM – Type 2 Diabetes Mellitus
HTN – Hypertension SBP –systolic blood pressure DBP – diastolic blood pressure
HREC – Human Research Ethics Committee
92
Chapter 2.
If the respective BP inclusion criteria were fulfilled for each group, blood glucose
concentrations for each study participant were then determined.
2.6 Blood Glucose Level (BGL) Determination

Following pre-study (baseline) BP measurement, the blood glucose level (BGL) of all
test participants was determined. Two-hour (2-hr) fasting blood glucose concentrations
were determined before cognitive testing using a sterile, single-use lancing device, which
included three depth settings (Accu-Chek Safe-T-Pro Plus, Roche Diabetes Care
Australia Pty Ltd), and measured using a reliable and validated blood glucometer device
(Accu-Chek Performa, Roche Diagnostics Australia Pty Ltd) (Figure 2.3). All blood
glucose concentration values were reported in millimoles per litre (mmol/L). This spot-
test was performed for all subjects as glycaemic events (hypo- and hyperglycaemia) are
associated with detectable changes in electroencephalography activity (Sommerfield,
Deary, & Frier, 2004; Cox et al., 2005; Graveling, Deary, & Frier, 2013; An et al., 2015).
Therefore, the impact of hypo- or hyperglycaemia could be negated and the relationship
between other moderating variables (e.g. glycaemic control and disease duration) could
be established. If blood glucose concentrations fell outside the normal recommended
blood glucose concentration range (Figure 2.4), participants were still included for further
study participation.
93
Chapter 2.
Figure 2.3. Blood glucometer device (left) and sterile, single-use lancing device
(right) used to determine blood glucose concentrations for all study
participants
!

#9<<9=?<5C@5B<9DB5==?< "
Figure 2.4. Normal 2-hour postprandial blood glucose concentration range. After 2-
hours of fasting, blood glucose concentrations falling between 3.5-
8mmol/L are typically considered normal, whereas concentrations outside
this range are considered abnormal (Diabetes Australia, 2020).
94
Chapter 2.
After the BGL was determined for each participant, they then completed the Lifestyle
Appraisal Questionnaire (LAQ) (Craig, Hancock, & Craig, 1996). Participants reporting
diagnosed T1DM or T2DM or HTN completed an additional questionnaire developed in-
house by the researcher for each disease type. These disease-specific questionnaires
solicited additional characteristics relevant to their respective chronic disease not already
obtained from the Lifestyle Appraisal Questionnaire (e.g. disease duration, glycosylated
haemoglobin (HbA1C), etc.) (Appendix 8.3 and 8.4).
2.7 Demographic Data Acquisition
2.7.1 Lifestyle Appraisal Questionnaire (LAQ) (Craig, Hancock, & Craig,

1996)
Lifestyle factors, perceived stress levels, and demographic characteristics of all
participants (non-clinical and clinical) were obtained using a reliable and validated
Lifestyle Appraisal Questionnaire (LAQ) (Craig, Hancock, & Craig, 1996). The LAQ is
a sensitive two-part questionnaire frequently administered in clinical and research
contexts that assesses lifestyle risk factors and perceived stress over an 8-week period
(Craig, Hancock, & Craig, 1996). It is also widely considered a reliable measure of long-
term health outcomes (Craig, Hancock, & Craig, 1996).
The LAQ comprises two parts: Part I consists of 22 questions that solicit lifestyle risk
factors associated with an increased risk of developing lifestyle-associated chronic
diseases (e.g. alcoholic beverage consumption, smoking patterns, sleep quality, lifestyle
disease history, drug intake). Additional lifestyle data, including body mass index (BMI),
diet, exercise patterns, and alternative relaxation techniques undertaken, are also
obtained. The maximum obtainable score is 73, with higher scores indicating an increased
likelihood of developing long-term lifestyle-related diseases, such as coronary heart
disease and diabetes mellitus (Craig, Hancock, & Craig, 1996). In contrast, Part II
assesses participant lifestyle pressures and perceived stress levels, evaluated by 25 Likert
scale type questions scored based on severity (0 – 3: 0 – almost never, 1 – sometimes, 2
– often, 3 – almost always). The maximum attainable score for Part II is 75, with higher
scores suggesting greater perceived stress levels (Craig, Hancock, & Craig, 1996).
95
Chapter 2.
Participants reporting T1DM or T2DM or HTN completed additional questionnaires

developed in-house. The purpose of these questionnaires was to solicit additional disease-
specific characteristics relevant to each condition not already obtained by the LAQ (e.g.
duration, age of disease onset, glycosylated haemoglobin (HbA1C), frequency of blood
glucose/blood pressure monitoring, alternative therapies used, medications taken and
frequency, and self-scored disease management) (Appendix 8.3 and 8.4). Current
literature suggests these factors moderate the relationship between each chronic disease
and cognition (Ryan, Geckle, & Orchard 2003; Roberts et al., 2008; Wessels et al., 2008).
Participants indicating no chronic condition (non-clinical cohort) were exempt from
completing these questionnaires and only completed the LAQ.
Following completion of the LAQ, participant responses to questions 1, 2, 7, 8 and 18

were reviewed to determine if the participant fulfilled the study inclusion criteria (Section
2.4). Question 18 of the LAQ, “Do you at present suffer from a chronic condition”, did
not apply for participants from the clinical sample. If participants fulfilled the study
inclusion criteria, a non-invasive, elastic 32-channel EEG cap with pre-determined
electrode positions complying with the standard International 10-20 system (Jasper, 1958)
for EEG electrode placement was attached to the participant’s scalp.
96
Chapter 2.
2.8 Electroencephalography Data Acquisition

Thirty channels of electroencephalography (EEG) recording were obtained using a
NeuroScan Synamps amplifier and Scan 4.3 recording software (Compumedics
NeuroScan, Charlotte, NC, USA). Brain electrical activity was recorded using a non-
invasive 32-channel elastic cap embedded with Ag/AgCl electrodes (Quik-Cap,
Compumedics, sampling rate: 1000 Hz) (Figure 2.5) in predetermined positions
conforming to the standard international 10-20 system of EEG electrode placement
(Jasper, 1958) (Figure 2.6). The scalp electrode positions examined were: (Fp1 (Fronto
polar 1, Fp2), (F7 (Frontal 7, F3, Fz, F4, F8), (FT7 (Fronto-temporal 7, FT8), (FC3 (Fronto-
central 3, FCZ, FC4), (T7 (Temporal 7, T8), (TP7 (Temporo-parietal 7, TP8), (C3 (Central
3, CZ, C4), (CP3 (Centro-parietal 3, CPZ, CP4), (P7 (Parietal 7, P3, PZ, P4, P8), and (O1
(Occipital 1, OZ, O2) (Figure 2.6).
Figure 2.5. The 32-channel non-invasive, elastic EEG cap used to record brain
electrical activity from study participants.
97
Chapter 2.
Figure 2.6. Topographic representation of the standard international 10-20 system of

electroencephalography (EEG). Areas circled in red indicate electrode
positions investigated in the present study. Odd numbers represent left
hemispheric positions; even numbers, right hemispheric positions. The ‘Z’
notation indicates positions occurring along the midline of the cranium.
Adapted and modified from Mert & Akan, (2018, p 5).
Key:
Fp – Fronto-polar F – Frontal FT – Fronto-temporal FC – Fronto-central
T – Temporal C – Central TP – Temporo-parietal CP – Centro-parietal
O – Occipital PO – Parieto-occipital Z – midline
Electrodes were referenced against a ground electrode and electrode impedance was
maintained below five kilo-ohms (KW) (Keil et al., 2014). Electroencephalography data
acquisition occurred while the participant was seated in a comfortable upright position in
a sound-attenuated, temperature- and lighting-controlled laboratory with minimal
interference. An additional electrode pair positioned above and below the orbit
participant’s left eye (VEOU and VEOL) recorded electro-oculogram (EOG) activity to
attenuate eye movement artifacts from the EEG signal (refer to section 2.10.1).
Participants were also instructed to minimise movement during EEG recordings to reduce
potential contamination of the electroencephalography signal from movement artifacts.
Following accurate EEG cap placement and filling of relevant EEG electrodes on the cap
with highly-conductive gel (Signa Gel, Parker Laboratories Inc, USA) (Figure 2.7), the
98
Chapter 2.
EEG signal was examined visually on the computer to determine whether adequate
signals from each electrode were being generated (Figure 2.8), as well as to remove any
potential artifacts evident (refer to section 2.10.1). If a poor tracing was generated, the
following adjustments were performed until a reasonable signal was observed: nearby
electrical equipment potentially interfering with the EEG signal was switched off;
adjustments to electrical leads and gel quantities in electrode positions yielding poor
signals were made; and the EEG cap was re-positioned. Once electrode impedance for
each electrode was below the specified threshold value indicated above, and signals from
each electrode position were considered acceptable, a baseline electroencephalography
recording was obtained.
Figure 2.7. Electrode gel used to fill electrodes in the EEG cap, sterile syringe and
blunted needle.
99
Chapter 2.
Figure 2.8. Example of an unprocessed EEG recording considered acceptable. The ‘X’
axis indicates time, whereas the ‘Y’ axis represents microvolts for the
electrode positions examined.
Electrophysiological data were obtained from each of the 30 electrode positions (see
section 2.8 for electrode locations) according to the standard international 10-20 system
(Jasper, 1958) over two study phases, baseline and active, each 5-minutes in duration.
The baseline phase involved the participant sitting quietly observing a blank computer
screen, whereas the active phase involved engagement with a cognitively-stimulating task
via a computer screen using pre-loaded software developed in-house, the Stroop Colour
Word Test (Stroop, 1935) (refer to section 2.8.1) (Figure 2.9). Prior to recording the active
Stroop test-linked EEG, a short practice run was conducted to determine whether the
participant understood the instructions provided about the Stroop test and responded
appropriately.
100
Chapter 2.
Figure 2.9. Screenshot depicting the Stroop Colour Word Test Program. This
assessment prompts participants to process matched and mismatched
stimuli (i.e. correctly identify the colour of the text shown on the screen as
fast as possible). For example, in the screenshot above, the correct answer
would be blue.
2.8.1 Stroop Colour Word Test (Stroop, 1935)

The Stroop Colour-Word Test was developed in 1935 and is a commonly administered
and reliable neuropsychological measure that assesses several cognitive modalities,
including cognitive flexibility, selective attention, psychomotor efficiency, and executive
cognitive control (Adleman et al., 2002; Homack & Riccio, 2004; Van der Elst et al.,
2006; Pilli et al., 2013). It is widely considered a sensitive indicator of executive cognitive
functioning, a broad cognitive domain concerned with multiple high-order cognitive
processes mediated by the frontal lobe. The instrument has been deployed extensively in
cognitive investigations in several areas of research, including frontal lobe function,
developmental changes in the frontal lobe, disruptions in cognitive function triggered by
neuropsychiatric disorders and, of relevance to this thesis, progressive changes in
psychomotor efficiency and executive control, two cognitive domains that are
consistently detrimentally affected by both DM and hypertension (Adleman et al., 2002;
Alvarez & Emory, 2006; Beratis et al., 2010).
101
Chapter 2.
A clinically-valuable characteristic of the tool is the generation of an observable “Stroop

interference effect”, elicited when participants must correctly identify text colour when
text is displayed in mismatched colours (i.e. the word GREEN printed in black ink) (Liotti
et al., 2000; Adleman et al., 2002; Van der Elst et al., 2006). This perceptible Stroop
interference effect, which is characterised by an elongation in response time and caused
by the simultaneous activation of two converging cortical pathways controlling attention
(MacLeod & MacDonald, 2000), is considered a sensitive indicator of selective attention
and executive cognitive functioning (Van der Elst et al., 2006). Both DM and HTN have
been consistently shown to affect cognitive flexibility and executive cognitive function
(Adleman et al., 2002; Alvarez & Emory, 2006; Beratis et al., 2010); therefore, the Stroop
test can be considered a suitable assessment for exploring disturbances in cognitive
function triggered by either of these chronic diseases. While the assessment is available
in several versions, the computerised version of the tool was utilised in the present
investigation. Higher accuracy in stimulus presentation has been reported using the
computer version (Pilli et al., 2013).
Following acquisition of EEG data (baseline and active), two reliable and validated
neurocognitive assessments, the Mini-Mental State Examination (MMSE) (Folstein,
Folstein, & McHugh, 1975) and the Cognistat (Kiernan et al., 1987), were then
administered.
2.9 Cognitive Assessment

Cognitive function in several domains of cognition was assessed subjectively using two
reliable and validated neuro-psychometric batteries; the Mini Mental State Examination
(MMSE) (Folstein, Folstein, & McHugh, 1975) and the Cognistat (Kiernan et al., 1987).
Both are frequently deployed cognitive screening tools, but key differences exist between
the two (discussed below). To negate any potential bias due to order effect, the cognitive
assessments were administered in a randomised order.
102
Chapter 2.
2.9.1 Mini-Mental State Examination (MMSE) (Folstein, Folstein, &

McHugh, 1975)
A reliable and validated neuro-psychometric tool available in several languages, the
Mini-Mental State Examination (MMSE) is a brief clinical measure frequently
administered in clinical and research environments that screens for early cognitive
impairment (Folstein, McHugh, & Folstein, 1975; Tombaugh & McHugh, 1992;
Pangman et al., 2000; Lancu & Olmer, 2006) (Figure 2.10). Administered in a command-
response based manner, and taking approximately 5-10 minutes, the assessment consists
of 11 questions that assess essential cognitive functions. The MMSE comprises two parts:
Part 1, which requires vocal responses, assesses cognitive domains of orientation to time
and place (10 points), registration (3 points), attention/calculation (5 points) and recall (3
points), whereas Part 2, examined via verbal and written responses, assesses language
(naming, repetition, reading, comprehension, visuoconstruction) (9 points) (Figure 2.11)
(Folstein, McHugh, & Folstein, 1975; Tombaugh & McHugh, 1992; Pangman, Sloan, &
Guse, 2000). The maximum obtainable score is 30, which represents the summed score
from each cognitive domain (Folstein, McHugh, & Folstein, 1975). Scores ≤ 23 typically
indicate probable cognitive dysfunction and have been linked with a subsequent dementia
diagnosis in approximately 79% of cases (Folstein, McHugh, & Folstein, 1975; Lancu &
Olmer, 2006; Marioni et al., 2011).
103
Chapter 2.
Figure 2.10. The Mini-Mental State Examination (MMSE). Adapted from Folstein,
McHugh, & Folstein (1975).
104
Chapter 2.
Figure 2.11. Stimulus sheet for assessing language in Part 2 of the MMSE.
Although the rapid administration of the assessment is widely praised, and broad
consensus exists among researchers that the psychometric properties of the assessment
are considered reasonable despite its low sensitivity (test-retest reliability values: 0.56-
0.99, Cronbach’s alpha: 0.54-0.96) (Tombaugh & McHugh, 1992), the cognitive tool has
been criticised for being susceptible to various demographic and cognitive variables.
Such variables include age, education level/attainment, and cultural background,
potentially resulting in misclassification of actual cognitive status (Tombaugh &
McHugh, 1992). Unlike sex, which has not been shown to influence MMSE performance,
age and education have both been consistently reported as powerful moderators affecting
MMSE scores (Crum et al., 1993). In a large population-based study conducted by Crum
et al. (1993), MMSE performance was strongly associated with both age and education
level, with higher education levels correlating with higher median test scores and stronger
cognitive performance in younger test populations. To address this limitation, authors
have suggested adjusting cut-off scores relative to the age and education level of the
population examined. This ensures that an unbiased outcome is achieved (Crum et al.,
1993; Galasko et al., 1996).
105
Chapter 2.
The widely-recommended 23-point cut-off score suggestive of potential cognitive

impairment has also attracted considerable debate, as some investigators argue this
recommended cut-off threshold may not yield optimal classification accuracy (van Gorp
et al., 1999; O’Bryant et al., 2008). Measurable improvements in both test accuracy and
sensitivity were observed by Van Gorp et al. (1999) when the cut-off score was increased
to ≤ 26/30. In agreement with this, O’Bryant et al. (2008) also reported improved
diagnostic accuracy in identifying dementia in highly-educated individuals when the cut-
off score was increased to ≤ 27/30. Therefore, on account of the improved diagnostic
accuracy and sensitivity reported, Van Gorp et al. (1999) and O’Bryant et al. (2008)
advise dismissing the recommended 23-point cut-off threshold to maximise diagnostic
accuracy, especially when assessing highly-educated populations.
2.9.2 Cognistat (Kiernan et al., 1987)

The Cognistat (formerly known as the Neurobehavioural Cognitive Status Examination
(NCSE) is another reliable, validated and widely-deployed brief cognitive examination
that quantitatively screens for cognitive dysfunction in several key cognitive domains
(Kiernan et al., 1987; Whiteside et al., 1996; Drane et al., 2003; Macaulay et al., 2003).
It is available in several languages and is routinely administered conjointly with or instead
of the MMSE, taking approximately 15-20 minutes to administer (Kiernan et al., 1987;
Logue et al., 1993; Engelhart, Eisenstein, & Meininger, 1994; Eisenstein et al., 2002).
The Cognistat consists of 10 subtests that examine a diverse range of cognitive domains:
orientation and attention, construction ability, memory, language, calculation, and
reasoning, the latter subdivided into two further modalities, similarities and judgment
(Schwamm et al., 1997; Engelhart, Eisenstein, & Meininger, 1994; Oehlert et al., 1997;
Eisenstein et al., 2002) (Figure 2.12). However, unlike the MMSE and other cognitive
screening tools that only yield a summed global score, the Cognistat measures domain-
specific cognitive performance (i.e. performance in individual cognitive domains),
yielding a graphical representation of cognitive performance (Figure 2.13). This provides
researchers with a quick snapshot and differentiated profile of overall patient cognitive
status and enables prompt recognition of potential early cognitive dysfunction (Logue et
al., 1993; Macaulay et al., 2003).
106
Chapter 2.
Figure 2.12. Participant completing the Construction domain subtest of the Cognistat.
This subtest requires participants to create the images shown in the
stimulus manual using the coloured square tiles provided. Permission to
reproduce the image has been obtained from the participant.
107
Chapter 2.
Figure 2.13. The Cognistat cognitive status profile. Scores beneath the average range
for each cognitive domain are graded (mild, moderate, severe) and indicate
the degree of cognitive impairment. Adapted from the Cognistat manual
(2007).
One characteristic that differentiates the Cognistat from other cognitive assessments is its
unique “screen-metric” construct. This identifies whether impairment is evident in a
particular cognitive domain, or if further cognitive evaluation is required (Oehlert et al.,
1997; Gupta & Kumar, 2009; Rice et al., 2015). First, the participant is asked a ‘screen’
question, representative of the cognitive domain being assessed. This question is typically
more difficult than questions presented subsequently in the ‘metric’ section. If the
participant answers the ‘screen’ question correctly, cognitive function for that particular
domain is considered intact, the maximum subtest score is awarded, and the examiner
advances to the next cognitive domain. Conversely, if the participant fails the ‘screen’
question the remaining ‘metric’ questions, which progressively increase in difficulty and
determine whether function is intact, are then administered (Oehlert et al., 1997; Nøkleby
et al., 2008; Rice et al., 2015).
108
Chapter 2.
Although the Cognistat demonstrates reasonable psychometric properties (high internal

consistency (Cronbach’s alpha: 0.94)) (Kiernan et al., 1987), limitations have been
identified with the popular clinical measure. While the Cognistat possesses reasonably
high sensitivity (sensitivity: 93%), Drane and Ossato (1997) found the assessment lacked
sound specificity. This effect was pronounced in memory and construction domains,
where it was found to yield an unacceptably high number of false positive results in
healthy elderly patients. Consequently, Drane and Ossato (1997) recommend extending
the average range score in each cognitive domain to address this limitation. Similar to the
MMSE, the Cognistat has also been shown to be susceptible to the effects of cognitive
variables (age and education level) (Drane et al., 2003). Drane et al. (2003) found that
age and education level strongly influenced Cognistat test performance. While age was
shown to primarily affect performance in memory and attention domains, both age and
education level influenced construction ability performance. Drane et al. (2003)
recommend considering age and education level when evaluating cognitive performance
to reduce the influence of these demographic variables on Cognistat performance. Drane
et al. (2003) also suggest disregarding the recommended normative cut-off point scores
for classifying impairment for non-cognitively healthy populations, arguing this may
potentially result in misclassification of a patient’s cognitive status.
The appropriateness, feasibility, and accuracy of the “screen-metric” approach of the

Cognistat in identifying cognitive deficits has also been scrutinised in recent decades.
Although the ‘screen’ question is typically of greater difficulty than subsequent ‘metric’
items, Oehlert et al. (1997) argue the recommended “screen-metric” format has the
potential to overlook possible cognitive deficits, particularly when screen questions are
only administered. In agreement with this view, Rice et al. (2015) found the “screen-
metric” approach failed to adequately detect cognitive deficits in a large sample of stroke
rehabilitation patients when screen questions were only administered.
109
Chapter 2.
It has also been reported that approximately 20% of normal cognitively-intact adults fail
screen items when the routine screen-metric format is deployed (Kiernan et al., 1987).
While administering all screen and metric items extends the average administration time
to roughly 30 minutes, Oehlert et al. (1997), Van Gorp et al. (1997) and Rice et al. (2015)
argue administrating the Cognistat in its entirety (screen and metric items) minimises the
likelihood of overlooking potential cognitive deficits, improves test reliability, and
enhances the overall predictability of the battery. Therefore, to reduce the possibility of
missing potential cognitive dysfunction, particularly the subtle cognitive dysfunction
associated with diabetes and hypertension, all questions of the Cognistat (screen and
metric) were administered in the present investigation.
Although some researchers question the diagnostic utility of the Cognistat as a screening
tool for cognitive impairment, most researchers laud the neuropsychological battery for
its conciseness, broad range of cognitive domains assessed, brief administration time, low
cost, and its multidimensional construct (Engelhart et al., 1994; Whiteside et al., 1996;
Eisenstein et al., 2002; Brown, Mapleston, & Nairn, 2011). As mentioned earlier,
dependent on the score obtained in each subtest, scores in the Cognistat are plotted on a
cognitive status profile. Performance is then subsequently graded on a scale: average,
mild, moderate or severe impairment (Figure 2.12) (Kiernan et al., 1987). This allows for
rapid identification of cognitively-intact areas and alerts clinicians to evidence of any
potential cognitive deficits present (Kiernan et al., 1987). Unlike the MMSE, each
cognitive domain in the Cognistat has a domain-specific impairment threshold score.
Scores in proximity to or beneath these thresholds indicate a degree/gradation (mild,
moderate, or severe) of cognitive impairment. Maximum obtainable scores, as well as
domain-specific cut-off scores indicative of impairment for each respective cognitive
domain, are presented below (Table 2.2). As previous research suggests Cognistat test
accuracy and sensitivity improves when administered in concert with other cognitive
screening tools (Schwamm et al., 1987; Macaulay et al., 2003), both the MMSE and
Cognistat were administered conjointly in the present study.
110
Chapter 2.
Table 2.2. Maximum achievable scores and domain-specific impairment threshold

scores for each cognitive domain of the Cognistat. Adapted and modified
from (Kiernan et al., 1987).
Grade of Cognitive Impairment

Maximum Impairment
Cognitive Domain Mild Moderate Severe
Score Threshold
Orientation 12 < 10 8 6 4
Attention 8 <6 5 3 1
Comprehension 6 <5 4 3 2
Language Repetition 12 < 11 9 7 5
Naming 8 <7 5 3 2
Construction 6 <4 3 2 0
Memory 12 < 10 8 6 4
Calculation 4 <3 2 1 0
Similarities 8 <5 4 3 2
Reasoning
Judgement 6 <4 3 2 1
Total 82 < 65 51 36 21
After completion of the cognitive assessment, three post-study BP measurements as well

as one post-study BGL measurement were recorded (as outlined in the experimental
protocol detailed in section 2.4) and this concluded the experimental testing (Figure 2.14).
Participants were then acknowledged for their participation in the study with a cognitive
profile (Appendix 8.5) and remuneration (clinical cohorts) (Appendix 8.8) and were
supplied with a copy of the signed consent form signed at the beginning of the study
(Appendix 8.1), with a copy also retained by the investigator. Finally, in accordance with
the UTS HREC requirements, the researcher completed a study summary sheet, reporting
on how the study was conducted and any unusual events that occurred throughout the
investigation (Appendix 8.6).
111
Chapter 2.
Figure 2.14. Present study experimental protocol
Cognitive Testing
Pre-Study*# Baseline EEG Active EEG Psychometrics Post-Study

Subject relaxes, Subject completes MMSE (Folstein et al.,
BP, HR observing blank Stroop Colour Word 1975) BP, HR
computer screen Test (Stroop, 1935) Cognistat (Kiernan et and BGL
and BGL
(5 mins) (5 mins) al., 1987)
(32-channel EEG) (32-channel EEG)
Key:
* – Lifestyle Appraisal Questionnaire

# – In-house Chronic Disease Questionnaire
BP – Blood Pressure
BGL – Blood Glucose Level
EEG – Electroencephalography
HR – Heart Rate
MMSE – Mini Mental State Examination
112
Chapter 2.
2.10 Data Processing and Analysis

Data analysed in the present cross-sectional investigation included:
• Demographic data (age, BMI, years of education, LAQ Part 1, LAQ Part 2)
(Craig, Hancock, & Craig, 1996)
• Clinical data (solicited by questionnaires developed in-house as part of this
research)
• Cardiovascular variables (pre-study and post-study SBP and DBP)
• Blood glucose concentrations (determined pre- and post-study)
• Electroencephalography data (baseline, active, (Stroop Test)) (Stroop, 1935)
• Cognitive function, assessed by the Mini-Mental State Examination (MMSE)
(Folstein, Folstein, & McHugh, 1975) and the Cognistat (Kiernan et al., 1987)
2.10.1 Electroencephalography Data Pre-Processing

Several artefacts can contaminate the electroencephalography signal (e.g. direct current,
movement, non-electrically shielded room, etc.), resulting in unreliable data
(Kaiboriboon et al., 2012). Therefore, all raw EEG data obtained (during baseline and
active recordings) was subjected to pre-processing (noise-reduction) prior to statistical
analysis (refer to Section 2.11) according to the steps below. Investigators suggest this
improves the signal-to-noise ratio (SNR) (Kaiboriboon et al., 2012).
1. Direct current (DC) interference or sources of high-frequency movement artefacts
(e.g. muscular, fast-paced movements) were eliminated using a Butterworth IIR
bandpass filter set at 1.5 and 50 Hz.
2. Ocular artefacts, such as blinking, were attenuated by applying an electro-
oculography algorithm (aligned-artefact average procedure).
3. 5-minute baseline and active EEG recordings were subsequently divided into 300
one-second epochs.
4. The individual epoch values were then examined for outliers, which were
removed using the modified Z-score statistic (epoch values ≥10 excluded)
(Maharaj, Lees, & Lal, 2019). Previous studies in our research unit have used this
method (Maharaj, Lees, & Lal, 2019). The modified Z-score statistic was
calculated using the following equation:
113
Chapter 2.
X = epoch value x̃ = median MAD = Median Absolute Deviation
The equation (below) can be used to derive the median absolute deviation:
X = epoch value x̃ = median
5. A fast-Fourier transform application (FFT) then enabled the derivation of average

EEG activity for each individual frequency band (delta: 1 – 4 Hz; theta: 4 – 8 Hz;
alpha: 8 –12 Hz; beta: 12 – 30 Hz; and gamma: 35 – 100 Hz) (Modi & Sahin,
2017).
All EEG values were recorded in microvolts per second squared (µV/s2).
114
Chapter 2.
2.11 Statistical Analysis
2.11.1 Power Analysis

Statistical power refers to the probability of rejecting correctly the null hypothesis when
it is false (i.e. not making a false negative) (Button et al., 2013). High statistical power
reduces Type II errors and increases the likelihood of identifying true effects and/or
relationships (Biau et al., 2008; Button et al., 2013). Cohen (1992) determined that the
minimum sample size required for the analyses in the present research, with adequate
power (0.8) and moderate to large effect size (0.5 – 0.8), is ~30. In groups where n ≤30
(T1DM, T2DM, and HTN), appropriate non-parametric analyses were performed.
Previous studies conducted in our research unit (Rothberg et al., 2016; Maharaj, Lees, &
Lal, 2019; Chalmers et al., 2020) and the literature (Pramming et al., 1988; Tallroth et
al., 1990; Bjorgaas et al., 1998; Howorka et al., 2000) have utilised smaller or similar
sample sizes to that reported in the present thesis.
2.11.2 Dependent Sample T-test

T-tests detect significant differences in means of paired (dependent) and unpaired
(independent) samples (Lund Research Ltd, 2019). Dependent sample t-tests (paired)
were conducted in the non-clinical group to determine significant differences between the
following variables: pre-study and post-study blood pressure (SBP and DBP); and pre-
study and post-study blood glucose level (BGL).
2.11.3 Wilcoxon Signed Rank Test

The Wilcoxon Signed Rank Test is a robust non-parametric equivalent of a dependent
sample t-test (Lund Research Ltd, 2019). It determines significant differences in the
medians of paired non-normally distributed data (skewed or ranked). This test was
performed to identify significant differences between pre-study and post-study
physiological variables (SBP, DBP, and BGL) in the clinical samples (T1DM, T2DM,
and HTN), with n ≤30.
115
Chapter 2.
2.11.4 Mann Whitney U Test

The Mann Whitney U Test is a non-parametric equivalent of an independent sample t-
test that identifies significant differences in the ranks of non-normally distributed data in
unpaired samples (Lund Research Ltd, 2019). This test was conducted to identify
significant differences in pre-study and post-study physiological variables (SBP, DBP,
and BGL) between the clinical samples (T1DM, T2DM, and HTN).
2.11.5 Partial Pearson’s Correlation

A partial Pearson’s correlation examines the strength and direction of an association
between two continuous dependent variables in a sample while controlling for
confounding variables (Lund Research Ltd, 2019). A partial Pearson’s correlation
controlling for age and BMI was applied to determine associations of BP and BGL with
cognitive function (EEG and cognitive measures, MMSE and the Cognistat) in the non-
clinical population. The literature indicates these lifestyle risk factors influence cognitive
outcomes (Kivipelto et al., 2018).
The Pearson’s correlation generates an r (rho) value (the Pearson’s correlation

coefficient). This value ranges between r = 1 and r = -1 and indicates the strength and
direction of the association. Positive r values indicate positive relationships (i.e. as one
variable increases, the other variable also increases), while negative r values indicate
negative associations (i.e. as one variable decreases, the other increases). No consensus
exists for specific cut-off values for different strengths of association, but the following
values are generally accepted: negligible (± 0.0 – 0.3), low (± 0.3 – 0.5), moderate (± 0.5
– 0.7), strong (± 0.7 – 0.9) (Hinkle et al., 2003).
2.11.6 Spearman’s Rank-Order Correlation

Spearman’s rank-order correlation is a non-parametric equivalent of Pearson’s correlation
(Lund Research Ltd, 2019). Similarly, it determines the strength and direction of an
association/relationship between two continuous variables in a sample. However, unlike
Pearson’s correlation, which uses individual data values, Spearman’s rank-order
correlation uses the ranks of data. This statistical test was applied to investigate
associations in study samples with n ≤30 participants (the clinical cohorts –T1DM,
T2DM, and HTN).
116
Chapter 2.
2.11.7 Multiple Analysis of Covariance (MANCOVA)

The multiple analysis of covariance (MANCOVA) is considered an extension of the
ANOVA and compares differences in means of multiple independent samples while
controlling for confounding variables (Lund Research Ltd, 2019). This test was applied
to determine differences in global and domain-specific cognitive variables between the
non-clinical and clinical cohorts while controlling for covariates (e.g. age and BMI). As
the MANCOVA is an omnibus test, post hoc Tukey tests were performed subsequently
to ascertain where significant differences, if any, occurred between the groups.
117
��
��
��
�
��
��
��
��
��
��
�
��
��
��
��
�
�
��
��
� �
��
�
Chapter 3.
3.1 Participant Summary

A total of ninety-four (n = 94) study participants, aged between 18-80 years, were
recruited from the local Sydney community for involvement in the present cross-sectional
investigation. The study cohort was comprised of the following four groups: (1) non-
clinical (n = 49, mean age: 31.3 ± 15.9 years), (2) Type 1 diabetes mellitus (T1DM) (n =
13, mean age: 35.1 ± 16.1 years), (3) Type 2 diabetes mellitus (T2DM) (n = 17, mean
age: 54.7 ± 12.1 years), and (4) hypertension (HTN) (n = 15, mean age: 61.0 ± 16.9 years).
All participants from the clinical populations (T1DM, T2DM, HTN) reported medication
use for their respective conditions (glucose-lowering or anti-hypertensive medication)
(see Section 8.9). Age and BMI differed significantly between the groups and were
therefore used as covariates in the final analyses.
3.2 Demographic Variables

Demographic characteristics (age, BMI, LAQ Part 1, LAQ Part 2, and years of education)
for all participants were obtained using a reliable and validated questionnaire (the
Lifestyle Appraisal Questionnaire (LAQ) (Craig, Hancock, & Craig, 1996). Age, BMI,
and LAQ Part 1 were determined to be significantly different between the groups (Table
3.1). The age of the populations ranged between 30-61 years, with the non-clinical group
being the youngest (31.3 ± 15.8 years) and the hypertension group (61.0 ± 16.9 years) as
the older cohort. The age of the T1DM (35.1 ± 16.1 years) and T2DM (54.7 ± 11.8 years)
cohorts fell between the non-clinical and HTN sample groups. Analysis revealed age
differed significantly between the following groups: non-clinical and T2DM (p<0.001);
non-clinical and hypertension (p <0.001); T1DM and T2DM (p = 0.01) and T1DM and
hypertension (p <0.001). The BMI of participants from the non-clinical (24.6 ± 5.1 kg/m2)
and T1DM (23.8 ± 1.60 kg/m2) groups resided within the healthy range (BMI: 18.5 –
24.9) (Australian Government Department of Health, 2020). In contrast, it fell within the
unhealthy range (BMI: > 25) (Australian Government Department of Health, 2020) for
both the T2DM (31.7 ± 1.4 kg/m2) and HTN groups (28.1 ± 1.50 kg/m2), with the T2DM
group demonstrating the highest BMI. A significant difference in BMI was also found
between the groups (p <0.001). Post hoc analysis showed this difference occurred
between the following groups: non-clinical and T2DM (p <0.001); non-clinical and
hypertension (p = 0.04) and T1DM and T2DM (p <0.001).
119
Chapter 3.
In relation to participant lifestyle risk factors (measured using the Lifestyle Appraisal
Questionnaire (LAQ Part 1)), scores ranged between 11-20 points, with the non-clinical
group scoring the lowest (11.7 ± 6.6) and the HTN population scoring the highest (19.7
± 4.4). A significant difference was observed in LAQ Part 1 score between the groups
(p<0.001). Post hoc analysis revealed this difference was between the non-clinical and
T2DM (p<0.001) and non-clinical and HTN (p <0.001) groups. While LAQ Part 2 scores
and years of education varied slightly between the groups, the differences were not
statistically significant.
120
Chapter 3.
Table 3.1. Key demographic characteristics (age, BMI, LAQ Part 1, LAQ Part 2,
and years of education) for all groups. P values are parametric
multivariate.
Demographic
Sample Group Mean ± SD p
Variable
Non-clinical 31.3 ± 15.8
T1DM 35.1 ± 16.1
Age (Yrs) <0.001*
T2DM 54.7 ± 11.8
HTN 61.0 ± 16.9
T1DM 23.8 ± 8.5
BMI (kg/m2) <0.001*
T2DM 31.7 ± 5.6
HTN 28.1 ± 5.2
T1DM 15.3 ± 6.9
LAQ Part 1 <0.001*
T2DM 19.4 ± 6.3
HTN 19.7 ± 4.4
T1DM 22.4 ± 10.9
LAQ Part 2 0.22
T2DM 19.9 ± 9.3
HTN 15.5 ± 12.5
Years of Education T1DM 21.0 ± 13.1
0.27
(Yrs) T2DM 21.7 ± 13.7
HTN 16.6 ± 4.9
Key:
Yrs – Years kg/m2 – kilograms per metre squared
LAQ – Lifestyle Appraisal Questionnaire T1DM – Type 1 diabetes mellitus
T2DM – Type 2 diabetes mellitus HTN – Hypertension
* – statistical significance p – p-value
121
Chapter 3.
3.2.1 Cardiovascular Variables (SBP, DBP, and HR)

Blood pressure (BP) and heart rate (HR) were recorded both before and after cognitive
testing (see section 2.5 under methods for BP protocol). The change (Δ) that occurred in
each variable (BP and HR) was calculated by subtracting the pre-study value from the
post-study value. Generally, blood pressure (SBP and DBP) increased after the study in
all groups, whereas heart rate (HR) decreased. Blood pressure (SBP and DBP) was found
to be highest in the HTN population and lowest in the non-clinical group (Table 3.2).
With respect to systolic blood pressure (SBP), a significant difference was found between
groups for both pre-study (p <0.001) and post-study SBP (p <0.001). For pre-study SBP,
post hoc analysis revealed differences between the following groups: non-clinical and
T1DM (p = 0.02), non-clinical and HTN (p <0.001), and T2DM and HTN (p = 0.02). For
post-study SBP, significant differences were found between the following groups: non-
clinical and T1DM (p <0.05), non-clinical and T2DM (p = 0.01), non-clinical and HTN
(p <0.001); T1DM and HTN (p = 0.02); and T2DM and HTN (p <0.001). Interestingly,
no significant difference was found within each of the groups between pre-study and post-
study SBP for all groups.
Similarly, significant differences between groups were observed for pre-study (p <0.05)
and post-study DBP (p <0.001). For pre-study DBP, post hoc analysis indicated
significant difference between the following groups: non-clinical and T1DM (p = 0.02),
non-clinical and T2DM (p = 0.03), and non-clinical and HTN (p <0.05). For post-study
DBP, it was found between the non-clinical and T1DM group (p <0.05) and non-clinical
and HTN (p <0.001) groups. Similar to SBP, no significant differences were observed
between pre-study and post-study DBP within the four groups.
122
Chapter 3.
Although pre-study HR varied slightly between the groups, significance was not reached;
however, an overall significant difference was found between the groups in post-study
HR (p <0.05). Post hoc analysis revealed this significance was between the following
groups: non-clinical and T1DM (p <0.05), non-clinical and T2DM (p = 0.01), T1DM and
HTN (p = 0.03), and T2DM and HTN (p = 0.04). Significant within-group differences
between pre-study and post-study HR were found in the following groups: non-clinical
(p <0.001), T2DM (p = 0.01), and HTN (p = 0.03).
123
Chapter 3.
Table 3.2. Pre-study and post-study cardiovascular variables (SBP, DBP and HR), as well as the change that occurred, for each group. P-
values are parametric multivariate.
Δ (post-study –
Cardiovascular Sample
Pre-study p Post-study p pre-study) p
Variable Group
Non-clinical 112.5 ± 12.1 111.9 ± 13.4 - 0.55 ± 7.3 0.60

T1DM 125.9 ± 18.5 126.5 ± 17.2 0.62 ± 12.3 0.97
SBP (mm Hg) <0.001* <0.001*
T2DM 120.2 ± 12.9 123.1 ± 11.5 2.7 ± 6.4 0.10
HTN 135.7 ± 21.0 140.1 ±19.8 4.4 ± 8.1 0.07
Non-clinical 73.7 ± 9.7 75.0 ± 9.4 1.3 ± 5.4 0.09
T1DM 81.2 ± 10.0 82.9 ± 10.1 1.7 ± 6.2 0.35
DBP (mm Hg) <0.05* <0.001*
T2DM 79.9 ± 6.3 80.3 ± 7.1 0.30 ± 4.5 0.73
HTN 82.9 ± 12.2 85.5 ± 10.0 2.5 ± 5.3 0.11
Non-clinical 70.0 ± 9.0 64.9 ± 7.4 - 5.1 ± 6.9 <0.001*
T1DM 74.2 ± 9.0 72.9 ± 8.7 - 1.3 ± 4.0 0.20
HR (bpm) 0.16 <0.05*
T2DM 75.2 ± 10.6 71.8 ± 9.5 - 3.4 ± 4.5 0.01*
HTN 70.5 ± 10.5 65.9 ± 8.7 - 4.6 ± 7.4 0.03*
Key:
SBP – systolic blood pressure DBP – diastolic blood pressure HR – heart rate mm Hg – millimetres of mercury
bpm – beats per minute Δ – change T1DM – Type 1 diabetes mellitus
T2DM – Type 2 diabetes mellitus * - statistical significance HTN – hypertension p – p-value
124
Chapter 3.
3.2.2 Blood Glucose Level (BGL)

Two-hour (2-hr) fasting blood glucose concentrations were determined for each
participant before cognitive testing (refer to section 2.6 under methodology for BGL
measurement protocol). The change (Δ) that occurred in BGL was calculated by
subtracting the pre-study value from the post-study value. All BGL values were reported
in millimoles per litre (mmol/L) (Table 3.3). Excluding the HTN population where BGL
increased marginally, BGL generally decreased after the study in all groups. It was within
the normal range (3.5 – 8 mmol/L) (ADA, 2020) for both the non-clinical and HTN
groups but was slightly elevated in the T1DM and T2DM groups, confirming diabetes
status.
Significant differences were found between pre-study (p <0.001) and post-study (p

<0.001) BGL in all the groups. For pre-study BGL, post hoc analysis revealed the
differences occurred between the following groups: non-clinical and T1DM (p <0.001),
non-clinical and T2DM (p <0.001), T1DM and HTN (p <0.001), and T2DM and HTN (p
<0.001). Conversely, for post-study BGL, the differences were found between the non-
clinical and T1DM (p <0.05), non-clinical and T2DM (p <0.001), and T2DM and HTN
(p <0.001). Analysis also revealed significant within-group differences between pre-study
and post-study BGL in non-clinical (p = 0.01) and T2DM groups (p = 0.01).
125
Chapter 3.
Table 3.3. Pre-study and post-study BGL, as well as the change in BGL that occurred, in all groups.
Physiological Sample Δ (post-study

Pre-Study p Post-Study p p
Variable Group – pre-study)
Non-clinical 5.2 ± 0.72 4.9 ± 0.58 - 0.26 ± 0.65 0.01*

BGL T1DM 8.0 ± 2.9 6.9 ± 2.6 - 1.07 ± 4.1 0.40
<0.001* <0.001*
(mmol/L) T2DM 9.0 ± 4.1 7.7 ± 3.2 - 1.36 ± 1.43 0.01*
HTN 5.2 ± 0.66 5.3 ± 0.58 0.1 ± 0.64 0.66
Key:
BGL – blood glucose level mmol/L – millimoles per litre Δ – change T1DM – Type 1 diabetes mellitus
T2DM – Type 2 diabetes mellitus HTN – Hypertension * – statistical significance p – p-value
126
Chapter 3.
3.2.3 Disease-specific variables (disease duration,

glycosylated haemoglobin)
Disease-specific variables, such as disease duration (chronicity) and age of disease onset,
were reported for the clinical populations (T1DM, T2DM, and HTN) in the present study
via a questionnaire developed in-house (see Appendix 9.3 and 9.4). Glycosylated
haemoglobin (HbA1C) was also provided. These variables were obtained as they have
been shown to moderate the relationship between these chronic diseases and cognition
(Biessels et al., 2014; Feinkohl et al., 2015; Biessels & Despa, 2018). Although frequency
and severity of hypoglycaemia (mild or severe) (for T1DM and T2DM) and duration of
disease (for HTN) were solicited, it could not be recalled reliably by study participants
and was therefore not reported. Table 3.4 summarises key disease-specific variables
sought from participants in the clinical samples.
Table 3.4. Disease-specific variables solicited from the clinical groups.
Variable Sample Group Value p
T1DM 6.9 %
HbA1C 0.27
T2DM 8.1 %
T1DM 17.8 ± 9.2

Disease Duration
0.02*
(Yrs)
T2DM 11.3 ± 5.9
T1DM 16.8 ± 14.9

Disease Onset
<0.001*
(Yrs)
T2DM 44.9 ± 14.7
Key:
HbA1C – glycosylated haemoglobin Yrs – years
T1DM – Type 1 diabetes mellitus T2DM – Type 2 diabetes mellitus
p – p-value * – statistical significance
127
Chapter 3.
The glycaemic control of patients with DM was slightly elevated (HbA1C: 7 – 8 %) (ADA,
2020), with HbA1C ranging between 7-8% (T1DM: 6.9%; T2DM: 8.0%) (Table 3.4). A
Wilcoxon signed rank test revealed no significant difference in glycosylated haemoglobin
between the T1DM and T2DM groups. On average, patients with T1DM generally had
diabetes for longer (mean duration: 17.8 ± 9.2 years) than those with T2DM (mean
duration: 11.3 ± 5.9 years) or HTN (7.5 ± 2.9 years). Most patients with T1DM developed
their condition after 7 years of age or earlier (disease onset: 16.8 ± 14.9 years) than those
with T2DM (disease onset: 44.9 ± 14.4 years), with analysis revealing a significant
difference in disease onset between T1DM and T2DM (p <0.001).
128
Chapter 4.
4. Associations between blood pressure, blood

glucose level and cognitive performance (Non-clinical
and Clinical)
This chapter reports the cognitive performance data (global and domain-specific) for all
four samples (non-clinical, T1DM, T2DM, and HTN), assessed using the reliable and
validated cognitive screening tools, the Mini-Mental State Examination (Folstein,
Folstein, & McHugh, 1975) and the Cognistat (Kiernan et al., 1987). It also reports the
performance of all groups for the Stroop Colour Word Test (Stroop, 1935), which was
used to simulate cognitive activity (measured using EEG) (see section 2.8 under
methodology for details). Associations between BP (SBP and DBP), BGL, and disease-
specific variables (disease duration and HbA1c level for the clinical groups) and cognitive
function are also reported for each group. All data are presented as mean ± SD.
4.1 Cognitive Performance
4.1.1 Global Cognitive Performance (Mini-Mental State Examination) (Folstein,

Folstein, & McHugh, 1975)
With respect to global cognitive performance, as assessed using the Mini-Mental State
Examination (MMSE), all groups performed above the impairment threshold for the
assessment (scores ≤ 23 indicative of cognitive dysfunction) (Table 4.1). Interestingly,
the T1DM group performed the strongest (28.5 ± 1.1), whereas the T2DM group
performed the worst (27.5 ± 2.9). The global cognitive performance scores of the non-
clinical and HTN groups resided between that of the T1DM and T2DM groups, scoring
28.3 ± 1.6 and 28.0 ± 1.5, respectively. Although global cognitive performance varied
slightly between the groups, it was not statistically significant.
129
Chapter 4.
Table 4.1. Mean scores obtained in the Mini-Mental State Examination by each of the
study sample groups.
Variable Sample Group Score p
T1DM 28.5 ± 1.1
MMSE 0.89
T2DM 27.5 ± 2.9
HTN 28.0 ± 1.5
Key:
MMSE – Mini-Mental State Examination T1DM – Type 1 diabetes mellitus
T2DM – Type 2 diabetes mellitus HTN – Hypertension
p – p-value
4.1.2 Domain-specific Cognitive Performance (Cognistat) (Kiernan et al., 1987)
In relation to domain-specific cognitive performance, as assessed using the Cognistat, all

groups performed above the impairment threshold (> 65) for the Cognistat, as well as for
each domain-specific impairment threshold (refer to section 2.9.2), confirming intact
cognitive function within the study cohort. Performance in most cognitive domains was
similar between the groups, with the non-clinical and T1DM groups demonstrating
consistently strong cognitive performance. Interestingly, performance differed
considerably in the memory domain, with the T1DM group performing the best (11.3 ±
0.95) and the HTN group performing the worst (9.2 ± 2.9). However, no significant
difference in domain-specific cognitive performance was observed between the groups
(Table 4.2).
130
Chapter 4.
Table 4.2. Mean scores obtained by each sample group (non-clinical, T1DM, T2DM,
and HTN) for individual domains of the Cognistat, as well as the total
Cognistat score.
Variable Sample Group Score p
T1DM 11.6 ± 0.77
Orientation 0.58
T2DM 11.3 ± 0.69
HTN 11.5 ± 1.5
T1DM 7.5 ± 0.66
Attention 0.92
T2DM 6.9 ± 1.1
HTN 7.3 ± 1.0
T1DM 5.8 ± 0.38
Comprehension 0.40
T2DM 5.8 ± 0.56
HTN 5.5 ± 0.64
T1DM 11.9 ± 0.28
Repetition 0.07
T2DM 11.8 ± 0.56
HTN 11.1 ± 1.36
T1DM 7.9 ± 0.28
Naming 0.73
T2DM 7.6 ± 0.87
HTN 7.5 ± 1.13
T1DM 5.7 ± 0.63
Construction 0.42
T2DM 5.4 ± 0.94
HTN 5.1 ± 0.88
Memory T1DM 11.3 ± 0.95 0.49

T2DM 9.4 ± 3.5
131
Chapter 4.
HTN 9.2 ± 2.9

T1DM 3.6 ± 0.65
Calculation 0.84
T2DM 3.6 ± 0.86
HTN 3.8 ± 0.41
T1DM 6.9 ± 1.3
Similarities 0.79
T2DM 6.6 ± 2.1
HTN 7.1 ± 1.1
T1DM 3.9 ± 0.76
Judgement 0.98
T2DM 4.1 ± 0.78
HTN 4.1 ± 1.2
T1DM 76.3 ± 2.90
Total Cognistat 0.76
T2DM 72.4 ± 6.7
HTN 72.7 ± 4.64
Key:
T1DM – Type 1 diabetes mellitus T2DM – Type 2 diabetes mellitus
HTN – Hypertension p – p-value
132
Chapter 4.
4.1.3 Stroop Colour Word Test (Stroop, 1935)

A significant difference was found in average response time (measured in milliseconds)
between the sample groups (p = 0.01) for matched stimuli in the Stroop Test (e.g. the
word BLUE printed in blue ink) (refer to Table 4.3). The non-clinical group displayed
the fastest average response time (1213.67 ± 240.9 ms), whereas the T2DM group
demonstrated the slowest response time (1779.9 ± 631.0 ms). The average response times
for the T1DM (1246.00 ± 171.27 ms) and HTN (1523.33 ± 364.96 ms) groups fell
between those of the non-clinical and the T2DM groups. Post hoc analysis also indicated
significant differences in response time for matched stimuli in the Stroop test between the
following groups: non-clinical and T2DM (p<0.001), T1DM and T2DM (p = 0.01), and
T2DM and HTN (p = 0.04).
A similar pattern in average response time was observed for mismatched stimuli in the
Stroop Test (e.g. the word BLUE printed in yellow ink). The non-clinical group showed
the fastest average response time (1440.25 ± 281.17 ms), whereas the T2DM group
demonstrated the slowest response time (2202.62 ± 765.84 ms). Similarly, response times
for the T1DM (1445.66 ± 227.94 ms) and HTN (1849.16 ± 387.09 ms) groups fell
between those of the non-clinical and T2DM groups. A significant difference in average
response time for mismatched stimuli in the Stroop test was found between the groups (p
< 0.001). Post hoc analysis revealed significant differences between the following groups:
non-clinical and T2DM (p <0.001), T1DM and T2DM (p <0.001), and T2DM and HTN
(p = 0.01).
Interestingly, no significant associations were found between the pre-study and post-
study physiological variables (BP and BGL) and the matched and mismatched aspects of
the Stroop Colour Word Test for the non-clinical and clinical groups. While no significant
associations were identified between disease-specific variables (HbA1C, age of disease
onset, disease duration) and the matched and mismatched aspects of the Stroop Colour
Word Test for the T1DM and HTN groups, a significant association was found between
age of disease onset and average response time for matched stimuli (p = 0.01) for the
T2DM group (Table 4.4) (Figure 4.1).
133
Chapter 4.
Table 4.3. Mean average response times obtained for each group (non-clinical,
T1DM, T2DM, and HTN) for matched and mismatched stimuli of the
Stroop Colour Word Test (Stroop, 1935).
Average Response
Variable Sample Group p
Time (ms)
Non-clinical 1213.67 ± 240.9
Matched T1DM 1246.00 ± 171.27
0.01*
Stimuli T2DM 1779.9 ± 631.0
HTN 1523.33 ± 364.96
Non-clinical 1440.25 ± 281.17
Mismatched T1DM 1445.66 ± 227.94
<0.001*
Stimuli T2DM 2202.62 ± 765.84
HTN 1849.16 ± 387.09
Key:
T1DM – Type 1 Diabetes Mellitus T2DM – Type 2 Diabetes Mellitus
HTN – Hypertension ms – milliseconds
* – statistical significance p – p-value
134
Chapter 4.
Table 4.4. Associations between disease-specific variables (HbA1C, age of disease

onset, and disease duration) and matched aspects of the Stroop Colour Word
Test.
Dependent Independent
Group p r
Variable Variable
HbA1C T1DM 0.76 - 0.10

T2DM 0.14 - 0.53
T1DM 0.76 - 0.09
Matched Disease Duration
T2DM 0.52 - 0.17
T1DM 0.09 0.49
Disease Onset
T2DM 0.01* 0.65
Key:
T1DM – Type 1 Diabetes Mellitus T2DM – Type 2 Diabetes Mellitus
HbA1C – glycosylated haemoglobin * – statistical significance
p – p-value r – rho value
135
Chapter 4.
p = 0.01; r = 0.65
Figure 4.1. Positive correlation between age of disease onset and average response
time for matched stimuli in the Stroop Colour Word Test for the Type 2
diabetes mellitus group.
Key:
ms – milliseconds p – p-value r – rho value
136
Chapter 4.
4.2 Associations between BP and cognition for the non-clinical sample group
A partial Pearson’s correlation was conducted to identify associations between pre-study
and post-study BP (SBP and DBP) and cognition (MMSE and Cognistat) in the non-
clinical group. No significant associations were found between SBP and DBP (pre-study
or post-study) and total MMSE score. There were also no significant associations between
pre-study and post-study BP variables (SBP and DBP) with individual cognitive domain
scores or the total Cognistat score.
4.3 Associations between BGL and cognition for the non-clinical sample group
No significant associations were found between pre-study BGL and total MMSE score in
the non-clinical group; however, post-study BGL was inversely associated with total
MMSE score (p = 0.03; r = - 0.32) (Figure 4.2). No significant associations were observed
between BGL (pre-study and post-study) and individual cognitive domain scores or the
total Cognistat score.
137
Chapter 4.
p = 0.03; r = - 0.32
Figure 4.2. Inverse correlation between post-study blood glucose level (BGL) and
total Mini-Mental State Examination score for the non-clinical group.
Key:
BGL – blood glucose level mmol/L – millimoles per litre
138
Chapter 4.
4.4 Associations between BP, BGL and cognition for the T1DM sample group
A Spearman’s Rank-Order correlation was performed to identify associations between
pre-study and post-study physiological variables (SBP, DBP, and BGL) and cognitive
function (MMSE and the Cognistat) in the T1DM group. No significant associations were
observed between SBP (pre-study and post-study) and cognitive performance; however,
pre-study DBP was significantly associated with performance in the similarities domain
of the Cognistat (p = 0.02; r = 0.62), while post-study DBP was significantly associated
with judgement performance in the Cognistat (p = 0.04; r = 0.56) (Figure 4.3). There were
no significant associations between BGL (pre-study and post-study) and individual
cognitive domain scores and the total Cognistat score.
139
Chapter 4.
p = 0.04; r = 0.56
Figure 4.3. Positive correlation between post-study diastolic blood pressure and
judgement performance in the Cognistat for the Type 1 diabetes mellitus
sample group.
Key:
DBP – diastolic blood pressure mm Hg – millimetres of mercury
With respect to disease-specific variables in the T1DM group, HbA1C level was
significantly associated with the comprehension domain of the Cognistat (p = 0.04; r =
0.58), but no other domains. The age of disease onset was also found to be significantly
associated with the similarities domain of the Cognistat (p = 0.02; r = 0.65). However, no
association was found between disease duration and global cognitive performance
(MMSE) or domain-specific cognitive performance.
140
Chapter 4.
4.5 Associations between BP, BGL and cognition for the T2DM sample group
A Spearman’s Rank-Order correlation was similarly applied to identify associations
between pre-study and post-study BP (SBP and DBP) and BGL and cognitive
performance (MMSE and the Cognistat) in the T2DM group. No significant associations
were found between pre-study SBP and cognitive measures, but post-study SBP was
significantly associated with construction performance in the Cognistat (p = 0.04; r =
0.52) (Figure 4.4). Similarly, no significant associations were observed between pre-
study DBP and cognitive function (global and domain-specific); however, post-study
DBP was significantly associated with the attention (p = 0.04; r = 0.52) and judgement (p
= 0.01; r = 0.65) (Figure 4.5) domains of the Cognistat, as well as the total Cognistat score
(p = 0.02; r = 0.61) (Figure 4.6).
141
Chapter 4.
p = 0.04; r = 0.52
Figure 4.4. Positive correlation between post-study systolic blood pressure and
construction performance in the Cognistat for the Type 2 diabetes
mellitus sample group.
Key:
SBP – systolic blood pressure mm Hg – millimetres of mercury
142
Chapter 4.
p = 0.04; r = 0.52
judgement performance in the Cognistat for the Type 2 diabetes mellitus
sample group.
Key:
143
Chapter 4.
p = 0.01; r = 0.65
Figure 4.6. Positive correlation between post-study diastolic blood pressure and total
Cognistat score for the Type 2 diabetes mellitus sample group.
Key:
Regarding BGL, pre-study BGL was significantly associated with the attention (p = 0.04;
r = - 0.52) (Figure 4.7) and memory domains (p = 0.04; r = 0.52) of the Cognistat. In
contrast, post-study BGL was significantly associated with memory performance in the
Cognistat (p <0.05; r = 0.70). No other significant associations were found between post-
study BGL and global and domain-specific cognitive performance. There were also no
significant associations observed between disease-specific variables (HbA1C, disease
duration, and age of disease onset) and global and domain-specific cognitive performance
for the T2DM group.
144
Chapter 4.
p = 0.04; r = - 0.52
Figure 4.7. Inverse correlation between pre-study blood glucose level and the
attention domain of the Cognistat for the Type 2 diabetes mellitus sample
group.
Key:
4.6 Associations between BP, BGL and cognition for the HTN sample group
A Spearman’s Rank Order correlation was also conducted to identify associations
between pre-study and post-study physiological variables (SBP, DBP, and BGL) and
global and domain-specific cognitive performance in the HTN group. No significant
associations were found between pre-study and post-study SBP and total MMSE score
and total score of the Cognsitat. Similarly, no significant associations were found between
pre-study and post-study DBP and the cognitive measures.
145
Chapter 4.
With respect to BGL, pre-study BGL was inversely associated with performance in the
construction (p = 0.01; r = - 0.67) and similarities (p= 0.03; r = - 0.56) (Figure 4.8)
domains of the Cognistat. Conversely, post-study BGL was significantly associated with
memory performance (p= 0.01; r = 0.69) in the Cognistat.
p = 0.03; r = - 0.56
Figure 4.8. Inverse correlation between pre-study blood glucose level and the
similarities domain of the Cognistat for the hypertension sample group.
Key:
146
Chapter 4.
4.7 Discussion: Cognitive Performance of Non-Clinical and Clinical

Populations
This chapter discusses findings concerning the cognitive performance (global and
domain-specific) of both the non-clinical and clinical sample groups, as assessed using
the reliable and validated cognitive screening tools, the Mini-Mental State Examination
(MMSE) (Folstein, Folstein, & McHugh, 1975) and the Cognistat (Kiernan et al., 1987).
Associations identified between pre-study and post-study physiological variables (SBP,
DBP, and BGL) and global and domain-specific cognitive performance, as well as the
different Stroop Colour Word Test variables (matched and mismatched stimuli), are also
discussed.
4.7.1 Cognitive Performance: Clinical Groups (T1DM, T2DM, HTN)

This study was the first to assess and compare global and domain-specific cognitive
performance between non-clinical and clinical groups (T1DM, T2DM, and HTN) using
the Mini-Mental State Examination (MMSE) and the Cognistat. The main findings from
this investigation are: (1) global and domain-specific cognitive performance does not
differ between the non-clinical and clinical cohorts; (2) higher blood glucose level (BGL)
in the non-clinical group is inversely associated with global cognitive performance; and
(3) blood pressure (SBP and DBP) is significantly correlated with global cognitive
performance and individual domains of cognition in clinical groups (T1DM, T2DM,
HTN).
Substantial evidence shows that patients with diabetes mellitus (T1DM or T2DM)
perform slightly worse in cognitive assessments compared to age-matched healthy
controls and demonstrate modest decrements in multiple cognitive domains (Kalmijn et
al., 1995; Brands et al., 2005, Kilander et al., 1998; Yaffe et al., 2014). In their seminal
meta-analysis, Brands et al. (2005) investigated the effect of T1DM on cognitive function
and the magnitude of deterioration in cognitive domains and found that patients with
T1DM exhibit subtle cognitive dysfunction in psychomotor speed and information
processing (Cohen’s d effect size: 0.3 – 0.7). Interestingly, the investigators reported that
memory and learning domains were preserved, suggesting T1DM does not affect brain
areas involved in long-term memory storage and retrieval, such as the hippocampus. This
finding is substantiated by neuroimaging studies, which reveal that patients with T1DM
display subtle abnormalities in cortical structure, such as reduced cortical grey matter
147
Chapter 4.
density and lower total grey volume, compared to age-matched healthy controls (Musen
et al., 2006; Wessels et al., 2006), but not brain regions responsible for memory. Although
statistical significance was not reached in the present study, memory performance for the
T1DM group in the Cognistat was largely preserved, supporting existing literature that
patients with T1DM demonstrate intact memory function (Brands et al., 2005). The
precise mechanisms accounting for this remain poorly elucidated, but exogenous insulin
has been implicated. Conversely, impairments in memory are commonly documented in
patients with T2DM (Monette et al., 2014; Biessels & Despa, 2018). This has been
ascribed to several abnormal molecular processes triggered by insulin resistance, which
current literature suggests induces cerebral insulin resistance and disrupts insulin
signalling (Sims-Robinson et al., 2010; Biessels & Despa, 2018).
No difference was observed in global or domain-specific cognitive performance between

the non-clinical group and the diabetes groups (T1DM and T2DM). Not all studies have
found differences in cognitive function between patients with diabetes and healthy
controls. Lawson et al. (1984) found no difference in cognitive function between patients
with T1DM (n = 48) with some degree of peripheral and autonomic neuropathy and age-
matched healthy controls (n = 40) despite assessing cognition using several reputable
neurocognitive batteries (Wechsler Adult Intelligence Scale (WAIS) (Wechsler, 1955),
Wechsler Memory Scale, Kimura Repeated Figures Test, and Repeated Words Test). The
investigators attributed this incongruity to methodological issues, such as the significant
heterogeneity in characteristics of the diabetes patients recruited (e.g. disease duration,
glycaemic control, frequency of hypoglycaemia). This, together with the numerous
neuro-psychometric batteries available for administration and inconsistency in cognitive
assessments administered, has frequently been raised by other researchers as a
longstanding limitation in this area of research that complicates comparability of data
between studies (Munshi et al., 2006; Roberts et al., 2008; Roriz-Filho et al., 2009;
Geijselaers et al., 2017). Geijselaers et al. (2017) recommend administering similar
neurocognitive batteries to assess the subtle changes in cognition linked to these
conditions in future investigations to address this limitation.
148
Chapter 4.
The absence of a difference in global and domain-specific cognitive performance between

the study groups in the present study conflicts with available evidence and may be
explained by several factors. The main confounding factors suggested in the literature for
diabetes include glycaemic control, glycaemic variability, age of disease onset, disease
duration, recurrent hypoglycaemia, and glucose-lowering agents, but the contribution of
each to cognitive dysfunction in diabetes is reportedly modest (Biessels et al., 2014;
Feinkohl et al., 2015; Biessels & Despa, 2018). First, the glycaemic control of patients
with diabetes recruited for the present study was generally fair (T1DM: 7.5%; T2DM:
8.0%) and fell within the standard, less-rigid range for glycaemic control (HbA1C: 7.5 –
8.0%). While the HbA1C level was slightly elevated in both groups, it was found not to
be associated with global cognitive performance, suggesting that deterioration in
cognitive function occurs at higher HbA1C values. Support for this view is provided by
Jacobson et al. (2007), who reported accelerated deterioration in psychomotor speed at
HbA1C values > 8.8%. Similarly, Cukierman-Yaffe and colleagues (2009) found that a
1% increase in HbA1C in patients with T2DM was associated with a 0.20-point lower
score in global cognitive performance (MMSE). Alarmingly, at high concentrations
(HbA1C > 10%) it has been linked to an increased risk of dementia based on risk score
models (Exalto et al., 2013) and also substantial morbidity and mortality (Ricks et al.,
2012).
Although most authors have reported negative associations between HbA1C level and
cognitive function, there are other investigators who have not. For example, in a large
sample of elderly Japanese patients with T2DM (n = 1173; mean age: 71.8 ± 4.6 years),
Akisaki et al. (2006) observed no association between glycosylated haemoglobin (mean
HbA1C: 7.9%) and global cognitive performance and information processing speed using
the MMSE and Stroop B test, respectively. Similarly, Christman et al. (2011) found no
association between glycosylated haemoglobin (mean HbA1C: 8.5%) and cognitive
function, assessed using three reputable neuropsychological assessments (Digit Symbol
Substitution Test (DSST), the Delayed Word Recall Test (DWRT), and the Wechsler
Adult Intelligence Scale-Revised (WAIS-R)) (Wechsler, (1955), in a large sample of
patients with T2DM (n = 516). These data collectively substantiate prior evidence that
cognitive function is impacted at higher HbA1C values (Jacobson et al., 2007;
Cukierman-Yaffe et al., 2009; Exalto et al., 2013). The contribution of other glycaemic
149
Chapter 4.
indices, such as fasting plasma glucose (FPG) and postprandial glucose (PPG), to
cognitive dysfunction should also be investigated in future studies.
Although poor glycaemic control (HbA1C ≥ 8%) is a well-established risk factor for
cognitive dysfunction and the subtle diabetes-associated cognitive decrements, the
strength of the relationship is reportedly weak (Geijselaers et al., 2017). Geijselaers et al.
(2017) reviewed numerous studies (cross-sectional and longitudinal) examining
associations between HbA1C and cognitive function and found that the association,
although significant, was relatively weak. Interestingly, Crane et al. (2013) observed that
higher than normal HbA1C, but not in the range diagnostic of diabetes, was implicated
with an increased risk of dementia in people without diabetes, suggesting elevated blood
glucose concentrations are a modifiable risk factor for cognitive impairment. This finding
is also supported by emerging literature, which argues that glycaemic variability, instead
of chronic hyperglycaemia alone, could be linked to dementia and possibly diabetes-
associated cognitive decrements (Rawlings et al., 2017). Given the glycaemic control of
patients with diabetes (T1DM and T2DM) in the present study was fair, this may have
accounted for the relatively intact global and domain-specific cognitive performance
observed. It could have also explained the positive association observed between HbA1C
and comprehension performance in the Cognistat for the T1DM group.
Early intensive glycaemic control is known to reduce the development and progression
of long-term adverse microvascular complications (The Diabetes Control and
Complications Trial Research Group, 1993; The UK Prospective Diabetes Study
(UKPDS) Group, 1998); however, it confers little benefit on macrovascular
complications (Holman et al., 2008; Hayward et al., 2015) and cognitive outcomes, with
any meaningful benefit of glycaemic control on macrovascular complications taking
approximately 10 years to manifest (UKPDS, 1998, DeFronzo et al., 2017). Conversely,
it can also precipitate unwanted hypoglycaemic episodes, potentially offsetting
meaningful vascular benefits mediated by aggressive glucose control (Herzog & Sherwin,
2012). The landmark Diabetes Control and Complications Trial (DCCT), which assessed
cognition as an endpoint and monitored cognitive functioning closely in two groups
(intensive glycaemic control vs conventional control), found intensive glycaemic control
did not improve cognitive outcomes despite HbA1C being 20 mmol/L lower in the
intensive group compared to the conventional control group after an 18.5 year follow-up
150
Chapter 4.
period. Therefore, it is clear that several risk factors beyond chronic hyperglycaemia
mediate the relationship and contribute to the development and progression of cognitive
dysfunction in patients with diabetes (T1DM and T2DM).
Another risk factor known to moderate the relationship between diabetes and cognition
is age of disease onset. Substantial literature indicates that age of disease onset of diabetes
is a strong predictor of long-term cognitive outcomes, especially in T1DM (Biessels,
Deary, & Ryan, 2008; Ryan et al., 2016). Neuroimaging and neuropsychological data
consistently reveal that early diabetes onset (< 7 years of age) is associated with subtle
changes in cognitive development and brain structure (cortical atrophy) that persists into
adulthood (Ferguson et al., 2005; Musen et al., 2006; Wessels et al., 2006). This early-
onset effect, known as the diathesis hypothesis, has been ascribed to the susceptibility of
the developing brain to diabetes-induced metabolic disturbances, notably glycaemic
events (hypo- and hyperglycaemia) (Ryan, 2006). These glycaemic fluctuations disrupt
blood-brain-barrier (BBB) permeability and integrity and the highly regulated internal
milieu of the CNS, disturbing several vital brain processes known to be labile during the
first four to five years of life, such as synaptogenesis and neuronal signalling (Ryan, 2006;
Sweeney et al., 2016). This results in the unregulated entry of plasma-derived substances
into the CNS, which is thought to interfere with brain development, leading to an
increased likelihood of cognitive dysfunction (Ryan, 2006; Sweeney et al., 2016).
Brain susceptibility to diabetes-induced metabolic derangements has been hypothesised

to occur at two critical life periods: (1) during childhood when the brain is developing,
and (2) during neurodegenerative processes linked to ageing (> 65 years of age) (Biessels,
Deary, & Ryan, 2008; Koekkoek et al., 2015). Conversely, the brain appears resistant to
metabolic disturbances/neuroglycopaenia outside these two key periods, with clinically
relevant diabetes-associated cognitive decrements mainly affecting specific sub-groups
of patients, such as those with advanced micro- and macrovascular complications
(Biessels, Deary, & Ryan, 2008). Most patients with T1DM and T2DM recruited for the
present study developed their condition after 7 years of age and did not report
complications (retinopathy, nephropathy, neuropathy). Taken collectively, these
diabetes-related characteristics may have accounted for the largely preserved and slightly
better global cognitive performance observed for the T1DM group in the Mini-Mental
State Examination (MMSE) and most domains of the Cognistat compared to the non-
151
Chapter 4.
clinical group and other clinical groups, supporting evidence that the brain is resistant to
metabolic insult outside the critical periods outlined above (youth to middle age). The
lack of clinically-relevant complications (micro- or macrovascular) in the T2DM group
may have also explained the absence of a difference in global and domain-specific
cognitive performance between the T2DM group and the other groups.
Whether recurrent hypoglycaemia contributes to cognitive dysfunction in patients with

DM remains contentious. The literature is mixed: some studies have found associations
between recurrent hypoglycaemia and cognitive decline (Exalto et al., 2013), whereas
others have not (Bruce et al., 2009). Some investigators have also suggested the
relationship is possibly bidirectional (Yaffe et al., 2013; Feinkohl et al., 2014). Frier
(2014) argues the relationship is age-dependent. Convincing evidence supporting this
observation is provided by Asvold et al. (2010), who reported that young children with
T1DM who experienced severe hypoglycaemic episodes and developed their condition
at < five years of age demonstrated poorer cognitive function than those with T1DM not
exposed to severe hypoglycaemia when retested as adults. These data suggest recurrent
hypoglycaemia elicits lasting effects on cognitive function and corroborate the diathesis
hypothesis discussed previously (Ryan, 2006). Hypoglycaemic episodes – mild or severe
– are the most common reversible and feared adverse effects of diabetes management;
however, accurate retrospective recall of prior hypoglycaemic episodes is poor (Frier,
2014). While severe hypoglycaemic episodes can be reliably recalled for up to one year,
mild episodes can only be accurately recalled for no longer than one week (Frier, 2014).
This complicates ascertainment of the precise contribution of hypoglycaemia to diabetes-
associated cognitive dysfunction and may account for the conflicting relationship
currently reported. Although frequency of hypoglycaemia was solicited in the present
investigation, it could not be reliably recalled by study participants; therefore, the
relationship between hypoglycaemia and cognition could not be established. Other
investigators have encountered this issue and suggest monitoring blood glucose using
continuous glucose monitoring (CGM) devices to address this limitation (Geijselaers et
al., 2017).
152
Chapter 4.
Although no differences in global and domain-specific cognitive performance were found

between the groups in the present study, these findings neither support nor refute the
clinical utility of the Mini-Mental State Examination (MMSE) (Folstein, Folstein, &
McHugh, 1975) or the Cognistat (Kiernan et al., 1987) for the detection of early cognitive
impairment or early cognitive dysfunction associated with both diabetes and
hypertension. Both are reliable, validated and established cognitive screening instruments
frequently deployed in clinical contexts for the rapid identification of incipient and/or
early cognitive impairment (Folstein, McHugh, & Folstein, 1975; Tombaugh & McHugh,
1992; Pangman et al., 2000; Lancu & Olmer, 2006). Notably, the Mini-Mental State
Examination demonstrates reasonable psychometric properties (test-retest reliability
values: 0.56-0.99, Cronbach’s alpha: 0.54-0.96) (Tombaugh & McHugh, 1992) and is
currently recommended in emerging clinical guidelines for screening elderly patients
with diabetes for subtle cognitive decrements (American Diabetes Association (ADA,
2020)). Implementation of its recommended cut-off score (scores ≤ 23) has also been
linked to a subsequent dementia diagnosis in approximately 79% of cases, and it has also
been the most frequently used cognitive assessment for screening cognitive dysfunction
in subjects with hypertension (Iadecola et al., 2016). Therefore, it is likely the small
sample size of the clinical populations in the present study, paired with the variables
described earlier, accounted for the lack of a difference in global and domain-specific
cognitive performance observed between the study groups.
No significant differences in global and domain-specific cognitive performance were

found between the non-clinical group and HTN group in the present study. This finding
conflicts with available literature and could be attributable to several factors outlined
earlier (see Chapter 1, Section 1.7.2), although it is plausible to suggest that the well-
controlled BP of the participants with hypertension, due to anti-hypertensive therapy,
primarily explains the discrepancy. Other studies have reported similar findings; Nilsson
et al. (1998) assessed cognitive function using the MMSE in treated hypertensive patients
(n = 123) and subjects without hypertension (n = 76) and found performance did not differ
between the groups, suggesting blood pressure-lowering therapy may confer favourable
cognitive benefits. Whether anti-hypertensive therapy delays or prevents cognitive
decline remains inconclusive and is beyond the scope of this thesis, but data from a
recently published meta-analysis suggest it lowers the risk of dementia and AD by 16 and
12%, respectively (Ding et al., 2020). The authors also reported similar effects between
153
Chapter 4.
different classes of anti-hypertensive therapy (Ding et al., 2020). Most patients with
hypertension recruited for the study also did not present with hypertension-related
complications (e.g. stroke, myocardial infarction). These are strong risk factors for
vascular cognitive impairment (VCI), which causes abrupt, stepwise declines in cognition
(van der Flier et al., 2018). Taken together, this may have accounted for the lack of a
difference in global and domain-specific cognitive performance between the non-clinical
and HTN group.
A significant difference in average response time was observed between the non-clinical
and clinical groups using a computerised version of the Stroop Colour Word Test, with
the clinical groups, on average, taking longer than the non-clinical group to process
matched and mismatched stimuli. This finding is in accordance with existing literature
(Ryan et al., 2003; Brands et al., 2006; Wessels et al., 2007), which consistently indicates
that information processing and executive cognitive functioning are common cognitive
modalities detrimentally impacted by both DM and HTN (Ryan et al., 2003; Brands et
al., 2006; Wessels et al., 2007; Graveling, Deary, & Frier, 2013). Ample evidence
indicates that abnormalities in white matter correlate with poorer performance in tasks of
information processing tasks, executive function, and memory (Gunning-Dixon et al.,
2000; Hedden & Gabrieli, 2004). Consistent with this observation, Wessels et al. (2007)
reported an association between white matter atrophy and decreased performance in
information processing, executive function, and attention performance in a small sample
of patients with T1DM (n = 10) with proliferative retinopathy (PN, Grade 5), measured
using 13 neuropsychological tasks (see Wessels et al., 2007 for all cognitive assessments
administered). Imaging studies also consistently show that DM (T1DM and T2DM) and
HTN are associated with modest changes in cerebral integrity, with hypertension
recognised as the most significant risk factor for small vessel disease (SVD) (Iadecola et
al., 2016). Therefore, the slower average response time observed in the clinical groups,
particularly the T2DM and hypertension groups, could potentially indicate some degree
of damage to underlying white matter, such as, at the molecular level, demyelination or
axonal loss induced by chronic ischaemia. This usually results from SVD, which causes
irreversible cerebrovascular damage (Prins & Scheltens et al., 2015; Van de Flier et al.,
2018).
154
Chapter 4.
The finding that patients with T2DM demonstrate longer average response times
compared to both patients with T1DM and HTN is novel and suggests other risk factors
common to T2DM may have contributed to the reduced speed of information processing.
First, the glycaemic control of the T2DM group was slightly elevated, potentially
contributing to the slowing of information processing. Another emerging risk factor
attracting considerable attention that may have contributed is cerebral insulin resistance
(Biessels & Despa, 2018). In the brain, insulin plays important roles in influencing crucial
cognitive functions such as learning and memory (Cholerton et al., 2013; Biessels &
Reagan, 2015); however, disrupted insulin signalling has been reported in the brains of
individuals with AD, but without T2DM (Arnold et al., 2018). Evidence from
experimental models has also shown that impaired insulin signalling promotes amyloid
beta aggregation and hyperphosphorylation of tau protein (Sims-Robinson et al., 2010).
Taken together, these findings raise the possibility that defective insulin signalling – a
core pathophysiological hallmark of T2DM – could contribute to diabetes-associated
cognitive dysfunction and possibly permit crosstalk and accelerate progression to AD,
accounting for the reduced information processing observed.
4.8 Associations: Non-clinical Group

In the non-clinical group, increasing BGL was found to be inversely associated with
global cognitive performance (MMSE); however, the strength of the association was
weak. This finding is in accordance with prior research. For example, disruptions in BGL,
such as hypo- and hyperglycaemia, are known to compromise brain glucose homeostasis,
leading to impairments in cognition (McNay & Cotero, 2010; Frier, 2014). The cognitive
sequelae resulting from hyperglycaemia typically include reduced information
processing, altered cognitive flexibility, and global cognitive dysfunction (Brands et al.,
2005). Fluctuations in BGL have also been linked to aggravating diabetes-associated
cognitive decrements and increasing the risk of dementia. Interestingly, Geijselaers et al.
(2017) hypothesise that dysglycaemia contributes to the development, but not
progression, of cognitive dysfunction in patients with diabetes, a hypothesis supported by
longitudinal investigations that found cognitive decline at baseline but not after
considerable follow-up (Jacobson et al., 2007).
155
Chapter 4.
Several putative mechanisms have been proposed to explain how elevated glucose
concentrations may impair cognitive function. In addition to the pathophysiological
mechanisms described earlier, elevated blood glucose induces hypoxia and interrupts
cerebral blood flow, depriving neurons of crucial glucose and oxygen, particularly in
frontal brain regions (Baglietto-Vargas et al., 2016; Morley, 2017). At the molecular
level, experimental evidence indicates hyperglycaemia also disturbs critical neurological
processes, such as axonal transport and myelination (Sims-Robinson et al., 2010; Morley,
2017). Thus, the weak inverse association reported between global cognition and
attention performance in the Cognistat potentially implies that neurons in frontal brain
regions may become vulnerable to the neurotoxic effects of elevated blood glucose
concentrations nearing hyperglycaemia, even for short periods of time, resulting in poorer
performance in tasks subserved by frontal lobe function.
No associations were found between blood pressure (SBP or DBP) and global and
domain-specific cognitive performance in the non-clinical group. Other studies assessing
cognitive function in subjects with and without hypertension have replicated this finding
(Harrington et al., 2000), suggesting the brain becomes vulnerable to damage at higher
BP. Although to date the relation between blood pressure and cognition remains largely
controversial, convincing evidence from cross-sectional and longitudinal studies
consistently suggests that high blood pressure (> 140mmHg/90mmHg) is associated with
poor cognitive performance (global) across all age groups (mid-life and late-life), with
the strongest association found for mid-life BP (Kilander et al., 1998; Yaffe et al., 2014).
Interestingly, an inverse relationship between BP and cognition has been documented in
the older populations (>85 years of age) (Richmond et al., 2011). As the blood pressure
(both SBP and DBP) of the non-clinical population was normotensive, this likely
explained the lack of associations identified between blood pressure and global and
domain-specific cognitive performance. The cross-sectional study design of the present
study also limits inferences about causality and cannot capture the minute-to-minute
variability of BP. Iaedecola et al., (2016) recommend longitudinal studies can establish
better temporal associations between blood pressure and cognition.
156
Chapter 4.
4.9 Associations: Clinical Group

The literature provides strong, consistent evidence that acute and chronic glycaemic
events (hypo- and hyperglycaemia) detrimentally affect performance in several cognitive
domains (McNay & Cotero, 2010; Brands et al., 2005; Frier, 2014). Hypoglycaemia
(BGL < 3mmol/L) causes neuroglycopenia, resulting in marked impairment in attention-
dependent activities, such as driving and executive function (Warren & Frier, 2005;
Inkster & Frier, 2012; Graveling, Deary, & Frier, 2013), whereas hyperglycaemia (BGL
> 11mmol/L) causes slowing in psychomotor efficiency and cognitive flexibility (Brands
et al., 2005; Cox et al., 2005). In the T1DM group, no associations were found between
BGL (pre-study and post-study) and global and domain-specific cognitive performance.
This could have been due to participant blood glucose concentrations, which, although
marginally elevated (8 mmol/L), resided within the normoglycaemic range, suggesting
that brain vulnerability to hyperglycaemia potentially occurs at higher BGL values (Cox
et al., 2005). It is noteworthy that in the non-clinical group, increasing BGL (>7 mmol/L)
was inversely associated with global cognitive performance, but not in the T1DM group.
This finding implies a degree of cerebral adaptation, suggesting the brain develops a
protective and adaptive mechanism to reduce further damage/insult (such as oxidative
stress and interruptions in cerebral blood flow) to vulnerable neuronal populations from
recurrent hyperglycaemia (Wyke, 1959; Hwang et al., 2019). Such an adaptive
mechanism purportedly attenuates the magnitude of impairment in cognition and could
explain the lack of an association identified (Zammitt et al., 2008; Graveling, Deary, &
Frier, 2013).
In the T2DM population assessed, BGL was found to be inversely associated with
attention performance in the Cognistat. This result is supported by the literature
suggesting that elevated glucose concentrations cause modest cognitive dysfunction in
frontal brain regions (Brands et al., 2005; Cox et al., 2005). However, it was also found
to be significantly correlated with memory performance in the Cognistat in the HTN
group, indicating that brain areas involved in memory, such as the hippocampus,
potentially require slightly elevated glucose for optimum function during cognitive
demand. In accordance with this observation, McNay et al. (2000) and McNay et al.
(2001) showed hippocampal function is impaired substantially by hypoglycaemia.
Although hippocampal glucocorticoid receptors are sensitive to the neurotoxic effects of
chronic hyperglycaemia, the current findings suggest that short periods of elevated
157
Chapter 4.
glucose appear beneficial for optimum functioning during cognitive demand, especially
memory performance.
Several associations were found between BP (SBP and DBP) and global and domain-
specific cognitive performance in the clinical groups. It is estimated that approximately
60-70% of patients with T2DM develop hypertension, primarily due to increased
reabsorption of glucose and sodium (DeFronzo et al., 2017), but the BP (SBP and DBP)
of participants with DM (T1DM and T2DM) and HTN in the present study resided within
the normotensive range and was well controlled (HTN group). Blood pressure in the
normal range supports optimum cognitive functioning, likely attributable to adequate,
uninterrupted cerebral perfusion and neurovascular coupling (Iadecola et al., 2016).
Taken together, this may have potentially accounted for the various significant
associations reported and lack of cognitive dysfunction demonstrated by these clinical
groups (T1DM, T2DM, and HTN). The present study also found that increasing DBP in
the HTN group was inversely associated with recall performance in the MMSE. The
literature also reports that higher BP (SBP and DBP) is associated with poor cognitive
function in global and domain-specific performance (Starr et al., 1993; Kilander et al.,
1998; Obisesan et al., 2008), although not all studies have reported such associations
(Farmer et al., 1987). Investigators link these mixed findings to the various
methodological limitations and confounding variables, which frequently complicate
determination of the precise relationship between BP and cognition (Obisesan et al.,
2009; Iadecola et al., 2016).
158
Chapter 4.
4.10 Conclusions: Cognitive Performance

Using established cognitive screening tools, the present study has yielded evidence to
suggest that cognitive performance (global and domain-specific), as assessed using the
Mini-Mental State Examination (MMSE) (Folstein, Folstein, & McHugh, 1975) and the
Cognistat (Kiernan et al., 1987), does not vary between non-clinical and clinical
populations (reject aims 1.1 and 1.2). This result contrasts with the available literature but
could be attributable to the methodological limitations outlined in Chapter 1 and disease-
specific variables, as discussed in this chapter, which have been shown consistently to
moderate the relationship between these chronic diseases and cognition (Munshi et al.,
2006; Geijselaers et al., 2017; Biessels & Despa, 2018). It may also be ascribed to the
small sample sizes of each group, particularly the clinical groups. However, significant
differences in average response time when processing congruent and incongruent stimuli
using the Stroop Colour Word Test were found. This finding is novel and both supports
and contributes to the established body of evidence indicating that both DM and HTN are
associated with slowing of information processing and impaired executive functioning. It
is also relevant clinically, as impairment in these cognitive domains can interfere with
crucial disease self-management tasks (e.g. blood glucose monitoring, medication dosing,
blood pressure monitoring etc.), leading to diminished self-management of DM and HTN.
This may subsequently result in more diabetes and hypertension-related hospitalisations
and increased frequency of complications, including diabetes and hypertension-
associated cognitive dysfunction (Srikanth et al., 2020).
It is clear that larger, adequately powered, and controlled investigations (cross-sectional

and longitudinal) using similar neurocognitive batteries are required to further understand
the pattern of the subtle cognitive dysfunction in these chronic diseases. Consistency in
cognitive assessments administered in future studies will improve the comparability of
data between studies (Geijselaers et al., 2017). It will also help researchers understand
the appropriateness of each cognitive battery for the screening of the subtle and slowly
progressing diabetes-associated cognitive decrements and hypertension-induced
cognitive dysfunction, a currently debated and still unanswered question. Given both DM
and HTN frequently affect cognitive domains of attention, information processing,
memory, and executive function, investigators recommend choosing cognitive screening
tools that assess these domains (Geijselaers et al., 2017) and have a high negative
predictive value (i.e. accurately exclude cognitive dysfunction) (Srikanth et al., 2020)).
159
Chapter 5.
5. Associations between Blood Pressure, Blood

Glucose Level and Electroencephalography (Non-
Clinical)
This chapter reports associations found between BP (SBP and DBP) and BGL and EEG
variables (delta, theta, alpha, beta, and gamma) during baseline and active recordings (see
section 2.8 in methodology) in the non-clinical cohort. The findings are presented
commencing with links to pre-study variables followed by post-study variables. As age
and BMI were found to be significantly correlated to dependent variables, a partial
Pearson’s correlation (controlling for age and BMI as covariates) was applied.
5.1 Associations between pre-study SBP and EEG activity (non-clinical)

Pre-study SBP was found to be significantly associated with alpha, beta, and gamma
activity during the baseline phase in the non-clinical group (Table 5.1). For alpha activity,
these associations with pre-study SBP were found at the following locations: T7 (p = 0.04;
r = 0.45) (Figure 5.1), C3 (p = 0.01; r = 0.60), CZ (p = 0.01; r = 0.63), TP7 (p = 0.03; r =
0.48), CP3 (p = 0.04; r = 0.47), CPZ (p = 0.02, r = 0.51), and CP4 (p = 0.02; r = 0.53). For
beta activity, they were found at F3 (p = 0.03; r = 0.51), FT8 (p = 0.02; r = 0.55) (Figure
5.2), and P8 (p = 0.02; r = 0.54). In contrast, links with pre-study SBP for gamma activity
were at TP8 (p = 0.03; r = 0.42), P7 (p = 0.04; r = 0.39), and P8 (p = 0.04; r = 0.39). No
significant associations were identified between pre-study SBP and theta or delta activity
during the baseline phase.
160
Chapter 5.
Table 5.1. Associations between pre-study SBP and EEG activity during the baseline
phase for the non-clinical group.
Dependent
Independent
variable Brain Area p r
variable
(Baseline)
T7 0.04* 0.45
C3 0.01* 0.60
Cz 0.01* 0.63
Alpha (α) TP7 0.03* 0.48
CP3 0.04* 0.47
CPZ 0.02* 0.51
Pre-Study SBP CP4 0.02* 0.53
F3 0.03* 0.51
Beta (β) FT8 0.02* 0.55
P8 0.02* 0.54
TP8 0.03* 0.42
Gamma (γ) P7 0.04* 0.39
P8 0.04* 0.39
Key:
SBP – systolic blood pressure F – Frontal T – Temporal
C – Central P – Parietal * – statistical significance
161
Chapter 5.
p = 0.04; r = 0.45
Figure 5.1. Positive correlation between pre-study systolic blood pressure and alpha
power at T7 during the baseline phase for the non-clinical group.
Key:
T – Temporal µV/s2 – microvolts per second squared
162
Chapter 5.
p = 0.02; r = 0.55
Figure 5.2. Positive correlation between pre-study systolic blood pressure and beta
power at FT8 during the baseline phase for the non-clinical group.
Key:
F – Frontal
163
Chapter 5.
Several significant associations were found between pre-study SBP and EEG variables
during the active phase (Table 5.2). These links with pre-study SBP were found in fast-
wave activities (alpha, beta, and gamma). For alpha activity, these associations were
found at: FC3 (p <0.001; r = 0.55) (Figure 5.3); T7 (p = 0.01; r = 0.41); C3 (p = 0.01; r =
0.39); CP3 (p < 0.05; r = 0.47); PZ (p = 0.01; r = 0.38); P8 (p = 0.01; r = 0.40), and O1 (p
= 0.02; r = 0.36). For beta activity, they were found at FP1 (p = 0.01; r = 0.42), FP2 (p =
0.01; 0.44), F7 (p = 0.04; r = 0.32); F3 (p =0.01; r = 0.42), FZ ( p = 0.04; r = 0.37), FC3 (p
<0.001; r = 0.52), FCZ ( p = 0.01; r = 0.40), TP7 ( p = 0.01; r = 0.42), CP3 ( p < 0.05; r =
0.50), CP4 (p <0.05; r = 0.46); PZ (p = 0.04; r = 0.32), P8 (p = 0.001; r = 0.52), O1 (p =
0.04; r = 0.34), and O2 (p = 0.04; r = 0.33). In contrast, pre-study SBP was associated
with gamma activity at the following locations: FP2 (p <0.001; r = 0.71); FC3 (p = <0.001;
r = 0.54) (Figure 5.4); T7 (p = <0.001; r = 0.55); C3 (p <0.001; r = 0.62); CZ (p = 0.001; r
= 0.44); TP7 (p = 0.01; r = 0.51); PZ (p = <0.001; r = 0.64). There were no significant
associations between pre-study SBP and slow-wave brain activities (theta and delta)
during the active phase.
164
Chapter 5.
Table 5.2. Associations between pre-study SBP and EEG activity during the active
Dependent
Independent
variable
(Active)
FC3 <0.001* 0.55
T7 0.01* 0.41
C3 0.01* 0.39
Alpha (α) CP3 <0.05* 0.47
PZ 0.01* 0.38
P8 0.01* 0.40
O1 0.02* 0.36
FP1 0.01* 0.42
FP2 0.01* 0.44
F7 0.04* 0.32
F3 0.01* 0.42
FZ 0.04* 0.37
FC3 <0.001* 0.52
FCz 0.01* 0.40
Pre-Study SBP Beta (β)
TP7 0.01* 0.42
CP3 <0.05* 0.50
CP4 <0.05* 0.46
PZ 0.04* 0.32
P8 0.001* 0.52
O1 0.04* 0.34
O2 0.04* 0.33
FP2 <0.001* 0.71
FC3 <0.001* 0.54
T7 <0.001* 0.55
Gamma (γ) C3 <0.001* 0.62
CZ 0.001* 0.44
TP7 0.01* 0.51
PZ <0.001* 0.64
Key:
C – Central P – Parietal p – p-value
* – statistical significance r – rho value
165
Chapter 5.
p = <0.001; r = 0.55
Figure 5.3. Positive correlation between pre-study systolic blood pressure and alpha
power at FC3 during the active phase for the non-clinical group.
Key:
SBP – systolic blood pressure F – Frontal
mm Hg – millimetres of mercury
166
Chapter 5.
p = <0.001; r = 0.54
Figure 5.4. Positive correlation between pre-study systolic blood pressure and gamma
power at FC3 during the active phase for the non-clinical group.
Key:
C – Central µV/s2 – microvolts per second squared
F – Frontal
167
Chapter 5.
5.2 Associations between post-study SBP and EEG activity (non-clinical)

No significant associations were found between post-study SBP and the EEG variables
during the baseline phase. However, multiple links were identified between post-study
SBP and EEG activity during the active phase, mainly in the fast-wave activities (alpha,
beta, and gamma) (Table 5.3). Post-study SBP was linked with alpha activity at FC3 (p =
0.01; r = 0.42) and T7 (p = 0.01; r = 0.51) during the active phase, while for beta activity
the associations were observed at several locations as follows: FP1 (p = 0.02; r = 0.37),
FP2 (p = 0.01; r = 0.38), F3 (p = 0.01; r = 0.41), FC3 (p = 0.01; r = 0.42), FCZ (p<0.05; r
= 0.46), TP7 (p = 0.01; r = 0.39), CP3 (p<0.001; r = 0.40), CP4 (p =0.01; r = 0.40), and P8
(p = 0.01; r = 0.40). In contrast, associations between post-study SBP and gamma activity
were at FP1 (p = 0.02; r = 0.37), FP2 (p = 0.01; r = 0.41) (Figure 5.5), FC3 (p = 0.02; r =
0.37), C3 (p = 0.01; r = 0.38), TP7 (p = 0.02; r = 0.37), TP8 (p = 0.01; r = 0.37), P7 (p
=0.01; r = 0.41), and PZ (p = 0.01; r = 0.42). Similar to pre-study SBP, no significant
associations were found between post-study SBP and slow-wave brain frequencies (theta
and delta) during the active phase.
168
Chapter 5.
Table 5.3. Associations between post-study SBP and EEG activity during the active
Dependent
Independent
variable
(Active)
FC3 0.01* 0.42
Alpha (α)
T7 0.01* 0.51
FP1 0.02* 0.37
FP2 0.01* 0.38
F3 0.01* 0.41
Beta (β) FC3 0.01* 0.42
FCZ <0.05* 0.46
TP7 0.01* 0.39
Post-Study SBP CP3 <0.001* 0.40
FP1 0.02* 0.37
FP2 0.01* 0.41
FC3 0.02* 0.37
C3 0.01* 0.38
Gamma (γ)
TP7 0.02* 0.37
TP8 0.01* 0.37
P7 0.01* 0.41
PZ 0.01* 0.42
Key:
C – Central P – Parietal p – p-value
* – statistical significance r – rho value
169
Chapter 5.
p = 0.01; r = 0.41
Figure 5.5. Positive correlation between post-study systolic blood pressure and
gamma power at FP2 during the active phase for the non-clinical group.
Key:
P – Parietal µV/s2 – microvolts per second squared
F – Frontal
170
Chapter 5.
5.3 Associations between pre-study DBP and EEG activity (non-clinical)

There were several significant associations between pre-study DBP and EEG variables
during the baseline phase for alpha and gamma activity (Table 5.4). Pre-study DBP was
associated with alpha activity at FZ (p = 0.02; r = 0.51), FC4 (p = 0.02; r = 0.54), T7 (p =
0.02; r = 0.51), C4 (p = 0.04; r = 0.45), TP7 (p = 0.01; r = 0.54), CP4 (p = 0.02; r = 0.52),
and PZ (p <0.05; r = 0.64) (Figure 5.6). In contrast, pre-study DBP was associated with
gamma activity at TP7 (p = 0.04; r = 0.40), TP8 (p = 0.04; r = 0.40), and P7 (p = 0.01; r =
0.44). There were no significant associations between pre-study DBP and other EEG
frequency bands (beta, theta, and delta) during the baseline phase.
Table 5.4. Associations between pre-study DBP and EEG activity during the baseline
Dependent
Independent
variable
(Baseline)
FZ 0.02* 0.51
FC4 0.02* 0.54
T7 0.02* 0.51
Alpha (α) C4 0.04* 0.45
TP7 0.01* 0.54
Pre-Study DBP
CP4 0.02* 0.52
PZ <0.05* 0.64
TP7 0.04* 0.40
Gamma (γ) TP8 0.04* 0.40
P7 0.01* 0.44
Key:
DBP – diastolic blood pressure F – Frontal T – Temporal
171
Chapter 5.
p = 0.05; r = 0.64
Figure 5.6. Positive correlation between pre-study diastolic blood pressure and
alpha power at PZ during the baseline phase for the non-clinical group.
Key:
172
Chapter 5.
5.4 Associations between post-study DBP and EEG activity (non-clinical)

Several significant associations were found between post-study DBP and EEG activity
during the baseline phase (Table 5.5). These links were found with gamma activity at
multiple locations: F4 (p = 0.04; r = 0.40), FC4 (p = 0.04; r = 0.39), TP7 (p = 0.01; r =
0.46), P7 (p = 0.02; r = 0.44), TP8 (p = 0.01; r = 0.46), and PZ (p = 0.01; r = 0.48). No
significant associations were found between post-study DBP and other EEG frequency
bands (alpha, beta, theta, and delta) during the baseline phase.
Table 5.5. Associations between post-study DBP and EEG activity during the baseline
Dependent
Independent
variable
(Baseline)
F4 0.04* 0.40
FC4 0.04* 0.39
TP8 0.01* 0.46
Post-Study DBP Gamma (γ)
P7 0.02* 0.44
TP8 0.01* 0.46
PZ 0.01* 0.48
Key:
173
Chapter 5.
Multiple significant associations were found between post-study DBP and EEG variables
during the active phase. These were mainly observed with fast-wave brain activities
(alpha, beta, and gamma), although an association was also found with theta activity
(Table 5.6). Partial Pearson’s correlation revealed post-study DBP links with alpha
activity at FC3 (p = 0.04; r = 0.32) and O2 (p = 0.04; r = 0.33), with beta activity at FC3
(p = 0.04; r = 0.32) and FCZ (p = 0.02; r = 0.35), and with gamma activity at FP2 (p =
0.03; r = 0.34) (Figure 5.7) and FC3 (p = 0.04; r = 0.31). Interestingly, for theta activity,
an inverse link with post-study DBP was found at P3 (p = 0.01; r = - 0.39). No significant
associations were found between post-study DBP and delta activity.
Table 5.6. Associations between post-study DBP and EEG activity during the active
phase for the non-clinical group
Dependent
Independent
variable
(Active)
FC3 0.04* 0.32
Alpha (α)
O2 0.04* 0.33
FC3 0.04* 0.32
Beta (β)
Post-Study DBP FCZ 0.02* 0.35
FP2 0.03* 0.34
Gamma (γ)
FC3 0.04* 0.31
Theta (θ) P3 0.01* - 0.39
Key:
C – Central P – Parietal O – Occipital
* – statistical significance p – p-value r – rho value
174
Chapter 5.
p = 0.03; r = 0.34
gamma power at FP2 during the active phase for the non-clinical group.
Key:
F – Frontal µV/s2 – microvolts per second squared
P – Parietal p – p-value
r – rho value
175
Chapter 5.
5.5 Associations between pre-study BGL and EEG activity (non-clinical)

With respect to pre-study BGL and EEG activity, significant associations were found
with beta activity at F3 (p = 0.03; r = 0.51) (Figure 5.8) and F4 (p = 0.02; r = 0.53) during
the baseline phase (Table 5.7). However, there were no significant associations between
pre-study BGL and other EEG variables (alpha, gamma, theta, and delta).
Table 5.7. Associations between pre-study BGL and EEG activity during the baseline
Dependent
Independent
variable
(Baseline)
F3 0.03* 0.51
Pre-Study BGL Beta (β)
F4 0.02* 0.53
Key:
BGL – blood glucose level F – Frontal * – statistical significance
176
Chapter 5.
p = 0.03; r = 0.51
Figure 5.8. Positive correlation between pre-study blood glucose level and beta power
at F3 during the baseline phase for the non-clinical group.
Key:
177
Chapter 5.
With respect to pre-study BGL and EEG activity during the active phase, significant
inverse associations were identified with slow-wave brain activities (theta and delta)
(Table 5.8). Pre-study BGL was inversely associated with theta activity at F3 (p = 0.04; r
= - 0.33) and FZ (p = 0.02; r = - 0.37) (Figure 5.9), while for delta it was at P7 (p = 0.03;
r = - 0.34). There were no significant associations between pre-study BGL and fast-wave
brain activities (alpha, beta, and gamma).
Table 5.8. Associations between pre-study BGL and EEG activity during the active
Dependent
Independent
variable
(Active)
F3 0.04* - 0.33
Theta (θ)
Pre-Study BGL FZ 0.02* - 0.37
Delta (δ) P7 0.03* - 0.34
Key:
BGL – blood glucose level F – Frontal P – Parietal
178
Chapter 5.
p = 0.02; r = - 0.37
Figure 5.9. Negative correlation between pre-study blood glucose level and theta
power at FZ during the active phase for the non-clinical group.
Key:
179
Chapter 5.
5.6 Associations between post-study BGL and EEG activity (non-clinical)

In contrast with pre-study BGL, no significant associations were found between post-
study BGL and EEG variables during the baseline phase. However, during the active
phase post-study BGL was significantly associated with theta activity at the following
three locations: F3 (p = 0.01; r = - 0.42), FZ (p = 0.01; r = - 0.43), and CZ (p = 0.04; r = -
0.34) (Figure 5.10) (Table 5.9).
Table 5.9. Associations between post-study BGL and EEG activity during the active
Dependent
Independent
variable
(Active)
F3 0.01* - 0.42
Post-BGL Theta (θ) FZ 0.01* - 0.43
CZ 0.04* - 0.34
Key:
BGL – blood glucose level F – Frontal C – Central
180
Chapter 5.
p = 0.04; r = - 0.34
Figure 5.10. Negative correlation between post-study blood glucose level and
theta power at CZ during the active phase for the non-clinical
group.
Key:
181
Chapter 5.
5.7 Discussion: Associations between BP and BGL and Electroencephalography

(EEG) (Non-Clinical)
This chapter discusses findings regarding the associations found between pre-study and
post-study physiological variables (SBP, DBP, and BGL) and electroencephalography
(EEG) data obtained during the baseline and active phases for the non-clinical group. The
electroencephalography signal reflects ongoing electrical activity of underlying glucose
and oxygen-dependent cortical pyramidal neurons, which are highly sensitive to small
changes in BP and BGL (Jordan, 2004); therefore, the EEG represents a promising
instrument for detecting changes in cognition linked to these variables.
5.7.1 Associations between BP and EEG

Data exploring associations between BP and EEG activity in normotensive individuals
are lacking. This study is the first to report associations between BP (both SBP and DBP)
and EEG activity across a broad age range in normotensive individuals using 32-channel
EEG recording. It is also one of only a few investigations to explore associations between
BGL and EEG activity during normoglycaemia. The main findings were: (1) BP was
significantly correlated with fast-wave brain activities (alpha, beta, and gamma) over all
cortical areas; and (2) BGL was significantly associated with fast-wave and slow-wave
brain activities primarily over the frontal and central brain areas.
It is well established that synchronous neuronal activity reflects cognitive function

(Uhlhaas & Singer, 2010; Hamm et al., 2015; Modi & Sahin, 2017; Solomon et al., 2017).
The present analysis revealed that increasing SBP and DBP, which resided within the
normotensive range (SBP: 112.5 ± 12.1 mm Hg; DBP: 73.7 ± 9.7 mm Hg), was associated
with fast-wave brain activities (alpha, beta, and gamma), although more associations were
identified for SBP. These associations for SBP and DBP were primarily observed over
frontal brain regions, but others were found scattered across the cortex. This finding is
novel and has not been reported previously. Although alpha activity is typically
associated with idle brain states (Berger, 1929; Adrian & Matthews, 1954), several lines
of evidence suggest they also underlie important cognitive processes, including
attentional and perceptual tasks, information processing, semantic memory, and
inhibitory control (Klimesch, 1997; Klimesch et al., 2007; Zoefel et al., 2011; Klimesch,
2012). Klimesch (1997) found increased alpha power correlated with memory and
attention and suggested that alpha waves indicate proper thalamo-cortical function. It is
182
Chapter 5.
noteworthy that most of these cognitive processes are also subserved primarily by the
frontal lobe (Cabeza & Nyberg, 1997; Wood & Grafman, 2003; Arnsten, 2009); thus, this
may explain the alpha correlations observed predominantly in frontal brain areas in the
present study.
Both SBP and DBP were also found to be significantly correlated with beta and gamma
activity. These associations were localised chiefly over the frontal and central brain areas.
This finding is novel but supports established literature that beta and gamma waves are
cortically-generated rhythms (Fries, 2009; Campisi & LaRocca; Modi & Sahin, 2017).
Beta and gamma rhythms have received considerable attention in the neurocognitive
literature and several investigators have shown that beta and gamma waves reflect high-
order cognitive processing, such as selective attention and consciousness (Fries, 2009;
Campisi & LaRocca; Modi & Sahin, 2017). Conversely, declines in beta and gamma
power have been reported in subjects with cognitive dysfunction and/or cognitive
impairment (AD, dementia) (Jelic et al., 1996; Huang et al., 2000; Jeong, 2004). Thus,
the significant associations identified between SBP and DBP and beta and gamma activity
imply that BP in the normotensive range augments coordinated rhythmic activity of
cortical neurons responsible for generating fast-wave activities, potentially resulting in
enhanced cortical activation.
Interestingly, fewer associations were identified between DBP and

electroencephalography activity than between SBP and EEG activity. This finding has
not been reported in prior studies and potentially raises the possibility that SBP exerts
stronger and more profound effects on cortical activity than DBP. In agreement with this,
evidence obtained from cross-sectional and longitudinal studies has consistently shown
that mid-life SBP is a strong predictor of long-term cognitive outcomes (Kilander et al.,
1998; Swan et al., 1998; Yaffe et al., 2014). However, in the seminal expert review of
Iadecola et al. (2016), the authors acknowledge that not all studies have robustly
compared evidence between SBP and DBP and cognition. The present analysis also found
a weak, inverse association between DBP and parietal theta activity, suggesting that DBP,
not SBP, may detrimentally affect vulnerable neuronal populations. Consistent with this
observation, Taylor et al. (2011) assessed the independent effects of SBP and DBP on
mortality in a large sample population (n = 13, 792, mean age: not reported) and found
that DBP was a stronger predictor of mortality in younger adults than SBP. The mean age
183
Chapter 5.
of the non-clinical group in the present study was relatively young (31.3 ± 15.8 years);
hence, this may have explained the outcome obtained. However, it is noteworthy that no
other studies to date have examined associations between DBP and brain electrical
activity in normotensive individuals. Therefore, this area of research warrants further
investigation.
The literature is not particularly helpful in elucidating the mechanisms underlying the
neurophysiological changes associated with BP. This is likely due to a lack of studies
exploring links between BP and EEG activity. Although mechanisms have been
proposed, multiple lines of evidence suggest that changes in cerebral blood flow (CBF)
influence oscillatory activity. Cerebral blood flow, which is regulated by highly-
specialised neurovascular units (NVUs), ensures underlying cortical pyramidal neurons
receive a continuous, uninterrupted supply of both glucose and oxygen for optimum
metabolic and electrical function (Hossmann, 1994; Foreman & Claassen, 2012; George
& Steinberg, 2015; Sweeney et al., 2018). Cerebral blood flow has also been closely
linked to EEG activity (Figure 5.11); increased fast-wave EEG activity has been reported
when CBF is in the normal range (35-50mL/100g/min), whereas declines in CBF have
been correlated with increased slow-wave activity (Hossmann, 1994; Jordan, 2004).
Other studies have also suggested that BP influences EEG activity via afferent fibres
projecting to brain regions involved in high-order cognitive operations, such as the
prefrontal cortex (PFC) and anterior cingulate (Goldstein & Silverman, 2006; Duschek et
al., 2007). These brain areas play important roles in orchestrating complex cognitive
functions, such as attention and executive function, which have been consistently linked
to fast-wave brain activities (Fries, 2009; Campisi & LaRocca, 2014; Modi & Sahin,
2017). Given the BP (SBP and DBP) of the non-clinical population assessed fell within
the normal range, it is conceivable that BP in the normotensive range is associated with
normal, uninterrupted CBF, which promotes reliable neuronal signalling and enhanced
cortical activation.
184
Chapter 5.
Figure 5.11. The relationship between cerebral blood flow and electroencephalography
activity. Cerebral blood flow in the normal range is associated with predominantly
normal EEG activity, whereas declines in CBF have been linked to increased slow-wave
activity. Adapted from Foreman & Claassen, (2012, p 2).
5.7.2 Associations between BGL and EEG activity

Evidence that blood glucose levels (BGL) could influence EEG activity was first reported
by Ross & Loeser (1951). Numerous studies have since demonstrated (primarily in
clinical populations – T1DM, T2DM) that glycaemic fluctuations (i.e. hypo and
hyperglycaemia) are associated with noticeable changes in cerebral electrical activity
(Ross & Loeser, 1951; Tallroth et al., 1990; Cox et al., 2005; Rachmiel et al., 2016). The
main changes reported in these studies are (1) sharp increases in slow-wave activity (delta
and theta) and (2) diminished fast-wave activity (alpha, beta, and gamma). These
pronounced changes have been primarily observed over anterior brain regions and
parieto-occipital areas, respectively, although some studies have found them diffusely
distributed across the cortex. While the neurophysiological changes associated with
glycaemic events have been explored, few studies have reported associations between
BGL and EEG activity during euglycaemia in non-clinical groups.
185
Chapter 5.
The present analysis showed that increasing BGL, within the euglycaemic range, was
significantly associated with beta activity in frontal brain regions. This finding is
consistent with evidence obtained from similar studies (Tallroth et al., 1990; Rachmiel et
al., 2016). The literature indicates that the EEG signal is dominated by fast-frequency,
low amplitude oscillations (alpha, beta, gamma) during euglycaemia, suggesting
continuous glucose supply to underlying cortical pyramidal neurons promotes enhanced
fast-wave neural activity (Elsborg et al.,1990) (Figure 5.12). Cortical pyramidal neurons
are known to be highly sensitive to perturbations in BGL and oxygen (Jordan, 2004) and
it is well established that neural tissue requires a continuous, uninterrupted supply of
glucose for optimal functioning (Sweeney et al., 2018). Although beta oscillations
typically underlie sensorimotor functions (Modi & Sahin, 2017), they have also been
implicated in various high-order cognitive operations, including attention (Engel et al.,
2000; Campisi & LaRocca, 2014), language processing (Weiss & Muller, 2012), working
memory, (Deiber et al., 2007), and hedonic processing (Marco-Pallares et al., 2015).
Consensus exists in the literature that beta waves are also cortically generated oscillations,
especially in frontal areas; hence, this may explain the correlations observed in the frontal
region in the present study (Modi & Sahin, 2017).
186
Chapter 5.
Figure 5.12. An example of an electroencephalogram tracing recorded during

euglycaemia and hypoglycaemia in the same individual. Adapted from
Elsborg et al., (1990, p 277).
The finding that euglycaemia was inversely associated with slow-wave brain activities
(theta and delta) in frontal and central brain regions in the present study also supports
prior literature (Tallroth et al., 1990; Brismar et al., 2002; Rachmiel et al., 2016) and
implies that neuronal populations are not deprived of glucose. A global loss of slow-wave
activity has been consistently reported during euglycaemia (Tallroth et al., 1990;
Bjorgaas et al., 1998; Brismar et al., 2002). Slowing of EEG activity has been attributed
to a shift in energy substrate (e.g. from glucose to amino acids and/or lactate) (Beall et
al., 2012) and a reduction in cerebral glucose metabolism (Lewis et al., 1974). Increased
slow-wave activity has also been repeatedly correlated with cognitive dysfunction and
cognitive impairment. In a 2.5-year longitudinal investigation, Coben et al. (1985)
reported increased theta and delta activity and diminished alpha and beta activity in
patients with Alzheimer’s Disease (AD). Similarly, Giaquinto & Nolfe (1986) observed
comparable changes in EEG activity in a cross-sectional investigation in subjects with
dementia (n = 47, mean age 71 ± 5 yrs, 32 male, 15 female). Similar electrophysiological
abnormalities have also been observed in patients with mild cognitive impairment (MCI)
(Jelic et al., 2000). Thus, data from the current study indicate that euglycaemia supports
187
Chapter 5.
optimal metabolic and electrical activities of sensitive underlying cortical pyramidal

neurons, resulting in enhanced fast-wave brain activity.
The inverse association reported between euglycaemia and slow-wave actvities in the
present study further supports the glycaemic thresholds proposed by prior studies. The
EEG has been shown to be characterised by slow-wave activity when BGL is low
(Pramming et al., 1988; Tallroth et al., 1990; Bjorgaas et al., 1998; Hyllienmark et al.,
2005) or abnormally elevated (Rachmiel et al., 2016). This pattern is widely thought to
reflect cortical dysfunction (Elsborg et al., 1990). Although hypoglycaemia is associated
with a slowing of brain activity, the blood glucose threshold at which marked changes in
EEG activity occur remains controversial. For example, some researchers have found
detectable increases in slow-wave activity at 2 mmol/L (Pramming et al., 1988), whereas
others have reported changes at lower concentrations (1.8 mmol/L) (Tallroth et al., 1990).
Others suggest these changes vary significantly between individuals and occur between
the concentration range of 1.6 – 3.4 mmol/L (Amiel et al., 1991; Juhl et al., 2010).
However, the findings of the present study clearly indicate that the
electroencephalography signal is sensitive to changes in BGL and can detect changes in
cognitive activity linked to both BP and BGL.
188
Chapter 5.
5.8 Conclusions
The present analysis determined consistent associations between BP and BGL and
oscillatory brain activity in a non-clinical population using non-invasive scalp
electroencephalography. Various associations were found, suggesting that both variables,
even in their normal range, influence ongoing cortical electrical activity. Several
associations were also identified in post-study SBP, DBP, and BGL, indicating that small
changes in these variables affect ongoing oscillatory brain activity. While changes in
BGL are understood to directly affect the metabolic activities of sensitive glucose-
dependent cortical pyramidal neurons, the mechanisms underlying BP-associated
changes in EEG activity remain less clear and are yet to be fully understood. Current
literature suggests that cerebral blood flow and reliable signalling to cognitively advanced
brain regions, such as the prefrontal cortex and anterior cingulate cortex, may explain the
observed EEG activity (Hossmann, 1994; Jordan, 2004). Neurochemical mechanisms
have also been implicated (Gomèz et al., 2004), although consensus in the literature is
lacking. Another largely unexplored question is whether SBP or DBP elicits stronger
effects on cortical activity. Therefore, this area warrants further investigation.
Although this study provides preliminary insight into the associations between BP, BGL
and EEG activity, it is clear that larger, adequately-powered investigations (cross-
sectional and longitudinal) are warranted for better elucidation of the EEG changes linked
to these variables, especially BP. Ample evidence suggests the brain becomes vulnerable
to metabolic insult/damage at specific glycaemic thresholds (Pramming et al., 1988;
Bjorgaas et al., 1998; Tallroth et al., 1990; Hyllienmark et al., 2005) but thresholds for
BP remain controversial. Given cognitive function broadly influences workplace
productivity, public safety, mental wellbeing, and everyday activities, this could have
broader implications, such as determining thresholds that support optimal cognitive
performance and those that cause deterioration in cognitive function.
189
Chapter 6.
6. Associations between Blood Pressure, Blood

Glucose Level and Electroencephalography (Clinical)
This chapter reports associations between BP (SBP and DBP), BGL and EEG variables
(delta, theta, alpha, beta, and gamma) during the baseline and active phases for the clinical
groups (T1DM, T2DM, and HTN). Associations found between disease-specific
variables (disease duration and HbA1C level) and EEG activity, are also reported. The
findings are presented commencing with links to pre-study variables followed by links to
post-study variables. A Spearman’s rank-order correlation was applied to determine
associations between pre-study and post-study physiological variables (SBP, DBP, and
BGL) and EEG activity (baseline and active phases) for all clinical groups.
6.1 Type 1 Diabetes Mellitus
6.1.1 Associations between pre-study SBP and EEG activity (T1DM)

Pre-study SBP was found to be significantly associated with beta and theta wave activity
during the baseline phase (Table 6.1). For beta activity, these links with pre-study SBP
were found at TP8 (p = 0.04; r = - 0.74), whereas for theta activity it was found at CPZ (p
= 0.04; r = 0.64). There were no significant associations between pre-study SBP and the
other EEG frequency bands (alpha, gamma, and delta).
190
Chapter 6.
Table 6.1. Associations between pre-study SBP and EEG activity during the baseline
phase for the T1DM group.
Dependent
Independent
variable
(Baseline)
Beta (β) TP8 0.04* - 0.74
Pre-Study SBP
Theta (θ) CPZ 0.04* 0.64
Key:
SBP – systolic blood pressure T – Temporal P – Parietal
C – Central p – p-value r – rho value
* – statistical significance
Conversely, several significant associations were found between pre-study SBP and EEG
variables during the active phase, mainly for slow-wave brain activities (theta and delta)
(Table 6.2). Pre-study SBP links were inversely associated with theta activity at the
following locations: F4 (p = 0.03; r = - 0.68), FT7 (p <0.05; r = - 0.83) (Figure 6.1), T7 (p
<0.05; r = - 0.82), P7 (p = 0.02; r = - 0.68), and O1 (p = 0.01; r = - 0.80). Inverse links
were also found with delta activity at F7 (p =0.02; r = - 0.72), F3 (p <0.001; r = - 0.90),
FT7 (p <0.05; r = - 0.80) (Figure 6.2), T7 (p <0.001; r = - 0.85), T8 (p = 0.02; r = - 0.73),
and O1 (p = 0.04; r = - 0.68). No significant associations were found between pre-study
SBP and fast-wave activities (alpha, beta, and gamma) during the active phase.
191
Chapter 6.
phase in the T1DM group.
Dependent
Independent
variable
(Active)
F4 0.03* - 0.68
FT7 <0.05* - 0.83
Theta (θ) T7 <0.05* - 0.82
P7 0.02* - 0.68
O1 0.01* - 0.80
Pre-Study SBP F7 0.02* - 0.72
F3 <0.001* - 0.90
FT7 <0.05* - 0.80
Delta (δ)
T7 <0.001* - 0.85
T8 0.02* - 0.73
O1 0.04* - 0.68
Key:
P – Parietal C – Central O – Occipital
192
Chapter 6.
p = 0.04; r = - 0.83
Figure 6.1. Negative correlation between pre-study systolic blood pressure and
theta power at FT7 during the active phase for the Type 1 diabetes
mellitus group.
Key:
T – Temporal
193
Chapter 6.
p = 0.04; r = - 0.80
Figure 6.2. Negative correlation between pre-study systolic blood pressure and delta
power at FT7 during the active phase for the Type 1 diabetes mellitus
group.
Key:
T – Temporal
194
Chapter 6.
6.1.2 Associations between pre-study DBP and EEG activity (T1DM)

There were several significant associations between pre-study DBP and EEG activity
during the baseline phase (Table 6.3). For gamma activity, a link with pre-study DBP was
found only at P3 (p = 0.01; r = 0.69). In contrast, inverse associations with theta activity
were found at FT7 (p = 0.01; r = - 0.70) (Figure 6.3), T7 (p = 0.03; r = - 0.62), CPZ (p =
0.03; r = - 0.65), and O2 (p = 0.01; r = - 0.73). No significant associations were found
between pre-study DBP and the other EEG frequency bands (alpha, beta, and delta)
during the baseline phase.
Table 6.3. Associations between pre-study DBP and EEG activity during the baseline
Dependent
Independent
variable
(Baseline)
Gamma (γ) P3 0.01* 0.69
FT7 0.01* - 0.70
Pre-Study DBP T7 0.03* - 0.62
Theta (θ)
CPZ 0.03* - 0.65
O2 0.01* - 0.73
Key:
P – Parietal C – Central O - Occipital
195
Chapter 6.
p = 0.01; r = - 0.70
Figure 6.3. Negative correlation between pre-study diastolic blood pressure and theta
power at FT7 during the baseline phase for the Type 1 diabetes mellitus
group.
Key:
F – Frontal T – Temporal
µV/s2 – microvolts per second squared
196
Chapter 6.
In the active phase, a few significant associations were found between pre-study DBP and
theta activity (Table 6.4). These pre-study DBP links were found at F4 (p = 0.01; r = -
0.81), OZ (p = 0.04; r = - 0.66), and O2 (p = 0.02; r = - 0.68). However, no significant
associations were found between pre-study DBP and the other EEG frequency bands
(alpha, beta, gamma, and delta) during the active phase.
Table 6.4. Associations between pre-study DBP and EEG activity during the active
Dependent
Independent
variable
(Active)
F4 0.01* - 0.81
Pre-Study DBP Theta (θ) OZ 0.04* - 0.66
O2 0.02* - 0.68
Key:
DBP – diastolic blood pressure F – Frontal O – Occipital
6.1.3 Associations between pre-study BGL and EEG activity (T1DM)

Significant associations were found between pre-study BGL and EEG activity during the
baseline phase (Table 6.5). These pre-study BGL links were found with theta activity at
FP1 (p =0.02; r = 0.68) (Figure 6.4) and OZ (p = 0.02; r = 0.64). However, there were no
significant associations between pre-study BGL and the other EEG variables (alpha, beta,
gamma, and delta) during the baseline phase.
197
Chapter 6.
Table 6.5. Associations between pre-study BGL and EEG activity during the baseline
Dependent
Independent
variable
(Baseline)
FP1 0.02* 0.68
Pre-Study BGL Theta (θ)
OZ 0.02* 0.64
Key:
O – Occipital p – p-value r – rho value
198
Chapter 6.
p = 0.02; r = 0.68
Figure 6.4. Positive correlation between pre-study blood glucose level and theta
power at FP1 during the baseline phase for the Type 1 diabetes mellitus
group.
Key:
r – rho value
199
Chapter 6.
Similarly, pre-study BGL was significantly associated with theta activity during the
active phase at the same two locations (Table 6.6) of FP1 (p =0.03; r = 0.72) and OZ (p =
0.02; r = 0.71) (Figure 6.5). Pre-study BGL was also significantly associated with delta
activity at TP8 (p = 0.02; r = 0.72). No significant associations were found between pre-
study BGL and fast-wave activities (alpha, beta, and gamma) during the active phase.
Dependent
Independent
variable
(Active)
FP1 0.03* 0.72
Theta (θ)
Pre-Study BGL OZ 0.02* 0.71
Delta (δ) TP8 0.02* 0.72
Key:
T – Temporal O – Occipital * – statistical significance
200
Chapter 6.
p = 0.02; r = 0.71
Figure 6.5. Positive correlation between pre-study blood glucose level and theta
power at OZ during the active phase for the Type 1 diabetes mellitus
group.
Key:
O – Occipital µV/s2 – microvolts per second squared
201
Chapter 6.
6.1.4 Associations between post-study SBP and EEG activity (T1DM)

A significant association was found between post-study SBP and EEG activity during the
baseline phase (Table 6.7). An inverse link was observed between post-study SBP and
theta activity at P3 (p = 0.02; r = - 0.66). No significant associations were found between
post-study SBP and other EEG variables (alpha, beta, gamma, and delta) during the
baseline phase.
Table 6.7. Associations between post-study SBP and EEG activity during the baseline
Dependent
Independent
variable
(Baseline)
Post-Study SBP Theta (θ) P3 0.02* - 0.66
Key:
SBP – systolic blood pressure P – Parietal * – statistical significance
Conversely, there were several significant negative associations between post-study SBP
and EEG activity during the active phase, mainly in slow-wave brain activities (theta and
delta) (Table 6.8). For theta activity, the links were found at FT7 (p = 0.03; r = - 0.67), P7
(p = 0.04; r = - 0.64), and P4 (p = 0.03; r = - 0.64), while for delta they were at FT7 (p =
0.01; r = - 0.75), and C3 (p = 0.02; r = - 0.68). There were no associations between post-
study SBP and fast-wave activities (alpha, beta, and gamma) during the active phase.
202
Chapter 6.
Dependent
Independent
variable
(Active)
FT7 0.03* - 0.67
Theta (θ) P7 0.04* - 0.64
Post-Study SBP P4 0.03* - 0.64
FT7 0.01* - 0.75
Delta (δ)
C3 0.02* - 0.68
Key:
SBP – systolic blood pressure F – Frontal P – Parietal
6.1.5 Associations between post-study DBP and EEG activity (T1DM)

Few significant associations were found between post-study DBP and EEG variables
during the baseline phase (Table 6.9). For gamma activity, a negative association with
post-study DBP was observed at FC3 (p = 0.01; r = - 0.69), while for theta it was at P3 (p
= 0.04; r = - 0.59). There were no significant associations between post-study DBP and
the other EEG variables (alpha, beta, or delta) during the baseline phase.
203
Chapter 6.
Table 6.9. Associations between post-study DBP and EEG activity during the baseline
Dependent
Independent
Variable Brain Area p r
variable
(Baseline)
Gamma (γ) FC3 0.01* 0.69
Post-Study DBP
Theta (θ) P3 0.04* - 0.59
Key:
SBP – diastolic blood pressure F – Frontal P – Parietal
Similarly, few significant associations were found between post-study DBP and EEG
activity during the active phase. These links with post-study DBP were found with
gamma, theta, and delta activities (Table 6.10). For gamma activity, the link with post-
study DBP was found at FC3 (p = 0.04; r = 0.60) (Figure 6.6). In contrast, association
with theta was observed at O2 (p = 0.02; r = - 0.67), and with delta at TP8 (p = 0.04; r = -
0.64). There were no significant associations between post-study DBP and the other EEG
variables (alpha and beta) during the active phase.
Table 6.10. Associations between post-study DBP and EEG activity during the
active phase for the T1DM group.
Independent Dependent Variable Brain
p r
variable (Active) Area
Gamma (γ) FC3 0.04* 0.60
Post-DBP Theta (θ) O2 0.02* - 0.67
Delta (δ) TP8 0.04* - 0.64
Key:
SBP – diastolic blood pressure F – Frontal C – Central
O – Occipital T – Temporal P – Parietal
204
Chapter 6.
p = 0.04; r = 0.60
gamma power at FC3 during the active phase for the Type 1 diabetes
mellitus group.
Key:
F – Frontal C – Central
µV/s2 – microvolts per second squared
205
Chapter 6.
6.1.6 Associations between post-study BGL and EEG activity (T1DM)

With respect to post-study BGL and EEG activity during the baseline phase, multiple
significant inverse associations were found in alpha, beta, and theta activity (Table 6.11).
For post-study BGL and alpha activity, this association was found at FP1 (p = 0.01; r = -
0.62), while for beta it was at both FP1 (p = 0.04; r = - 0.70) and TP8 (p = 0.02; r = - 0.81).
In contrast, Spearman’s rank-order correlation revealed post-study BGL links with theta
activity at multiple locations as follows: FP1 (p = 0.04; r = - 0.63), FP2 (p = 0.01; r = -
0.69), F7 (p = 0.03; r = - 0.65), F3 (p = 0.02; r = - 0.65), F4 (p = 0.03; r = - 0.62), FC4 (p =
0.04; r = - 0.59), FT8 (p = 0.03; r = - 0.63), C3 (p = 0.04; r = - 0.59), TP7 (p = <0.05; r = -
0.77) (Figure 6.7), CP3 (p = 0.02; r = - 0.64), TP8 (p = 0.04; r = - 0.59), and P8 (p = 0.03;
r = - 0.64). No significant associations were identified between post-study BGL and
gamma or delta activity during the baseline phase.
206
Chapter 6.
Table 6.11. Associations between post-study BGL and EEG activity during the
baseline phase for the T1DM group.
p r
variable (Baseline) Area
Alpha (α) FP1 0.04* - 0.62
FP1 0.04* - 0.70
Beta (β)
TP8 0.02* - 0.81
FP1 0.04* - 0.63
FP2 0.01* - 0.69
F7 0.03* - 0.64
F3 0.02* - 0.65
Post-Study
F4 0.03* - 0.62
BGL
FC4 0.04* - 0.59
Theta (θ)
FT8 0.03* - 0.63
C3 0.04* - 0.59
TP7 <0.05* - 0.77
CP3 0.02* - 0.64
TP8 0.04* - 0.59
P8 0.03* - 0.64
Key:
207
Chapter 6.
p = 0.02; r = - 0.77
Figure 6.7. Negative correlation between post-study blood glucose level and theta
power at TP7 during the baseline phase for the Type 1 diabetes mellitus
group.
Key:
r – rho value
208
Chapter 6.
Similarly, several significant inverse associations were found between post-study BGL
and EEG activity during the active phase (Table 6.12). These were primarily observed in
slow-wave brain activities (theta and delta), although one was found in beta. The link in
post-study BGL with beta activity was observed at TP8 (p = 0.02; r = - 0.81). For theta
activity, the links were found at F3 (p = 0.03; r = - 0.65), FZ (p = 0.02; r = - 0.67); TP7 (p
= 0.02; r = - 0.70); TP8 (p = 0.01; r = - 0.75), and P8 (p <0.05; r = - 0.78) (Figure 6.8). In
contrast, associations of post-study BGL with delta activity were found at multiple
locations as follows: FZ (p = 0.02; r = - 0.70), F4 (p = 0.02; r = - 0.70), FC3 (p = 0.04; r –
0.63), P8 (p = 0.02; r = - 0.69), O1 (p = 0.01; r = - 0.78), and O2 (p = 0.04; r = - 0.64).
active phase for the T1DM group.
p r
Beta (β) TP8 0.02* - 0.81
F3 0.03* - 0.65
FZ 0.02* - 0.67
Theta (θ) TP7 0.02* - 0.70
TP8 0.01* - 0.75
Post-Study P8 <0.05* - 0.78
BGL FZ 0.02* - 0.70
F4 0.02* - 0.70
FC3 0.04* - 0.63
Delta (δ)
P8 0.02* - 0.69
O1 0.01* - 0.78
O2 0.04* - 0.64
Key:
209
Chapter 6.
p = 0.02; r = - 0.78
Figure 6.8. Negative correlation between post-study blood glucose level and theta
power at P8 during the active phase for the Type 1 diabetes mellitus
group.
Key:
210
Chapter 6.
6.1.7 Associations between disease-specific variables and EEG activity (T1DM)

As shown in Table 6.13 (below), several significant associations were found between
disease-specific variables (HbA1C and disease duration) and EEG variables during the
baseline phase for the T1DM group. Glycosylated haemoglobin was significantly
associated with beta activity at C3 (p = 0.02; r = 0.79) and TP8 (p = 0.04; r = 0.78), and
with theta activity at P3 (p = 0.03; r = 0.66) and P4 (p = 0.02; r = 0.67). In contrast, disease
duration was inversely associated with gamma activity at FT7 (p = 0.03; r = - 0.62).
Table 6.13. Associations between disease-specific variables (HbA1c, disease

duration) and EEG activity during the baseline phase for the T1DM
group.
p r
C3 0.02* 0.79
Beta (β)
TP8 0.04* 0.78
HbA1C
P3 0.03* 0.66
Theta (θ)
P4 0.02* 0.67
Disease
Gamma (γ) FT7 0.03* - 0.62
Duration
Key:
HbA1c – glycosylated haemoglobin C – Central T – Temporal
P – Parietal * – statistical significance p – p-value r – rho value
211
Chapter 6.
Similarly, multiple significant associations were found between HbA1C and slow-wave
frequency brain waves (theta and delta) during the active phase (Table 6.14). For theta,
these associations with HbA1C were found at the following locations: FT7 (p = 0.03; r =
0.69), T7 (p = 0.01; r = 0.77), P7 (p = 0.02; r = 0.73), P4 (p = 0.04; r = 0.65), O1 (p = 0.03;
r = 0.72), and O2 (p = 0.03, r = 0.70), while links for delta were found at F3 (p = 0.01; r =
0.78), FT7 (p = 0.02, r = 0.72), T7 (p < 0.05; r = 0.86) (Figure 6.9), C3 (p = 0.03; r = 0.69),
TP7 (p = 0.03; r = 0.70), and CP3 (p = 0.04; r = 0.65). There were no significant
associations between disease duration and the EEG variables (delta, theta, alpha, beta, or
gamma) during the active phase.
Table 6.14. Associations between disease-specific variables (HbA1C, disease

duration) and EEG activity during the active phase for the T1DM group.
p r
FT7 0.03* 0.69
T7 0.01* 0.77
P7 0.02* 0.73
Theta (θ)
P4 0.04* 0.65
O1 0.03* 0.72
O2 0.03* 0.70
HbA1c
F3 0.01* 0.78
FT7 0.02* 0.72
T7 <0.05 0.86
Delta (δ)
C3 0.03* 0.69
TP7 0.03* 0.70
CP3 0.04* 0.65
Key:
HbA1c – glycosylated haemoglobin F – Frontal T – Temporal
P – Parietal O – Occipital C – Central p – p-value
r – rho value * – statistical significance
212
Chapter 6.
p = 0.02; r = 0.86
Figure 6.9. Positive correlation between glycosylated haemoglobin and delta power
at T7 during the active phase for the Type 1 diabetes mellitus group.
Key:
HbA1C – glycosylated haemoglobin T – Temporal
µV/s2 – microvolts per second squared p – p-value
r – rho value
213
Chapter 6.
6.2 Type 2 Diabetes Mellitus (T2DM)
6.2.1 Associations between pre-study SBP and EEG activity (T2DM)

Multiple significant inverse associations were found between pre-study SBP and slow-
wave activity during the baseline phase (Table 6.15). The link in pre-study SBP with theta
activity was observed at FP1 (p = 0.01; r = - 0.65). In contrast, associations with delta
activity were at FZ (p = 0.04; r = - 0.58), FT8 (p = 0.03; r = - 0.66), T8 (p <0.01; r = - 0.86),
CP4 (p =0.01; r = - 0.74), TP8 (p = 0.04; r = - 0.59), and O2 (p = 0.01; r = - 0.74) (Figure
6.10). No significant associations were found between pre-study SBP and fast-wave
frequencies (alpha, beta, and gamma) during the baseline phase.
Table 6.15. Associations between pre-study SBP and EEG activity during the
p r
Theta (θ) FP1 0.01* - 0.65
FZ 0.04* - 0.58
FT8 0.03* - 0.66
Pre-Study SBP T8 <0.01* - 0.86
Delta (δ)
CP4 0.01* - 0.74
TP8 0.04* - 0.59
O2 0.01* - 0.74
Key:
C – Central P – Parietal O – Occipital
214
Chapter 6.
p = 0.01; r = - 0.74
Figure 6.10. Negative correlation between pre-study systolic blood pressure

and delta power at O2 during the baseline phase for the Type 2
Key:
O – Occipital µV/s2 – microvolts per second squared
215
Chapter 6.
Several significant associations were found between pre-study SBP and gamma, theta,
and delta activity during the active phase (Table 6.16). For gamma, the links with pre-
study SBP were observed at three locations: C3 (p = 0.02; r = 0.60), CP4 (p = 0.03; r =
0.59), and P3 (p = 0.02; r = 0.62) (Figure 6.11), while for theta the links were found at
FP2 (p = 0.04; r = - 0.58) and C3 (p = 0.02; r = - 0.62). For delta, a single link with pre-
study SBP was found at TP8 (p =0.02; r = - 0.60). Similarly, no significant associations
were found between pre-study SBP and the fast-wave frequency bands (alpha and beta)
during the active phase.
p r
C3 0.02* 0.60
Gamma (γ) CP4 0.03* 0.59
P3 0.02* 0.62
Pre-Study SBP
FP2 0.04* - 0.58
Theta (θ)
C3 0.02* - 0.62
Delta (δ) TP8 0.02* - 0.60
Key:
SBP – systolic blood pressure C – Central P – Parietal
F – Frontal T – Temporal * – statistical significance
216
Chapter 6.
p = 0.02; r = 0.62
Figure 6.11. Positive correlation between pre-study SBP and gamma power at
P3 during the active phase for the Type 2 diabetes mellitus group.
Key:
217
Chapter 6.
6.2.2 Associations between post-study SBP and EEG activity (T2DM)

Multiple significant inverse associations were found between post-study SBP and EEG
activity during the baseline phase (Table 6.17). Most associations were observed with
slow-wave frequency bands (theta and delta), although some were found with the fast-
wave activities (beta). For beta activity, these links with post-study SBP were found at F4
(p = 0.04; r = - 0.54), CZ (p = 0.02; r = - 0.62), and C4 (p = 0.04; r = - 0.55) (Figure 6.12),
and for theta at CZ (p = 0.02; r = - 0.63). In contrast, associations with delta activity were
found at the following locations: FP1 (p = 0.04; r = - 0.65), F3 (p =0.03; r = - 0.59), CZ (p
= 0.01; r = - 0.71), T8 (p = 0.01; r = - 0.66), CP4 (p = 0.04; - 0.59), and OZ (p = 0.04; r =
- 0.59).
Table 6.17. Associations between post-study SBP and EEG activity during the
p r
F4 0.04* - 0.54
Beta (β) CZ 0.02* - 0.62
C4 0.04* - 0.55
Theta (θ) CZ 0.02* - 0.63
FP1 0.04* - 0.65
Post-Study SBP
F3 0.03* - 0.59
CZ 0.01* - 0.71
Delta (δ)
T8 0.01* - 0.66
CP4 0.04* - 0.59
OZ 0.04* - 0.59
Key:
SBP – systolic blood pressure F – Frontal C – Central
P – Parietal O – Occipital T – Temporal
218
Chapter 6.
p = 0.04; r = - 0.55
Figure 6.12. Negative correlation between post-study systolic blood pressure

and beta power at C4 during the baseline phase for the Type 2
Key:
219
Chapter 6.
Similarly, several significant inverse associations were found between post-study SBP
and EEG theta and delta activities during the active phase (Table 6.18). Spearman’s rank-
order correlation revealed links with post-study SBP for theta activity only at CZ (p =
0.04; r = - 0.55), while for delta inverse associations were found at FCZ (p = 0.04; r = -
0.54), CZ (p = 0.02; r = - 0.61), TP8 (p = 0.01; r = - 0.65), and O1 (p = 0.03; r = - 0.59).
There were no significant associations between post-study SBP and fast-wave brain
activities (alpha, beta, or gamma) during the active phase.
p r
Theta (θ) CZ 0.04* - 0.55
FCZ 0.04* - 0.54
Post-Study SBP CZ 0.02* - 0.61
Delta (δ)
TP8 0.01* - 0.65
O1 0.03* - 0.59
Key:
SBP – systolic blood pressure C – Central F – Frontal
T – Temporal O – Occipital * – statistical significance
220
Chapter 6.
6.2.3 Associations between pre-study and post-study DBP and EEG activity
(T2DM)
Few significant inverse associations were found between pre-study DBP and theta and
delta activity during the baseline phase (Table 6.19). Pre-study DBP was associated with
theta activity at CPZ (p = 0.04; r = - 0.54), and with delta activity at CPZ (p = 0.04; - 0.58)
and CP4 (p = 0.04; r = - 0.60). Similar to pre-study and post-study SBP, there were no
significant associations between pre-study DBP and fast-wave brain activities (alpha,
beta, or gamma) during the baseline and active phases. Further, no associations were
found between post-study DBP and EEG activity during either the baseline or active
phase.
Table 6.19. Associations between pre-study DBP and EEG activity during the
p r
Theta (θ) CPZ 0.04* - 0.54
Pre-Study DBP CPZ 0.04* - 0.58
Delta (δ)
CP4 0.04* - 0.60
Key:
DBP – systolic blood pressure C – Central P – Parietal
6.2.4 Associations between pre-study and post-study BGL and EEG activity
(T2DM)
Few significant associations were found between pre-study BGL and slow-wave
frequency bands (theta and delta) during the baseline phase (Table 6.20). Pre-study BGL
was associated with theta activity at CPZ (p = 0.04; r = 0.54) and with delta activity at O1
(p = 0.01; r = 0.69).
221
Chapter 6.
Table 6.20. Associations between pre-study BGL and EEG activity during the
p r
Theta (θ) CPZ 0.04* 0.54
Pre-Study BGL
Delta (δ) O1 0.01* 0.69
Key:
BGL – blood glucose level C – Central P – Parietal
O – Occipital p – p-value
r – rho value * – statistical significance
Similarly, few significant associations were found between pre-study BGL and EEG
activity during the active phase (Table 6.21). Pre-study BGL was linked with delta
activity at TP8 (p <0.05; r = 0.73) (Figure 6.13), O1 (p = 0.02; r = 0.61), and O2 (p = 0.01;
r = 0.65). There were no significant associations between pre-study BGL and the other
EEG frequency bands (alpha, beta, gamma, or theta) during the active phase. No
significant associations were also found between post-study BGL and EEG variables
(delta, theta, alpha, beta, or gamma) during both the baseline and active phases.
p r
TP8 <0.01* 0.76
Pre-Study BGL Delta (δ) O1 0.03* 0.61
O2 <0.01* 0.76
Key:
BGL – blood glucose level T – Temporal P – Parietal
222
Chapter 6.
p = 0.02; r = 0.76
Figure 6.13. Positive correlation between pre-study blood glucose level and
delta power at TP8 during the active phase for the Type 2 diabetes
mellitus group.
Key:
P – Parietal
223
Chapter 6.
6.2.5 Associations between HbA1C and EEG activity (T2DM)

Several significant associations were identified between glycosylated haemoglobin
(HbA1C) and EEG activity during the baseline phase (Table 6.22). These links with
HbA1C were found for gamma, theta, and delta, although most were observed for the
slow-wave brain activities (theta and delta). Associations of HbA1C with gamma activity
were found at CP4 (p = 0.01; r = - 0.83) and P3 (p = 0.04; r = - 0.72), whereas links with
theta activity were at FP1 (p = 0.01; r = 0.80), FCZ (p = 0.01; r = 0.90), and PZ (p =0.02;
r = 0.85). Associations with delta activity were observed at multiple locations as follows:
FP1 (p = 0.04; r = 0.71), F7 (p = 0.01; r = 0.81), FT7 (p = 0.01; r = 0.87), FCZ (p < 0.05; r
= 0.91), FC4 (p = 0.04; r = 0.74), FT8 (p = 0.02; r = 0.79), P7 (p = 0.01; r = 0.84), P3 (p =
0.01; r = 0.87), PZ (p = 0.01; r = 0.83), P8 (p = 0.04; r = 0.74), and O2 (p = 0.01; r = 0.81).
224
Chapter 6.
Table 6.22. Associations between glycosylated haemoglobin (HbA1C) and EEG

activity during the baseline phase for the T2DM group.
p r
CP4 0.01* - 0.83
Gamma (γ)
P3 0.04* - 0.72
FP1 0.03* 0.80
Theta (θ) FCZ 0.01* 0.90
PZ 0.02* 0.85
FP1 0.04* 0.71
F7 0.01* 0.81
FT7 0.01* 0.87
HbA1C
FCZ <0.05* 0.91
FC4 0.04* 0.74
Delta (δ) FT8 0.02* 0.79
P7 0.01* 0.84
P3 0.01* 0.87
PZ 0.01* 0.83
P8 0.04* 0.74
O2 0.01 0.81
Key:
HbA1C – glycosylated haemoglobin F – Frontal C – Central
P – Parietal T – Temporal O – Occipital
Conversely, only few significant associations were found between HbA1C and EEG
activity during the active phase. (Table 6.23). Glycosylated haemoglobin was associated
with theta activity at CP4 (p = 0.02; r = - 0.78), O1 (p = 0.03; r = - 0.77), and OZ (p = 0.03;
r = - 0.77). It was also associated with beta activity at PZ (p = 0.01; r = 0.87).
225
Chapter 6.
Table 6.23. Associations between glycosylated haemoglobin (HbA1C) and EEG

activity during the active phase for the T2DM group.
p r
CP4 0.02* - 0.78
Gamma (γ) O1 0.03* - 0.77
HbA1c
OZ 0.03* - 0.77
Theta (θ) PZ 0.01* 0.87
Key:
HbA1C – glycosylated haemoglobin C – Central P – Parietal
T – Temporal p – p-value r – rho value
6.2.6 Associations between disease duration and EEG activity (T2DM)

No significant associations were found between disease duration and EEG activity during
the baseline phase. However, several links were found between disease duration and EEG
activity during the active phase, particularly with gamma and delta activities (Table 6.24).
For gamma activity, these links with disease duration were found at F7 (p = 0.04; r =
0.56), F4 (p = 0.02; r = 0.62), FT7 (p = 0.03; r = 0.60) (Figure 6.14), and TP7 (p = 0.03; r
= 0.56). In contrast, associations with delta activity were observed at CP3 (p = 0.04; r =
0.52) and O1 (p = 0.01; r = 0.67). No significant associations were found between disease
duration and the other EEG activities (alpha, beta, or theta) during the active phase.
226
Chapter 6.
Table 6.24. Associations between disease duration and EEG activity during the
active phase for the T2DM group
p r
F7 0.04* 0.56
F4 0.02* 0.62
Gamma (γ)
Disease FT7 0.03* 0.60
Duration TP7 0.03* 0.56
CP3 0.04* 0.52
Delta (δ)
O1 0.01* 0.67
Key:
F – Frontal P – Parietal T – Temporal C – Central
227
Chapter 6.
p = 0.03; r = 0.60
Figure 6.14. Positive correlation between disease duration and gamma power
at FT7 during the active phase for the Type 2 diabetes mellitus
group.
Key:
F – Frontal T – Temporal µV/s2 – microvolts per second squared
228
Chapter 6.
6.3 Hypertension
6.3.1 Associations between pre-study SBP and EEG activity (HTN)

Only one significant association was found between pre-study SBP and EEG activity
during the baseline phase (Table 6.25) for the HTN group. This link between pre-study
SBP and theta activity was found at PZ (p = 0.03; r = 0.60) (Figure 6.15). There were no
significant associations between pre-study SBP and the other EEG frequency bands
(delta, alpha, beta, or gamma) during the baseline phase.
Table 6.25. Associations between pre-study SBP and EEG activity during the
baseline phase for the HTN group.
p r
Pre-Study SBP Theta (θ) PZ 0.03* 0.60
Key:
SBP – systolic blood pressure P – Parietal * – statistical significance
229
Chapter 6.
p = 0.03; r = 0.60
Figure 6.15. Positive correlation between pre-study systolic blood pressure

and theta power at PZ during the baseline phase for the
hypertension group.
Key:
Significant associations were found between pre-study SBP and EEG activity during the
active phase (Table 6.26). These active phase associations between pre-study SBP and
beta activity were at F3 (p = 0.02; r = - 0.77) and PZ (p = 0.04; r = - 0.70). However, there
were no significant associated between pre-study SBP and the other EEG variables (delta,
theta, alpha, or gamma) during the active phase.
230
Chapter 6.
phase for the HTN group.
p r
F3 0.02* - 0.77
Pre-Study SBP Beta (β)
PZ 0.04* - 0.70
Key:
SBP – systolic blood pressure F – frontal P – Parietal
6.3.2 Associations between post-study SBP and EEG activity (HTN)

As shown in Table 6.27, several significant associations between post-study SBP and
EEG activity were found during the baseline phase. These links of post-study SBP were
primarily found with alpha and slow-wave brain activities (theta and delta). Post-study
SBP was associated with alpha activity at several locations as follows: FT7 (p = 0.04; r =
0.63), FC4 (p = 0.01; r = 0.71), FT8 (p 0.01; r = 0.83), CP3 (p = 0.03; r = 0.64), CP4 (p <
0.001; r = 0.74), TP8 (p < 0.001; r = 0.80), and P7 (p = 0.03; r = 0.63). In contrast,
associations with theta activity were at P3 (p = 0.03; r = 0.60), P4 (p = 0.03; r = 0.60), and
O2 (p = 0.01; r = 0.66), and with delta activity at FC3 (p = 0.04; r = 0.57), CZ (p = 0.03; r
= 0.59), and P4 (p = 0.04; r = 0.57). No significant associations were identified between
post-study SBP and fast-wave oscillations (beta or gamma) during the baseline phase.
231
Chapter 6.
Table 6.27. Associations between post-study SBP and EEG activity during the
p r
FT7 0.04* 0.63
FC4 0.01* 0.71
FT8 0.01* 0.83
Alpha (α) CP3 0.03* 0.64
CP4 <0.05* 0.74
TP8 <0.001* 0.80
Post-Study SBP P7 0.03* 0.63
P3 0.03* 0.60
Theta (θ) P4 0.03* 0.60
O2 0.01* 0.66
FC3 0.04* 0.57
Delta (δ) CZ 0.03* 0.59
P4 0.04* 0.57
Key:
SBP – systolic blood pressure F – frontal C – Central
P – Parietal p – p-value r – rho value
O – Occipital * – statistical significance
Conversely, few significant associations were identified between post-study SBP and
EEG activity during the active phase (Table 6.28). Post-study SBP was linked with theta
activity at TP7 (p = 0.04; r = 0.90) and CPZ (p = 0.04; r = 0.71), and with delta activity at
CPZ (p = 0.04; r = 0.62). Similarly, there were no significant associations between post-
study SBP and fast-wave oscillations (alpha, beta, or gamma) during the active phase.
232
Chapter 6.
p r
TP7 0.04* 0.90
Theta (θ)
Post-Study SBP CPZ 0.04* 0.71
Delta (δ) CPZ 0.04* 0.62
Key:
SBP – systolic blood pressure T – Temporal P – Parietal
6.3.3 Associations between pre-study DBP and EEG activity (HTN)

With respect to DBP, few significant associations were found between pre-study DBP
and EEG activity during the baseline phase (Table 6.29). Pre-study DBP links with beta
activity were observed at CP3 (p = 0.01; r = 0.71), and with theta activity at TP7 (p = 0.02;
r = 0.67) (Figure 6.16) and OZ (p = 0.01; r = 0.76). There were no significant associations
between pre-study DBP and the other EEG frequency bands (delta, gamma, or alpha)
during the baseline phase. Additionally, no significant associations were observed
between pre-study DBP and EEG variables (delta, theta, alpha, beta, or gamma) during
the active phase.
233
Chapter 6.
Table 6.29. Associations between pre-study DBP and EEG activity during the
p r
Beta (β) CP3 0.01* 0.71
Pre-Study DBP TP7 0.02* 0.67
Theta (θ)
OZ 0.01* 0.76
Key:
DBP – diastolic blood pressure C – Central T – Temporal
P – Parietal p – p-value r – rho value
234
Chapter 6.
p = 0.02; r = 0.67
Figure 6.16. Positive correlation between pre-study diastolic blood pressure

and theta power at TP7 during the baseline phase for the
hypertension group.
Key:
T – Temporal P – Parietal
r – rho value
235
Chapter 6.
6.3.4 Associations between post-study DBP and EEG activity (HTN)

Two significant associations were identified between post-study DBP and EEG activity
during the baseline phase (Table 6.30). Post-study DBP was associated with theta activity
at TP7 (p = 0.04; r = 0.60) and OZ (p = 0.01; r = 0.79). However, there were no significant
associations between post-study DBP and the other EEG variables (delta, alpha, beta, or
gamma) during the baseline phase.
p r
TP7 0.04* 0.60
Post-DBP Theta (θ)
OZ 0.01* 0.79
Key:
DBP – diastolic blood pressure T – Temporal P – Parietal
O – occipital p – p-value r – rho value
Conversely, several significant inverse associations were found between post-study DBP
and EEG activity (beta and gamma) during the active phase (Table 6.31). Post-study DBP
was linked with beta activity at F3 (p = 0.03; r = - 0.72), F4 (p = 0.04; r = - 0.83), and FCZ
(p = 0.04; r = - 0.70), while associations with gamma activity were found at FT7 (p = 0.02;
r = - 0.65) (Figure 6.17), FC3 (p = 0.02; r = - 0.65), T7 (p = 0.02; r = - 0.70), C3 (p = 0.04;
r= - 0.62), CPZ (p = 0.03; r = - 0.63), and O1 (p – 0.03; r = - 0.64). No significant
associations were found between post-study DBP and the other EEG variables (delta,
theta, or alpha) during the active phase.
236
Chapter 6.
active phase for the HTN group.
p r
F3 0.03* - 0.72
Beta (β) F4 0.04* - 0.83
FCZ 0.04* - 0.70
FT7 0.02* - 0.65
Post-DBP FC3 0.02* - 0.65
T7 0.02* - 0.70
Gamma (γ)
C3 0.04* - 0.62
CPZ 0.03* - 0.63
O1 0.03* - 0.64
Key:
DBP – diastolic blood pressure F – Frontal C – Central
P – Parietal T – Temporal O – Occipital
237
Chapter 6.
p = 0.02; r = - 0.65
Figure 6.17. Negative correlation between post-study diastolic blood pressure

and gamma power at FT7 during the active phase for the
hypertension group.
Key:
F – Frontal T – Temporal
r – rho value
238
Chapter 6.
6.3.5 Associations between pre-study BGL and EEG activity (HTN)

No significant associations were found between pre-study BGL and EEG activity during
the baseline phase. However, links between pre-study BGL and beta activity were
identified at FT7 (p = 0.02; r = 0.77) and FC4 (p = 0.04; r = 0.74) during the active phase
(Table 6.32). No significant associations were found between pre-study BGL and the
other EEG variables (delta, theta, alpha, or gamma) during the active phase.
p r
FT7 0.02* 0.77
Pre-Study BGL Beta (β)
FC4 0.04* 0.74
Key:
BGL – blood glucose level F – Frontal T –Temporal
6.3.6 Associations between post-study BGL and EEG activity (HTN)

Similarly, no significant associations were found between post-study BGL and EEG
activity during the baseline phase. However, associations during the active phase were
observed between post-study BGL and beta activity at TP7 (p = 0.04; r = - 0.79) and P8
(p < 0.05; r = - 0.93) (Table 6.33). There were no associations between post-study BGL
and the other EEG frequency bands (alpha, gamma, theta, or delta) during the active
phase.
239
Chapter 6.
active phase for the HTN group.
p r
Post-Study TP7 0.04* - 0.79
Beta (β)
BGL P8 <0.05* - 0.93
Key:
BGL – blood glucose level T – Temporal P – Parietal
240
Chapter 6.
6.4 Discussion: Associations between BP and BGL and Electroencephalography

(EEG) (Clinical Samples)
This chapter explores the associations between pre-study and post-study physiological
variables (SBP, DBP, and BGL) and EEG data for the clinical groups (T1DM, T2DM,
and HTN). The findings are discussed sequentially, commencing with the T1DM group.
Associations found between disease-specific variables (disease duration, HbA1C) for
each group, if any, are also discussed. The main findings are: (1) high BP is primarily
correlated with slow-wave brain activities (theta and delta), (2) poor glycaemic control
(T1DM and T2DM) is associated with slow-wave brain activities (theta and delta), and
(3) disease duration in DM (both T1DM and T2DM) is associated with both fast-wave
and slow-wave frequencies.
6.4.1 Associations between BP and EEG (T1DM and T2DM)

Literature exploring associations between BP and EEG activity is scarce. This
longstanding neglect has been attributed to various factors described earlier (see Chapter
1, section 1.7.2). This study is the first to report associations between BP (SBP and DBP)
and EEG activity in individuals with T1DM and T2DM. The present analysis revealed
that increasing SBP, but within the normotensive range for the T1DM and T2DM groups,
was inversely correlated with slow-wave EEG activities (theta and delta) in the frontal,
temporal, and parietal brain regions. Although no studies have replicated this outcome in
subjects with DM (T1DM and/or T2DM), this finding suggests that BP in the
normotensive range supports optimal metabolic and electrical activities of underlying
cortical pyramidal neurons, resulting in enhanced connectivity and fast-wave activity
(Hossmann, 1994; Foreman & Claassen, 2012; George & Steinberg, 2015; Sweeney et
al., 2018). In accordance with this observation, neuropsychological studies consistently
show that patients with normal BP perform better overall in neurocognitive assessments
than those with hypertension (Harrington et al., 2000; Wu et al., 2016). Cortical
pyramidal neurons are known to be sensitive to small changes in BP (Jordan, 2004) and
are dependent on a continuous, uninterrupted supply of glucose and oxygen for optimal
function (Sweeney et al., 2018). While slow-wave activity has been linked to several
cognitive processes (e.g. working and semantic memory) (Sauseng et al., 2010) in healthy
subjects, it has also been consistently reported in subjects with cognitive impairment
(Jelic et al., 1996; Jelic et al., 2000) and is widely considered to reflect cortical network
dysfunction (Uhlhaas & Singer, 2010). Others have suggested it may indicate neuronal
241
Chapter 6.
hypersynchronisation (Tomkins et al., 2008). Thus, it is conceivable that BP in the

normotensive range supports the metabolic and electrical activities of sensitive cortical
neurons. This could be due to providing adequate CBF, which has been closely linked to
EEG activity, and ensures activated cortical regions receive sufficient nutrients during
metabolically-demanding cognitive tasks (Hossmann, 1994; Jordan, 2004; Foreman &
Claassen, 2012; George & Steinberg, 2015; Sweeney et al., 2018).
Similar to the non-clinical group, more correlations were identified between SBP and
cerebral electrical activity than between DBP and cerebral electrical activity in the T1DM
and T2DM groups. This finding has not been documented in any prior investigations and
again potentially implies that SBP exerts stronger effects on brain oscillatory activity than
does DBP. Clear evidence exists in the literature that SBP is associated with adverse long-
term cognitive outcomes, particularly during mid-life (Kilander et al., 1998; Swan et al.,
1998; Yaffe et al., 2014). Interestingly, others have reported that slight elevations in DBP
(10 mm Hg), but not SBP, are associated with an increased risk (7%) of cognitive
impairment (Tsivgoulis et al., 2009). Iadecola et al. (2016) advise that not all studies have
compared evenly the evidence between SBP and DBP. There is also a lack of studies
specifically investigating associations between DBP and brain electrical activity.
Therefore, this area of research requires further exploration.
6.4.2 Associations between BP and EEG (Hypertension)

The present study showed that increasing SBP was significantly associated with slow-
wave brain activities (theta and delta) over the central and parietal brain regions in the
HTN group. This finding is novel and implies neuronal populations in central and parietal
areas are preferentially damaged by high SBP. Aberrant oscillatory activity has been
linked to disruptions in connectivity and coordinated rhythmic activity of underlying
cortical neurons (Buzsaki & Draguhn, 2004; Babiloni et al., 2011a; Hamm et al., 2015;
Modi & Sahin, 2017). Marked changes in BP, such as hypotension, have been associated
with detectable and reversible changes in cortical electrical activity (Mani & Townsend,
1954; Weisz et al., 2002; Duschek et al., 2006), but no studies have investigated
associations between hypertension and EEG activity. Chronic arterial hypertension is
known to disrupt vulnerable arteries, such as the middle cerebral artery and
lenticulostriate arteries (Sörös et al., 2013), which supply crucial nutrients (glucose and
O2) to neurons in frontal, central, and parietal regions (Sörös et al., 2013). Accumulating
242
Chapter 6.
evidence also indicates that hypertension damages neurovascular units (NVUs), which
are critical vascular regulatory mechanisms responsible for regulating cerebral blood flow
(Jennings et al., 2005; Iadecola et al., 2016). Interruptions in cerebral blood flow lead to
hypoperfusion during periods of high metabolic demand, depriving sensitive cortical
neurons of vital nutrients (glucose and O2). Changes in CBF have also been closely linked
to changes in oscillatory activity (Hossmann, 1994; Jordan, 2004). Therefore, it is
plausible that the mechanisms described above may explain the increased slow-wave
brain activity observed over the central and parietal regions.
Hypertension has also been linked to disrupting the integrity of the highly protective
blood-brain barrier (BBB) (Rosenberg, 2012; Huisa et al., 2015; Sweeney et al., 2018).
The mechanisms underlying this have yet to be clearly elucidated but are thought to
involve pericyte degeneration (Armulik et al., 2011) and hypoxia (Rosenberg et al.,
2016). It is known that damage to the BBB results in increased permeability and leakiness,
allowing the unregulated entry of potentially toxic substances into the sensitive CNS
parenchyma (Iadecola et al., 2016; Sweeney et al., 2018). This has been hypothesised to
disturb reliable neuronal signalling and synaptic function, triggering neuroinflammation
(Sweeney et al., 2018). BBB disruption has also been associated with
electrophysiological abnormalities, including observed increases in slow-wave activity
(Tomkins et al., 2008). Similar changes in EEG activity have been reported in patients
with early cognitive impairment (MCI) and advanced cognitive dysfunction (AD and
dementia) (Jelic et al., 1996; Jelic et al., 2000). As converging evidence suggests
neuroinflammation is a clinical hallmark of both MCI and AD, it is conceivable that the
increased slow-wave activity associated with SBP may potentially be an indicator of the
early stages of neuroinflammation.
The increased slow-wave brain activity observed at BP ≥ 135/90 mm Hg in the HTN

group in the present study also implies cerebral tissue becomes susceptible to the
deleterious effects of high blood pressure early (i.e. before anti-hypertensive therapy is
recommended). Although glycaemic thresholds for when changes in EEG activity occur
have been established, consensus is lacking in the literature concerning thresholds for BP
at which EEG is altered. It has been suggested that optimal BP in the brain is ≤ 130/80
mm Hg (Sörös et al., 2013). This is corroborated by data from large-scale studies, which
suggest that cortical tissue is more vulnerable than the cardiac system (Yusuf et al., 2004;
243
Chapter 6.
O’Donnell et al., 2010). In accordance with this, one large meta-analysis examining
cross-sectional studies also concluded that the rate of vascular and overall mortality
increases when BP exceeds 115/75 mm Hg (Lewington et al., 2002). Based on the data
obtained, the present study supports both Sörös et al. (2013) and Lewington et al. (2002),
suggesting susceptibility to hypertension-associated cognitive dysfunction occurs earlier
(BP ≥ 135/90 mmHg) than anti-hypertensive therapy is currently recommended (BP ≥
140/90 mmHg) (Unger et al., 2020). However, it is clear that further research
investigating the electroencephalography changes in patients with HTN is required to
more clearly define BP thresholds that detrimentally affect cognition. Such research will
validate the electroencephalography changes observed in the present study and could
have broader clinical implications, such as ascertaining the precise BP thresholds when
the brain becomes vulnerable to damage from HTN or elevated BP. It could also
determine the suitability of both BP and the EEG as non-invasive biomarkers of early
cognitive dysfunction.
Interestingly, higher SBP was found to be significantly associated with increased alpha
activity in frontal and central brain regions. This finding is novel. The literature generally
suggests increased alpha power correlates strongly with attentional processes, working
memory, and overall global cognitive status (Klimesch, 1996; Huang et al., 2000;
Knyazev, 2007), whereas declines in alpha activity reflect cortical dysfunction and/or
cognitive decline (Huang et al., 2000; Babiloni et al., 2010; Babiloni et al., 2011; van der
Hiele et al., 2011; Babiloni et al., 2016). However, not all studies have replicated these
findings: Alexander et al. (2006) reported increased alpha power in healthy subjects with
subjective memory complaints (n = 100, mean age: 64.9 years, age range: 52-88 years,
gender breakdown not reported) across the cortex from 26 electrode positions. On the
other end of the BP spectrum, increased alpha power in hypotension has been theorised
to reflect preparedness to react (Duschek et al., 2006). It has been suggested that increases
in the power of EEG frequency bands precede early stages of cognitive dysfunction and
reflect compensatory/adaptive mechanisms to preserve cognitive function (Pijnenburg et
al., 2004; Smith et al., 2007). Similar changes have been reported in neuroimaging studies
(fMRI), where patients with MCI demonstrate enhanced cortical activation compared to
those with AD (Sarter & Bruno, 2000). Thus, the present study data could suggest a
potential early compensatory mechanism by cerebral tissue, at high SBP, to shield
vulnerable neuronal populations from potential degeneration.
244
Chapter 6.
Strong inverse associations were observed between increasing DBP and fast-wave
activity (beta and gamma). These were primarily localised to frontal and central brain
regions. This finding is novel and suggests that high DBP also elicits detrimental effects
on cortical electrical activity. Although SBP is considered a stronger predictor of adverse
cognitive outcomes than DBP, prior studies have shown DBP is also associated with poor
cognitive function, especially in younger populations (Tsivgoulis et al., 2009). Potential
neurodegeneration of brain areas responsible for generating beta and gamma-band
oscillations may also explain the outcome obtained, as the synchronisation of neural
oscillations depends upon the integrity of underlying synaptic connections (Uhlhaas &
Singer, 2010). Hypertension is associated with reductions in both grey and white matter
and global cortical neurodegeneration (Jennings et al., 2012; Sörös et al., 2013; Iadecola
et al., 2016; Iadecola et al., 2019). Reductions in grey matter have been related to
decreases in the amplitude of cognitive event-related potentials (ERPs) (McCarley et al.,
2002). Studies exploring resting EEG activity in subjects with neurodegenerative diseases
(e.g. AD and dementia), which are characterised by widespread cortical atrophy, have
also consistently demonstrated diminished beta and gamma power (Jelic et al., 1996; Jelic
et al., 2000). Therefore, the reduced fast-wave brain activity may indicate disrupted
cortico-cortical connections and early damage/neurodegeneration to brain areas
responsible for generating these oscillations. At the neurotransmitter level, it may also
suggest disturbances in GABA (γ-aminobutyric acid) -ergic interneurons (Cobb et al.,
1995). These have been implicated in generating cortical fast-wave oscillations (Cobb et
al., 1995) by acting as a pacemaker. Whether SBP or DBP elicits stronger effects on
cortical electrical activity has not been explored in prior investigations and remains
unknown; therefore, it should be assessed in future studies (cross-sectional and
longitudinal).
245
Chapter 6.
The literature is not particularly helpful in elucidating the mechanisms underlying

hypertension-associated electrophysiological abnormalities. This could be attributed to
the longstanding neglect of research exploring the neurophysiological changes linked to
hypertension. Unlike hypoglycaemia, which the literature suggests can cause permanent,
irreversible changes in cerebral activity, studies suggest BP causes reversible changes in
EEG activity (Mani & Townsend, 1954; Weisz et al., 2002; Duschek et al., 2006). This
complicates ascertainment of the precise electroencephalography changes associated with
hypertension. Current available literature suggests interruptions in CBF primarily account
for the changes in EEG activity due to reduced perfusion to metabolically-active cortical
regions (Hossmann, 1994; Jordan, 2004). Others implicate neurochemical mechanisms
(Gomèz et al., 2004). It is clear that future studies investigating the neurophysiological
changes associated with hypertension are urgently warranted.
6.4.3 Associations between BGL and EEG (T1DM and T2DM)

Numerous studies have shown that changes in BGL (both hypoglycaemia and
hyperglycaemia) are characterised by noticeable changes in EEG activity (Greenblatt,
Murray, & Root, 1946; Izzo, Schuster, & Engel, 1953; Eeg-Olofsson, 1977; Brismar et
al., 2002; Hyllienmark et al., 2005). The main electrophysiological abnormalities
reported are: (1) marked increases in slow-wave activity (theta and delta) and (2) sharp
reductions in fast-wave brain activities (alpha, beta, and gamma) (Eeg-Olofsson, 1977;
Brismar et al., 2002; Hyllienmark et al., 2005). These changes have been primarily found
over anterior and parieto-occipital brain regions, respectively (Izzo, Schuster, & Engel,
1953; Eeg-Olofsson, 1977; Brismar et al., 2002; Hyllienmark et al., 2005; Cooray et al.,
2011a).
The present analysis revealed that increasing BGL was associated with slow-wave theta
power over anterior brain areas (FP1). This result is consistent with existing evidence
(Hauser et al., 1995; Faigle, Sutter, & Kaplan, 2013; Rachmiel et al., 2016). Available
literature indicates there is a shift to slow-wave frequencies (theta and delta) during
hyperglycaemia, with metabolically-demanding brain regions such as the prefrontal
cortex being most vulnerable (Tallroth et al., 1992; Frier, 2014); however, the blood
glucose threshold at which marked changes in EEG activity occur remains unclear.
Rachmiel et al. (2016) observed pronounced changes in EEG activity at a BGL of 11
mmol/L. Conversely, the present study found increases in slow-wave activity at BGL as
246
Chapter 6.
high as 8 mmol/L and a loss of slow-wave activity at BGL of 7 mmol/L. These findings
imply a critical glucose threshold and provide suggestive evidence, requiring further
confirmation, that vulnerable neuronal populations become susceptible to
hyperglycaemia-induced damage within this narrow range. It is well established that
cortical pyramidal neurons rely upon a continuous supply of glucose, with small changes
in BP or BGL rapidly disrupting CNS homeostasis and impairing function (Jordan, 2004;
Frier, 2014). The dissimilar outcome obtained between the present study and the
investigation of Rachmiel et al. (2016) could be attributed to cerebral adaptation and a
maladaptive counterregulatory response. Graveling et al. (2013) suggest the brain
develops an adaptive mechanism in patients with poor metabolic control to attenuate the
magnitude of damage to cortical tissue. Such adaptation could explain the
electrophysiological abnormalities reported by Rachmiel et al. (2016) at higher BGL. It
may also indicate potential damage to glucose-sensing neurons of the ventromedial
hypothalamus (VMH), arcuate nucleus (ARC), and paraventricular nucleus (PVN), which
play important roles in initiating the counterregulatory response (Song & Routh, 2006).
Damage to these neurons results in maladaptive responses to high glucose, resulting in
impaired awareness of hyperglycaemia.
The finding that glycosylated haemoglobin (HbA1C) was correlated with slow-wave
oscillations in both T1DM and T2DM groups is in agreement with available literature
(Tsalikian et al., 1981; Hauser et al., 1995; Hyllienmark et al., 2005) and contributes
additional evidence that poor glycaemic control is associated with electrophysiological
abnormalities. Poor glycaemic control is a well-established risk factor for cognitive
dysfunction (Ryan et al., 2003; Geijselaers et al., 2017; Biessels & Despa, 2018) and has
been previously correlated with increases in slow-wave activity (Tsalikian et al., 1981;
Hauser et al., 1995; Hyllienmark et al., 2005). Hauser et al. (1995) found poor metabolic
control was correlated with increased theta/delta power in young subjects with T1DM (n
= 44, mean age: not reported, mean HbA1C: 8.3%, diabetes duration: 5.9 years).
Similarly, Hyllienmark et al. (2005) found HbA1C correlated with increased delta activity
and concluded that poor metabolic control (as measured using HbA1C) is a risk factor for
abnormalities in EEG activity (n = 35, mean HbA1C: 7.2% (range: 4 – 10%), mean age:
17.1 ± 1.7 years (range: 14 – 19 years), disease duration: 7.6 ± 4.6 years, age of disease
onset: 9.6 ± 4.6 years (range: 1.6 – 17 years), gender breakdown: 16 females, 19 males).
247
Chapter 6.
Not all studies have reported associations between glycaemic control and EEG activity.
Soltész & Acsadi (1989) found no relationship between metabolic control and EEG
slowing despite patients having high glycosylated haemoglobin level (mean HbA1C:
11.3%). However, Soltész & Acsadi (1989) only visually inspected
electroencephalography changes and this may have accounted for the discrepancy. While
evidence from previous studies in our research unit and the broader literature indicates
slowing of EEG activity is typically observed in drowsy and fatigued behavioural states
(Lal & Craig, 2002; Campisi & LaRocca, 2014; Modi & Sahin, 2017), increases in slow-
wave activity are consistently found in patients with cognitive dysfunction and/or
cognitive impairment (Jelic et al., 1996; Jelic et al., 2000). This has led researchers to
believe that slow-wave activity reflects underlying cortical disruption between distal
regions and/or possible cognitive decline (Jelic et al., 1996; Jelic et al., 2000; Modi &
Sahin, 2017). Collectively, these data suggest that (1) poor glycaemic control is a risk
factor for electrophysiological abnormalities, and (2) the brain becomes vulnerable to
insult from HbA1C concentrations as high as 7.2%; however, emerging evidence suggests
patients with high HbA1C, but not in the range diagnostic of diabetes, have an increased
risk of dementia (Crane et al., 2012). This should be investigated in subsequent studies.
Importantly, the consistency of the data obtained in the present study raises the possibility
that the EEG is a suitable non-invasive measure for detecting the subtle cognitive
dysfunction triggered by diabetes.
The association between disease duration and electrophysiological abnormalities

currently remains controversial. Inconsistent findings have been obtained: some
investigators have shown it is associated with changes in evoked potentials related to
cognitive function (P300 component) (Tallroth et al., 1990), but most have found no
relationship at all (Mooradian et al., 1988; Hauser et al., 1995; Hyllienmark et al., 2005).
The present analysis showed disease duration (T1DM: 17.8 ± 9.2 years, T2DM: 11.3 ±
5.9 years) was significantly correlated with beta and delta power. This result is novel.
Increases in the power of EEG frequency bands have been suggested to precede early
stages of cognitive dysfunction and reflect initial compensatory/adaptive mechanisms to
preserve cognitive function (Pijnenburg et al., 2004; Smith et al., 2007). Comparable
changes have been observed in studies assessing cortical activation in individuals with
MCI and AD using functional magnetic resonance imaging (fMRI) (Sarter & Bruno,
2000). Hence, it is conceivable the increased beta power may represent an early
248
Chapter 6.
compensatory/adaptive mechanism by the brain to avert further damage to vulnerable

neurons. Alternatively, it could suggest neurodegeneration of brain regions that generate
beta-waves, such as cortico-cortical circuits (Uhlhaas & Singer, 2010; Modi & Sahin,
2017), due to repeated exposure to glycaemic events. On the other hand, the increased
delta power likely reflects underlying cortical disruption. However, it is noteworthy that
the duration of diabetes in the present study was greater than other studies; hence, this
may explain the absence of correlations found in earlier investigations.
Little is known about how transient hyperglycaemia influences brain oscillatory activity.
While it is associated with interruptions in cerebral blood flow (Morley, 2017), which
have been previously linked to EEG activity, hyperglycaemia has also been implicated in
disrupting the highly-regulated microenvironment of the CNS (Sweeney et al., 2018). At
the molecular level, data from animal studies indicate it results in impaired axonal
transport, demyelination, and ion channel dysfunction (Tomlinson & Gardiner, 2008).
There is also evidence that hyperglycaemia has direct inhibitory effects on orexigenic
hypothalamic neurons involved in modulating wakefulness and vigilance (Burdakov et
al., 2006; Sakurai, 2014). Inhibition of these wakefulness-promoting neurons could
explain the loss of fast-wave oscillations. Others ascribe the changes in EEG activity
during hyperglycaemia to hyperosmolarity, electrolyte disturbances, and ketoacidosis
(Misra & Kalita, 2018). Therefore, additional research is required to clarify the precise
mechanisms by which hyperglycaemia causes the observed abnormal EEG activity.
An important unanswered question is whether the electrophysiological abnormalities

commonly reported in patients with DM (T1DM and T2DM), and those found in HTN
(discussed earlier), relate to the effects of medication use for the conditions or whether
the effects can be reversed by therapy. Convincing evidence from a large, recently-
published meta-analysis (n = 31,090, age: > 55 years, follow-up: 7 – 22 years) examining
whether blood pressure-lowering medication reduces dementia risk showed anti-
hypertensive therapy was associated with a 16% and 12% reduction in AD and dementia,
respectively (Ding et al., 2020). Similar effects were reported between different classes
of anti-hypertensive medication, suggesting individual drug classes do not demonstrate
clinical differences (Ding et al., 2020). The risk of vascular cognitive impairment (VCI)
was not assessed. Although medication use/type was not found to be significantly
correlated with any variables, the possible confounding effects of medication cannot be
249
Chapter 6.
excluded. Interestingly, no studies have explored the impact of glucose-lowering

medications on oscillatory brain activity in patients with T2DM. Most patients with
T2DM in the present study were taking commonly-prescribed anti-diabetic medications
(e.g. metformin, dipeptidyl peptidase 4-inhibitors (DPP4-is)). Observational studies
suggest that these medications may exert beneficial effects on cognition by influencing
critical brain processes (e.g. metabolism, inflammation, and regeneration) (Patrone et al.,
2014; Orkaby et al., 2017), but no conclusive evidence exists that these therapies modify
the risk of cognitive dysfunction in T2DM (De Galan et al., 2009; Areosa Sastre et al.,
2017). Srikanth et al. (2020) suggest anti-hyperglycaemic agents that improve insulin
sensitivity and glucose uptake in the brain (e.g. metformin, intranasal insulin, and
glucagon-like peptide 1 receptor agonists (GLP-1RAs)) may be promising avenues for
future research. Whether glucose-lowering medications can reverse aberrant oscillatory
activity also remains unknown. It is clear this area of research requires further
exploration.
250
Chapter 6.
6.5 Conclusions
The results of this study provide evidence to suggest that raised blood glucose
concentrations and blood pressure (both SBP and DBP) are associated with changes in
oscillatory brain activity, which can be consistently detected using non-invasive scalp
electroencephalography. Several associations between the disease-specific variables
(HbA1C and disease duration) were also found, not only supporting limited existing
evidence, but also adding novel findings to the current literature. Collectively, the data
from the present study indicate: (1) BP and BGL are risk factors for and contribute to
electrophysiological abnormalities in both DM (T1DM and T2DM) and HTN, and (2) the
EEG may be a promising non-invasive biomarker for reliably and accurately detecting
early changes in cognition linked to DM and HTN. Prior studies have demonstrated the
clinical potential of the electroencephalography signal in identifying early disturbances
in cortical activity (Tallroth et al., 1992; Hauser et al., 1995; Jelic et al., 2000;
Hyllienmark et al., 2005). While several mechanisms have been proposed to account for
the abnormal EEG activity, the precise mechanisms remain poorly elucidated, especially
with respect to hypertension. Given the established link between both DM and HTN and
cognitive dysfunction, this area of research urgently requires further investigation.
Although novel data were obtained, as reported in this thesis, larger, adequately-powered
studies (cross-sectional and longitudinal) are required to validate current data and provide
a better understanding of the precise electrophysiological changes associated with both
DM (T1DM and T2DM) and HTN. Such studies will validate the current preliminary
abnormalities in EEG activity reported herein. The diagnostic specificity of altered neural
oscillations should also be interpreted cautiously. Uhlhaas & Singer (2010) caution that
aberrant oscillatory activity could reflect pathophysiological processes. Others implicate
inflammation and oxidative stress as the cause of abnormal oscillatory brain activity
(Mehvari et al., 2016). Subsequent longitudinal studies investigating EEG activity in
subjects with DM and HTN would clarify whether the altered neural activity indicates
degeneration of underlying brain structures or early stages of these pathophysiological
processes. Another limitation in the existing literature is the limited EEG montage
systems utilised. Future studies should investigate brain oscillatory activity using the
standardised international 10-20 system (Jasper, 1958) and a comprehensive montage
system (e.g. 32-channel EEG), similar to the present investigation. This will enable
meaningful comparisons between investigators and provide uniform coverage of the
251
Chapter 6.
scalp, better highlighting specific brain regions most susceptible to early deterioration. It
is also critical that disease-specific variables (e.g. age of disease onset, disease duration,
frequency of hypoglycaemic episodes) continue to be reported in as much detail as
possible. Clear evidence exists in the literature that these variables influence cognitive
outcomes (Munshi et al., 2006, Roberts et al., 2008, Roriz-Filho et al., 2009), although
the contribution of each is reportedly small (Biessels & Despa, 2018). Biessels & Despa
(2018) and Biessels and Whitmer (2019) also suggest future randomised controlled trials
(RCTs) should explore cognitive outcomes at least as a secondary endpoint. This may
identify medications that, in addition to providing meaningful glycaemic benefits, elicit
beneficial effects on cognition, which could contribute to reducing the substantial
socioeconomic and emotional costs linked to treating diabetes-related complications.
252
Chapter 7.
7. Limitations, Future Directions, and

Conclusions
7.1 Limitations and Future Directions

The present, cross-sectional investigation examined the relationship(s) between blood
pressure (SBP and DBP) and blood glucose level (BGL) and cognitive function in clinical
(T1DM, T2DM, and HTN) and non-clinical samples using scalp electroencephalography
(EEG) and psychometric batteries (MMSE and the Cognistat). Although reliable and
validated cognitive measures were used, and novel associations were identified,
limitations need to be discussed to inform future directions. This could help advance the
assessment of cognitive function in future studies.
Human physiological parameters are highly dynamic and influenced by many variables.
Although decrements in cognitive function triggered by both diabetes and hypertension
are observable in cross-sectional studies, the cross-sectional study design affords only a
snapshot of cognitive function. This limits inferences about causality. While measures
were taken to reduce variability in the data obtained (repetition of BP measurements,
adequate rest periods between BP measurements, five-minute EEG recordings,
administration of two neuro-psychometric batteries, enforcement of experimental
constraints, noise and temperature-controlled and sound-attenuated laboratory, exclusion
of participants drinking >16 standard drinks per day or taking psychotropic medication),
future studies would benefit from longitudinal study designs obtaining 24-hour BP and
BGL via ambulatory blood pressure monitors and continuous glucose monitoring (CGM)
systems. This would reduce fluctuations in physiological variables and allow researchers
to monitor long-term trajectories in cognitive function, enabling evaluation of any change
in cognition. It could also allow more robust assessment and understanding of the pattern
of decline of diabetes-associated cognitive decrements and hypertension-induced
cognitive dysfunction.
253
Chapter 7.
The present study recruited participants with clinically-diagnosed diabetes mellitus

(T1DM or T2DM; HbA1C ≥ 6.5%) and HTN (BP > 140/90 mm Hg), with or without
complications, controlled or uncontrolled with medication. These characteristics are
representative of the general population, but future studies would benefit from recruiting
individuals from high-risk populations (e.g. patients with pre-diabetes or who are pre-
hypertensive). Epidemiological studies suggest an estimated 3-11% of patients with pre-
diabetes (HbA1C ≥ 5.7 – < 6.5%) convert to overt T2DM annually (DeFronzo et al., 2015;
ADA, 2020). Pre-diabetes is associated with an increased risk of developing T2DM and
mortality (Abraham & Fox, 2013). It has also been linked to cognitive dysfunction: for
example, in a large, population-based cohort of elderly subjects, Kalmijn et al. (1995)
found elderly patients with impaired fasting glucose (IFG) performed worse in the MMSE
compared to age-matched subjects with normoglycaemia. Data also revealed that patients
with hyperinsulinaemia, a clinical pathophysiological hallmark of pre-diabetes,
demonstrated sub-optimal cognitive performance (Kalmijn et al., 1995). Therefore,
screening patients with pre-diabetes for early, subtle cognitive dysfunction using
objective cognitive measures and appropriate cognitive screening tools would be highly
advantageous. Although current clinical guidelines discourage routine screening of
patients with diabetes in the general population for potential cognitive dysfunction, due
to this being labour-intensive and there being no disease-modifying therapies available to
avert progression to dementia, early screening could enable early instatement of risk
factor reduction countermeasures recommended to delay progression to overt,
irreversible cognitive impairment (e.g. cardiovascular risk factor management and
individualised diabetes care) (Kalmijn et al., 1995; Biessels & Despa, 2018; Biessels &
Whitmer, 2019).
It is estimated that approximately 45.8% of all diabetes cases in adults worldwide are
undiagnosed (Beagley et al., 2014). Given the link between pre-diabetes and cognitive
dysfunction, studies exploring cognition using objective neurological measures in young
individuals with pre-diabetes could also have significant implications for global health
care. Such studies could determine whether pre-diabetes is associated with any early
reversible changes in oscillatory brain activity or early deterioration in specific cognitive
domains long before irreversible deficits in cognition have manifested. This could alert
general practitioners of individuals at high risk of developing diabetes-associated
cognitive decrements to commence early, aggressive therapeutic intervention to maintain
254
Chapter 7.
euglycaemia. It could also allow the initiation of robust risk-reduction countermeasures

(outlined above), which may delay progression to overt T2DM and potentially diabetes-
associated cognitive decrements. This may subsequently contribute to reducing the
substantial socioeconomic burden linked to diabetes-related complications on individual
patients, carers, and the healthcare system.
Glycosylated haemoglobin (HbA1C) is widely considered the ‘gold-standard’ indicator of

long-term glycaemic control; however, some researchers argue it does not accurately
reflect minute-to-minute fluctuations in blood glucose concentrations (Kovatchev, 2017;
ADA, 2020). It is also influenced by erythrocyte-related disorders and is insensitive to
hypoglycaemic episodes (Kovatchev, 2017; ADA, 2020). Hypoglycaemic episodes –
mild or severe – have been associated with permanent, irreversible changes in cognitive
function (Frier, 2014). They have also been linked to predicting the development of
cognitive impairment in the future (Yaffe et al., 2013). The landmark Diabetes Control
and Complications Trial (DCCT) found approximately 8% of severe hypoglycaemic
episodes could be determined from HbA1C (DCCT, 1993). This issue is compounded by
unreliable retrospective recall of severe and mild hypoglycaemic episodes, with recall
lasting up to one year and one week, respectively (Frier, 2014). Emerging literature also
suggests that glycaemic fluctuations in mid-life could contribute to cognitive decrements
and dementia (Rawlings et al., 2017). Therefore, future research may benefit from
measuring real-time blood glucose concentrations during cognitive assessment
(physiological and cognitive assessment) using continuous glucose monitoring (CGM).
To date, few studies have evaluated cognitive outcomes in patients while recording BGL
continuously using CGM. These studies may provide stronger visualisation of real-time
changes in blood glucose concentration occurring during cognitive stimulation. They may
also identify how minute-to-minute fluctuations in BGL influence cognitive function and
could assist investigators to understand better the precise changes in EEG activity that
occur in response to variations in these variables.
255
Chapter 7.
Acquisition of neuroimaging data via established neuroimaging modalities (e.g.

functional magnetic resonance imaging (fMRI)) should be considered for future cognitive
studies. These technologies are expensive but would facilitate the identification of brain
regions activated during cognitive tasks and hence complement and validate any
abnormalities observed in EEG recordings and the cognitive tools. Such data could
support EEG as a potential non-invasive biomarker of early cognitive dysfunction and
the MMSE and the Cognistat in clinical contexts as the preferred cognitive screening
tools for screening the subtle cognitive decrements associated with DM and HTN. It could
also afford prompt determination of brain areas vulnerable to early insult from both DM
and HTN in susceptible individuals.
Scalp EEG recordings are the most cost-effective and common method for recording
cortical activity; however, the signal may be attenuated by signal-distorting tissue, such
as the skull and intermediary neural tissue (Ritter & Villringer 2006; Buzaki et al., 2012).
It also cannot adequately detect electrical activity generated by neuronal populations in
deep-brain structures (e.g. the hippocampus). Neuroimaging data consistently indicate
that the hippocampus is detrimentally affected by T2DM (Ritter & Villringer 2006;
Buzaki et al., 2012). One modified version of the EEG that addresses these limitations
and could be deployed in future studies is electrocorticography (ECoG). This technique
records brain activity directly from the cerebral cortex via stainless-steel electrodes
implanted subdurally, bypassing underlying signal-distorting tissue and improving spatial
resolution (Buzaki et al., 2012). Therefore, future investigations may consider
implementing this technique to more accurately record brain activity in patients with DM
and HTN. The improved spatial resolution may also assist researchers to pinpoint earlier
the brain regions susceptible to insult, enabling earlier therapeutic and lifestyle risk factor
management intervention.
At the time of writing, this study was the first to report associations between BP and brain
oscillatory activity; thus, the reproducibility of the EEG signal to consistently identify
electrophysiological abnormalities associated with diabetes and hypertension, chiefly the
latter, should be strongly considered in future longitudinal investigations. Brismar et al.
(2005) found the EEG demonstrated high test-retest reliability in detecting reductions in
fast-wave brain activity (beta and gamma) over repeat measurements at three and nine
months post-baseline (correlation coefficient: 0.91 (alpha) and 0.92 (beta) between first
and second visit in adolescent subjects with T1DM receiving multiple insulin therapy
256
Chapter 7.
(MIT). Further investigation of this reproducibility will assist investigators determine the
appropriateness of the EEG as a screening instrument for monitoring the subtle changes
in cognition linked to both diabetes and hypertension.
Numerous cognitive batteries are available to assess cognitive function, but no clear
consensus exists in the literature concerning the suitable cognitive screening tools to
detect the subtle cognitive decrements triggered by DM and HTN. These are commonly
undetected by formal neurocognitive testing until frank and irreversible damage has
occurred. The diverse range of cognitive assessments available also complicates
comparability of data between studies. The Mini-Mental State Examination (MMSE)
(Folstein et al., 1975) and the Cognistat (Kiernan et al., 1987) are established cognitive
tools routinely deployed in clinical practice for screening cognitive impairment; however,
they do not robustly assess all cognitive domains (such as executive function and long-
term memory) or all modalities affected adversely by both DM and HTN (Srikanth et al.,
2020). This can result in potential cognitive decrements being overlooked and an
incomplete representation of broader cognitive performance. One cognitive tool sensitive
to executive function and recommended by the US National Institute of Neurological
Disorders and Stroke, is the Montreal Cognitive Assessment (MoCA) (Nasreddine et al.,
2005). Previous investigators using this tool have reliably identified executive
dysfunction, a cognitive domain detrimentally impacted by both DM and HTN (Dong et
al., 2010; Pendlebury et al., 2010). Thus, the present study could have also benefited from
investigating diabetes or hypertension-associated cognitive dysfunction using this
assessment tool. However, it should be stressed that the modern, around-the-clock
healthcare system permits little time for comprehensive assessment of cognition in day-
to-day practice. It has also been suggested that screening all patients with diabetes for
potential early cognitive dysfunction would be labour-intensive (Biessels & Whitmer,
2019) and impractical (Srikanth et al., 2020), as prevention strategies for dementia in
middle-aged individuals with diabetes mirror those with known cardiovascular risk
factors (i.e. optimising glycaemic control, lipid concentrations, BP, and diet and
exercise). Therefore, neuro-psychometric batteries sensitive to the subtle cognitive
dysfunction associated with DM and HTN with a high negative predictive value should
be prioritised. Consistency in neuropsychological assessments administered in future
studies will also improve the comparability of data between investigators (Geijselaers et
al., 2017).
257
Chapter 7.
Although novel findings were reported in this thesis, it is noteworthy that this was a
preliminary/pilot study. The sample sizes of the clinical populations were also small. This
limits the generalisability of the findings and the number of adjustments performed during
data analysis. It also may have potentially given rise to chance findings (Type I and Type
II error). However, appropriate statistical analyses (non-parametric) were conducted to
accommodate the smaller sample sizes (clinical populations) and consistent findings were
obtained, likely reflecting true results. The recruited population may also not necessarily
reflect a true representation of the general population. Recruitment from the local Sydney
community could have inadvertently introduced bias, as this would have attracted
participants in close proximity to the study location. Future studies should aim to recruit
larger sample sizes, preferably age and BMI-matched, from as many geographical
locations as possible. This would strengthen the translatability of the data and would also
facilitate meaningful and direct between-group comparisons.
Several disease-specific (diabetes and hypertension linked) variables were obtained in the
present study (e.g. glycosylated haemoglobin (HbA1C), disease duration (chronicity), age
of disease onset, medication, etc.). Current literature indicates these factors moderate the
relationship between these conditions and cognitive function (Munshi et al., 2006;
Roberts et al., 2008; Roriz-Filho et al., 2009). One diabetes-related variable that may
have been imprecisely reported by study participants was frequency and severity of
hypoglycaemia. Hypoglycaemia is a common, reversible, adverse effect of glucose-
lowering therapy in patients with diabetes (T1DM and T2DM) associated with noticeable
changes in cognitive function (EEG activity and cognitive performance) (Frier, 2014).
However, retrospective recall of hypoglycaemic events is often poor: severe
hypoglycaemic episodes can be recalled accurately for up to one year, whereas mild
episodes can only be recalled with accuracy for no longer than one week (Frier, 2014).
This complicates precise ascertainment of the contribution of hypoglycaemia to cognitive
dysfunction and may explain the controversial relationship currently reported between
hypoglycaemia and adverse cognitive outcomes. Improvement in recall of
hypoglycaemic episodes can be achieved by monitoring blood glucose concentrations
continuously using CGM. Analysis of such data could reveal dynamics of blood glucose
fluctuations, such as the type of hypoglycaemic event experienced (mild or severe). It
could also enable potential prediction of upcoming adverse glycaemic events based on
258
Chapter 7.
patterns from previous glycaemic events (Kovatchev, 2017) and help researchers
understand better the relationship between hypoglycaemia and cognitive dysfunction.
Blood pressure was obtained using a reliable and validated automated non-invasive blood
pressure monitor (see Chapter 2, section 2.7) in accordance with recommendations
outlined by the International Society of Hypertension Global Hypertension Practice
Guidelines (i.e. quiet sitting position, five-minute rest period, repeat measurements,
determination of average BP) (Unger et al., 2020). While brachial blood pressure is the
preferred method of blood pressure measurement in clinical practice and research, some
researchers suggest it may not accurately reflect cerebral blood pressure, as the brain is
located upstream from the measurement point (Cohen & Townsend, 2017). Thus, cerebral
blood pressure may vary slightly from BBP. Future investigations exploring cognitive
function in patients with hypertension would benefit from measuring cerebral perfusion
pressure (CPP) in addition to BBP. This could help researchers determine the variability
in BP that exists between the CNS and systemic circulation and understand better when
the brain becomes vulnerable to hypertension-associated cognitive dysfunction.
The present investigation assessed cognitive function in the more prevalent and common
forms of diabetes mellitus (T1DM and T2DM) and HTN (essential hypertension).
However, various other types of diabetes and hypertension exist, including monogenic
diabetes syndromes (maturity-onset diabetes of the young), gestational diabetes, drug or
treatment-induced diabetes (e.g. prolonged glucocorticoid use), and treatment-resistant
diabetes/hypertension (Oparil et al., 2018; ADA, 2020). No studies to date have examined
cognitive function in these forms. Although less common than the well-established sub-
types, future studies investigating cognitive function in these rarer sub-types may provide
important novel insights into cognitive dysfunction. It could also assist researchers to
develop a profile of cognitive disposition unique to each form and enable comparisons of
patterns of cognitive decline between the different types.
259
Chapter 7.
Numerous glucose-lowering agents are available to assist clinicians to optimise

glycaemic control and to treat and manage T2DM. Although the anti-hyperglycaemic
benefits of these medications are well established, it is currently unknown whether they
confer neuroprotective effects. Sodium glucose co-transporter 2 inhibitors (SGLT2is, e.g.
empagliflozin, dapagliflozin, and ertugliflozin) and glucagon-like peptide-1 receptor
agonists (GLP-1RAs, e.g. semaglutide, dulaglutide, and liraglutide) are new classes of
anti-hyperglycaemic agents that have demonstrated favourable cardio-renal outcomes
independent of glycaemic control, including clinically-meaningful blood pressure
reductions and weight loss (Zinman et al., 2015). Current clinical guidelines recommend
cardiovascular risk factor management and individualised diabetes management
regimens, suitable for the patient and their capabilities, as standard of care for patients
suspected of having diabetes-associated cognitive decrements (Biessels & Despa, 2018;
Biessels & Whitmer, 2019; Srikanth et al., 2020). Given these medications have
demonstrated favourable cardio-renal outcomes independent of glycaemic control in
large randomised controlled trials (RCTs), it is plausible these therapies may confer
potential beneficial effects on cognition. Large, prospective, long-duration RCTs
exploring whether these medications preserve cognitive function, perhaps as a secondary
endpoint, would be invaluable. They could also contribute to guiding global health care
decisions concerning choice of anti-diabetic medication for patients with diabetes,
particularly those with established cognitive decrements.
Finally, the present study could have benefited from recording sensory-evoked event-
related potentials (ERPs). These are EEG-derived recordings reflective of cortical activity
in response to specific stimuli (e.g. auditory and visual) thought to indicate function of
neural circuits (Modi & Sahin, 2017). The P300 wave (positive spike in brain activity 300
milliseconds (ms) after presentation of stimulus), which is elicited by auditory stimuli, is
the most commonly investigated ERP and is posited to underlie attention and auditory
processing (Mulert et al., 2004; Howe, Bani-Fatemi, & De Luca, 2014). Appreciable
evidence suggests that abnormalities in the P300 waveform, notably decreased amplitude
or increased latency, reflect disrupted neuronal connectivity in frontal and/or parietal
brain circuits and inattentiveness (Howe, Bani-Fatemi, & De Luca, 2014; Modi & Sahin,
2017). Early studies have shown that patients with DM (T1DM and T2DM) demonstrate
prolonged P300 wave latencies and decreased amplitude, especially those with
longstanding, poorly-controlled diabetes (Tsalikian et al., 1981; Mooradian et al., 1988;
260
Chapter 7.
Tallroth et al., 1990). Similar P300 abnormalities have also been reported in subjects with
hypertension, with higher BP associated with more marked changes (de Quesada-
Martinez et al., 2005). Given both DM and HTN are associated with noticeable
abnormalities in the P300 brain potential, future investigations exploring cognitive
function in subjects with DM or HTN could benefit from recording and analysing other
ERPs linked to cognitive processes. The present study also recruited study participants
across a broad age range (18-80 years). The impact of ageing on underlying neural
architecture and hence neuronal oscillations is well established (Vlahou et al., 2014);
therefore, it should be controlled in future investigations, as in this study, to reduce its
moderating effects on EEG activity.
7.2 Conclusions
Diabetes mellitus (T1DM and T2DM) and hypertension (HTN) are prevalent, chronic
diseases associated with subtle cognitive dysfunction and an increased risk of cognitive
impairment (65% increased risk of dementia for T1DM (Smolina et al., 2015); 1.5 to 2.5
times increased risk for T2DM (Biessels, 2006; Cheng et al., 2014); risk of dementia in
hypertension unknown). These subtle cognitive decrements, which affect all age groups
differently and progress insidiously, can interfere with and complicate daily disease self-
management tasks (e.g. managing meals and medication, recognising hypoglycaemia,
etc.), especially in elderly populations (> 65 years of age). Although cognitive
dysfunction is being increasingly recognised in clinical practice as a complication of both
DM and HTN, with recommendations for managing patients reporting cognitive
complaints now included in professional clinical guidelines, clinicians still have difficulty
addressing diabetes and hypertension-associated cognitive complaints with patients
(Biessels & Whitmer, 2019). Awareness of cognitive dysfunction still also reportedly lags
behind that of other well-known diabetes and hypertension-linked complications (e.g.
retinopathy, nephropathy, neuropathy, stroke, etc.) but guidelines for evaluating and
diagnosing cognitive dysfunction in DM are developing, especially for T2DM (Figure
7.1). Important questions also remain, such as the selection of cognitive screening tools
to be used to detect the subtle cognitive dysfunction triggered by both conditions and the
target groups that should be screened. The frequency of screening and when screening
should commence, and whether early screening programmes could avert adverse
cognitive outcomes, also remain unclear.
261
c o g niti v e f u n cti o n h a v e alr e a d y b e e n n ot e d ( fi g ur e). T h e pr o c e s s f or d et e cti o n of c o g niti v e d y sf u n cti o n oft e n
W h et h er c o g niti v e s cr e e ni n g s h o ul d e xt e n d t o p e o pl e b e gi n s i n e pi s o di c cli ni c al c ar e, w h e n a p er s o n wit h
wit h di a b et e s w h o ar e y o u n g er t h a n 6 5 y e ar s of a g e i s l e s s di a b et e s or t h eir f a mil y m e m b er r e p ort s a c o n c er n a b o ut
cl e ar. T o d et er mi n e p e o pl e li k el y t o b e n e fit fr o m c h a n g e s i n t hi n ki n g or c o g niti v e f u n cti o n t o a g e n er al
s cr e e ni n g, i n di vi d u al s a g e d 4 5 – 6 0 y e ar s c a n b e str ati fi e d p h y si ci a n ( e g, pri m ar y or f a mil y p h y si ci a n, g eri atri C hciaa pt
n, er
or 7.
i nt o hi g h-ri s k or l o w-ri s k gr o u p s f or t h e i n ci d e n c e of i nt er ni st).4 8 D et e cti o n i n v ol v e s t w o p h a s e s: ri s k e v al u ati o n
d e m e nti a 2 0 – 4 0 y e ar s l at er, u si n g a pr e di cti o n s c or e s u c h a n d di a g n o si s. G ui d eli n e r e c o m m e n d ati o n s s u g g e st at
a s t h e C ar di o v a s c ul ar Ri s k Fa ct or s, A gi n g, a n d I n ci d e n c e l e a st a bri ef c o g niti v e a s s e s s m e nt (if a m or e c o m pr e h e n si v e
Di a g n osis of t y p e 2 Ris k f act ors f or c o g niti v e d ysf u ncti o n or T a r g et e d ass ess m e nt Sc or es s u g g est l o w
di a b et es r e p ort e d c o nc er ns ¥ D et ail e d m e dic al hist or y a n d e x a mi n ati o n pr o b a bilit y of
¥ S elf-r e p ort e d or i nf or m a nt-r e p ort e d c o nc er ns ¥ Bri ef c o g niti v e t esti n g ( e g, M o ntr e al C o g niti v e c o g niti v e d ysf u ncti o n
a b o ut c o g niti v e f u ncti o n Ass ess m e nt)
¥ O n e or m or e u n e x pl ai n e d f all
¥ Hist or y of r ec urr e nt h y p o gl yc a e mi a
¥ Diffi c ult y wit h di a b et es s elf- m a n a g e m e nt,
i ncl u di n g err ors i n s elf- a d mi nistr ati o n of dr u gs
¥ S y m pt o ms of d e pr essi o n, str ess, or b ot h
E xtr a vi gil a nc e i n t h os e a g e d 6 5 y e ars a n d ol d er
N o e vi d e nc e f or t h e Ass ess m e nt s u g g ests hi g h er pr o b a bilit y of M o nit or e v er y

b e n e Þt of l ar g e-sc al e c o g niti v e d ysf u ncti o n 1Ð 2 y e ars
p o p ul ati o n scr e e ni n g
of c o g niti v e f u ncti o n
i n p e o pl e wit h di a b et es
C o nsi d er s p eci alist r ef err al f or n e ur o ps yc h ol o gic al E xcl u d e or tr e at r e v ersi bl e c a us es, p artic ul arl y if
e v al u ati o n, l a b or at or y e v al u ati o n, i m a gi n g, or a r a pi d o ns et
c o m bi n ati o n of t h es e ass ess m e nts t o e n a bl e di a g n osis ¥ D eliri u m
of mi n or a n d m aj or n e ur oc o g niti v e dis or d ers, a n d ¥ M et a b olic or e n d ocri n e
u n d erl yi n g c a us es ¥ M et a b olic or e n d ocri n e d ysf u ncti o n
Fi g u r e 7. 1 . Fr a m e w or k f or di a g n osi n g a n d e v al u ati n g c o g niti v e d ysf u n cti o n i n T y p e

Fi g ur e: C o nc e pt u al fr a m e w or k f or t ar g et e d ris k e v al u ati o n a n d di a g n osis of c o g niti v e f u ncti o n i n p e o pl e wit h t y p e 2 di a b et es
A d a pt e d fr o m Si ncl air et al. 4 3
2 di a b et es m ellit us . A d a pt e d fr o m Sri k a nt h et al., ( 2 0 2 0, p 5 3 7).
w w w.t h el a nc et.c o m/ di a b et es- e n d ocri n ol o g y V ol 8 J u n e 2 0 2 0 537
C urr e nt gl o b al esti m at es s u g g est a p pr o xi m at el y 1 3 5. 6 milli o n ( 1 9. 3 %) p e o pl e w orl d wi d e ,

o v er t h e a g e of 6 5 y e ars, h a v e di a b et es (I D F, 2 0 1 9). T his n u m b er is e x p e ct e d t o d o u bl e
b y 2 0 4 5, risi n g t o 2 7 6. 2 milli o n ( 1 9. 6 %), pri m aril y attri b ut a bl e t o a n a g ei n g p o p ul ati o n
(I D F, 2 0 1 9). Si mil ar p att er ns i n pr e v al e n c e h a v e b e e n r e p ort e d f or h y p ert e nsi o n. Gi v e n
t h e tr e n ds i n pr e v al e n c e pr e di ct e d f or b ot h D M a n d H T N, a n d t h e i n cr e as e d c o-o c c urr e n c e
of b ot h di a b et es a n d h y p ert e nsi o n a n d c o g niti v e i m p air m e nt d u e t o i n cr e asi n g n u m b er s
of el d erl y p ati e nts, i t is cl e ar t h at o bj e cti v e n o n-i n v asi v e bi o m ar k ers of e arl y c o g niti v e
d ysf u n cti o n ar e ur g e ntl y r e q uir e d. I d e ntifi c ati o n of s uit a bl e s cr e e ni n g t o ols t o all o w
s cr e e ni n g f or t h e s u btl e c o g niti v e d ysf u n cti o n li n k e d t o t h es e c o n diti o ns w o ul d als o b e
a d v a nt a g e o us . S u c h m e as ur es c o ul d h el p cli ni ci a ns i d e ntif y p ati e nts at i n cr e as e d ris k of
d e v el o pi n g di a b et es -ass o ci at e d c o g niti v e d e cr e m e nts a n d h y p ert e nsi o n -ass o ci at e d
c o g niti v e d ysf u n cti o n. It will als o all o w e arl y stri n g e nt gl y c a e mi c a n d bl o o d pr ess ur e
c o ntr ol, a d e q u at e c ar di o v as c ul ar ris k f a ct or m a n a g e m e nt, a n d i n di vi d u alis e d di a b et es
c ar e. C urr e nt lit er at ur e a n d g ui d eli n es r e c o m m e n d t h es e str at e gi es t o r e d u c e t h e
d e v el o p m e nt of l o n g -t er m mi cr o- a n d m a cr o v as c ul ar c o m pli c ati o ns a n d t o i m pr o v e
q u alit y of lif e ( Q o L) (Bi ess els & D es p a, 2 0 1 8; Bi ess els et al., 2 0 2 0; A D A, 2 0 2 0). T his
m a y c o ntri b ut e t o r e d u ci n g n ot o nl y t h e s u bst a nti al s o ci o e c o n o mi c b ur d e n li n k e d t o
di a b et es a n d h y p ert e nsi o n -r el at e d c o m pli c ati o ns o n c ar ers, i n di vi d u al p ati e nts, a n d t h e
h e alt h c ar e s yst e m, but als o d el a y pr o gr essi o n to pr o gr essi v e, irr e v ersi bl e
n e ur o d e g e n er ati v e dis e as es, s u c h as A D a n d d e m e nti a.
262
Chapter 7.
The present investigation revealed that BP (SBP and DBP) and BGL are associated with
observable changes in oscillatory brain activity, which could be consistently detected
using non-invasive scalp EEG. The main changes found in EEG activity in response to
BP and BGL were: (1) noticeable increases in slow-wave brain activity (theta and delta),
marked over frontal and parietal brain regions, and (2) reductions in fast-wave brain
activities (alpha, beta, and gamma), which were evident over central and parietal brain
areas. These changes in cortical activity were particularly pronounced when BP and BGL
reached certain thresholds (BP: > 135/90 mm Hg; BGL: > 8mmol/L), potentially
suggesting the brain becomes vulnerable to insult from diabetes and hypertension-
associated cognitive dysfunction at these thresholds. Multiple associations were also
found between disease-specific variables (HbA1C, disease duration, and age of disease
onset) in the clinical groups and slow-wave activities (theta and delta). While Uhlhaas &
Singer (2010) advise the diagnostic specificity of altered neuronal oscillations should be
interpreted cautiously, the data from the present study suggest: (1) the EEG signal can
reliably and accurately detect changes in ongoing brain oscillatory activity linked to small
changes in BGL and BP, and (2) that the EEG could be a suitable neurophysiological
measure for detecting the subtle and slowly-progressing cognitive decrements linked to
both conditions, potentially identifying EEG as a non-invasive biomarker of early
cognitive dysfunction. These are commonly undetected by formal cognitive assessment
until frank and irreversible, due to their pernicious nature.
Although novel data were obtained in the present study, it is clear that larger, adequately-
powered investigations (cross-sectional and longitudinal) are required to understand
better the precise electrophysiological abnormalities that occur in patients with DM
(T1DM and T2DM) and HTN, especially the latter. Such studies will validate the
preliminary changes in EEG activity reported herein and help determine the main disease-
specific variables (e.g. disease duration, HbA1C, frequency of hypoglycaemia, etc.) that
contribute to the electrophysiological abnormalities commonly observed in patients with
DM and HTN. Continued investigation of the changes in EEG activity associated with
these conditions may enable early recognition of brain regions susceptible to insult,
allowing initiation and continuation of robust risk-reduction measures (e.g.
cardiovascular risk factor management, reduced dietary salt intake, etc.) currently
suggested by professional clinical guidelines. This may subsequently delay progression
263
Chapter 7.
to progressive, irreversible neurodegenerative conditions, such as AD and dementia, and

reduce the substantial socioeconomic costs linked to these conditions.
264
8. Appendices
8.1 Consent Form (Non-clinical and Clinical)
UNIVERSITY OF TECHNOLOGY, SYDNEY

CONSENT FORM
I __________________________agree to participate in the research project ‘Investigating cognitive

function in clinical and non-clinical samples using electroencephalography (EEG) and psychometric
assessment: A comparative study (Approval no: UTS HREC REF NO. 2014000110) being conducted by
George Kalatzis in the Neuroscience Research Unit, University of Technology, Sydney (UTS). Funding
for this research project has been provided by the School of Life Sciences (UTS).
I understand the purpose of this study is to explore associations between chronic medical disorders and
cognitive function. This has implications for identifying disease states that profoundly impair cognitive
function, those that accelerate cognitive decline, prompting larger human-based studies, and potentially
identifying EEG as a non-invasive biomarker of early cognitive decline.
I understand that participation in this research will involve resting (quiet sitting) and cognitive measures
(active mental stimulation). I am also aware that study participation will involve measurements of blood
pressure and blood glucose level and brain activity through non-invasive techniques, as well as the
completion of questionnaires on lifestyle factors, brain (cognitive) function. I understand this experimental
protocol will inflict minimal risk and/or inconvenience.
I also understand the study will involve screening for blood pressure. If classified as normal (non-clinical),
under circumstances I may be found to have high blood pressure (>160/100mmHg) throughout any period
of the study, involvement in the study will discontinue and I will be offered to be escorted to a doctor and/or
advised to consult a medical professional. If my blood pressure is found to be greater than 140/90 mmHg,
I will be notified to consult a doctor. If my blood pressure is found to be greater than 160/100 mmHg prior
to testing, I understand I will be excluded from further study participation.
Lastly, I also understand the study will involve measurement of blood glucose levels. This will be achieved
using a sterile, single-use lancet and reliable and validated blood glucometre and will ONLY be measured
before and after cognitive testing and will inflict minimal injury/pain.
I am aware I can contact the investigator (George Kalatzis) on or the principal supervisor
(Associate Professor Sara Lal) (02) 9514-1592 or [email protected]) if I have any concerns regarding
the research. I am also aware that I am able to withdraw my participation from this research project at any
time, without consequences, and without providing a reason.
I agree that George Kalatzis has answered my questions.
I agree that the data collected in this project may be published in a form that does not identify me in any
way.
________________________________________ ____/____/____
Signature (participant)
________________________________________ ____/____/____
Signature (researcher or delegate)
NOTE:
This study has been approved by the University of Technology, Sydney Human Research Ethics Committee (HREC). If you have any
complaints or reservations about any aspect of your participation in this research which you cannot resolve with the researcher, you may contact
the Ethics Committee through the Research Ethics Officer (ph: 02 9514 9772; [email protected]) and quote the UTS HREC
reference number. Any complaint you make will be treated in confidence and investigated fully and you will be informed of the outcome.
265
8.2 Emergency Protocol
General Emergency Protocol

*ALWAYS CALL SECURITY FIRST*
UTS Contacts
1. Dial/call UTS Security: dial “6” on an internal UTS phone or 9514 1192
2. Dial/call 000
3. Dial/call student medical services (9514 1177)
4. Dial/contact principal supervisor (Sara Lal – 9514 1592)
If required:
UTS Medical Centre

Student Services Unit
Tower Building 1, Level 6, UTS
Ph: (02) 9514 1177
Opening Hours:
Monday: 8:30am – 5:30pm
Tuesday: 8:30am – 5:15pm
Wednesday: 8:30am – 5:00pm
Thursday: 8:30am – 3:45pm
Friday: 8:30am – 4:45pm
Saturday and Sunday: Closed)
(Note: opening hours are approximate)
Broadway General Practice (External medical centre)

Level 1, Broadway Shopping Centre,
Bay Street, NSW, 2007
Ph: (02) 8245 1500
Opening Hours:
Monday – Wednesday: 8:30am – 7:00pm
Thursday: 8:30am – 8:00pm
Friday: 8:30am – 7:00pm
Saturday: 9:00am – 6:00pm
Sunday: 10:00am – 6:00pm
266
Student/Researcher Protocol
Inclusion Criteria
The inclusion criteria for the present study was based on the Lifestyle Appraisal
Questionnaire (LAQ) (Craig, Hancock, & Craig, 1996). Participants must meet the
following inclusion criteria to be eligible for participation: no severe concomitant disease,
no history of alcoholism and drug abuse, and no psychosis, psychological or intellectual
problems likely to limit compliance.
Before commencement of any human-related research study, after the

participant/volunteer has had a 10-minute sitting (rest) period, record three (3) sitting BP
measurements from the participant’s right arm. A standard sphygmomanometre or
reliable and validated digital BP monitor (Omron, etc.) should be used to record BP
measurements.
After the measurements, if the average of the three BP readings are >160/100mmHg or
>160mmHg for systolic alone or >100mmHg for diastolic BP alone, the participant will
not be included in the research study (see consent form in section 9.1 above) and will be
thanked for their time and offer to be escorted to the nearest medical centre. The
student/researcher must advise participant of their BP and encourage them to seek
medical attention.
In the clinical samples (see section 8.1 above) if refused to be escorted to a medical centre,
the participant may continue with the study (so long as they feel well enough to do so);
however, they are still advised to see a GP regarding their elevated BP.
Similarly, three BP readings are to be recorded at the end of the study (if the participant
qualified and underwent the study). If BP readings are >160/100mmHg or >160mmHg
for systolic alone or >100mmHg for diastolic BP alone, the participant is offered to be
escorted to the nearest medical centre and advised to see a GP regarding their BP.
Note: In any case BP is >140/90mmHg, advise the participant to consult their GP.
NOTE:
According to the Australian Heart Foundation (AHF) (www.heartfoundation.org.au)
new hypertension guidelines (2008):
Normal BP: < 120/80 mmHg
High to normal BP: 120-139/80-89 mmHg
Grade 1 (mild) hypertension: 140-159/90-99 mmHg
Grade 2 (moderate) hypertension: 160-179/100-109 mmHg
Grade 3 (severe) hypertension: > 180/110 mmHg
267
8.3 Chronic Disease Questionnaire (Diabetes Mellitus)
Neuroscience Research Unit

Diabetes Questionnaire
Name: Age:
Gender: Ethnicity:
1. Please circle which type of diabetes you have:
a) Type 1
b) Type 2
2. How long have you been diagnosed with diabetes?
3. Does anyone in your immediate family such as your siblings, parents or grandparents
have a confirmed diagnosis of diabetes mellitus?
4. How regularly do you monitor your blood glucose levels?
5. In the last 3 months, have you had your haemoglobin a1c/glycosylated haemoglobin
(HbA1C) measured by your doctor? If so, please list.
6. Do you at present take any medication(s) to control your diabetes? If so, please list
these medication(s).
7. How frequently do you take the medications listed above?
268
8. Over the years you have been diagnosed with diabetes have you experienced any severe
hypoglycaemic (low blood glucose) episodes that have caused disturbance to your daily
activities?
9. Do you engage in any physical/recreational activity to manage better your diabetes? If

so, please list all physical activities you undertake along with a rough approximation of
the time you spend on each activity per week.
10. What other measures do you undertake to control your blood glucose levels (BGL)
(e.g. diet, stringent glucose monitoring, limit alcohol consumption etc.)
11. Have you developed any other medical issues (blindness, kidney issues, tingling in
extremities, heart attack, stroke) from your diabetes?
12. On the scale below, please indicate how well you think you manage your diabetes (0=
poor, 10= excellent)
1 2 3 4 5 6 7 8 9 10
269
8.4 Chronic Disease Questionnaire (Hypertension)
Neuroscience Research Unit

Hypertension Questionnaire
Name: Age:
Gender: Ethnicity:
1. Based on Australian hypertension guidelines, please circle which category best

describes your degree of hypertension.
a) £120mmHg/80mmHg
b) ³120mmHg/80mmHg (High-normal)
c) ³140mmHg/90mmHg (Grade 1)
d) ³160mmHg/100mmHg (Grade 2)
e) ³180mmHg/110mmHg (Grade 3)
2. How regularly do you monitor your blood pressure?
3. How do you measure your blood pressure (e.g. manual sphygmomanometre, self-
reported automatic BP monitor, measured by physician?)
4. How often do you have your blood pressure measured and examined by your doctor?
5. Does anyone in your immediate family such as your siblings, parents or grandparents
have a confirmed diagnosis of hypertension?
270
6. How long have you been diagnosed with hypertension?
7. Do you at present take any medication(s) to control your blood pressure? If so, please
list these medication(s).
8. How frequently do you take the medications listed above?
9. What other measures (e.g. physical activity, meditation exercises, restrict sodium
intake, limit smoking) do you undertake to control your blood pressure?
10. Have you developed any adverse outcomes (e.g. stroke, coronary artery disease)
from your hypertension? If others, please list.
11. On the scale below, please indicate how well you think you manage your high blood
pressure (0= poor, 10= excellent)
1 2 3 4 5 6 7 8 9 10
271
��
��
��
��
��
��
��
��
��
��
��
��
��
��
��
�
8.6 Study Summary Sheet
Summary of Research
(to be completed immediately after each lab study)
Date: ________________
Researcher: ______________
Participant: ______________
1. Provide a brief summary of the study (tick one of the following):

□ The study went smoothly
□ There were some issues
□ There were major issues
2. General account and summary of the study (detail in a few lines or more):
______________________________________________________________________
______________________________________________________________________
______________________________________________________________________
__________________________________________________________________
3. Were there any ‘out-of-the-ordinary’ events or issues in this lab study? Yes / No
If yes, provide more details:
______________________________________________________________________
______________________________________________________________________
______________________________________________________________________
__________________________________________________________________
4. Was there an emergency situation in the lab? Yes / No

If yes, provide more details:
______________________________________________________________________
______________________________________________________________________
______________________________________________________________________
__________________________________________________________________
Note:
If you answered YES to Question 3, you must notify a senior researcher and/or
responsible academic/deputy responsible academic immediately. If you answered YES to
Question 4, you SHOULD have followed the emergency protocol and you MUST report
the incident using HIRO (Hazard and Incident Reporting Online) via the UTS Safety and
Wellbeing website (https://www.uts.edu.au/about/safety-wellbeing/hazard-and-incident-
response/hiro-support). Subsequently you must then notify a senior researcher or
responsible/deputy responsible academic as soon as possible.
273
N E U R O S CI E N C E
R E S E A R C H U NI T
D o Y O U h a v e DI A B E T E S ?
Do Y O U w a nt t o c o ntri b ut e t o i m p ort a nt r e s e ar c h ?
Crit eri a : 1 8- 8 0, Di a b et es m ellit us ( T y p e 1 or T y p e 2)
G e or g e K al at zi s
G e or g e. K al at zis @st u d e nt. uts. e d u. a u
* A L L d at a o bt ai n e d i s tr e at e d c o nfi d e nti all y, r e m ai n s s e c ur e a n d c o m pl et el y a n o n y m o u s
N O T E: T hi s st u d y h a s b e e n a p pr o v e d b y t h e U ni v er sit y of T e c h n ol o g y, S y d n e y H u m a n R e s e ar c h Et hi c s C o m mitt e e ( H R E C: 2 0 1 4 0 0 0 1 1 0). If y o u h a v e

any co m pl ai nt s or r e s er v ati o n s a b o ut a n y a s p e ct of y o ur p arti ci p ati o n i n t hi s r e s e ar c h, w hi c h y o u c a n n ot r e s ol v e wit h t h e r e s e a r c h er, y o u m a y c o nt a ct t h e
Et hi c s C o m mitt e e t hr o u g h t h e R e s e ar c h Et hi c s Offi c er ( p h: 0 2 9 5 1 4 9 6 1 5, R e s e ar c h. Et hi c s @ ut s. e d u. a u) a n d q u ot e t h e U T S H R E C r ef e r e n c e n u m b er.
Any co m pl ai nt y o u m a k e will b e tr e at e d i n c o nfi d e n c e a n d i n v e sti g at e d f ull y a n d y o u will b e i nf or m e d of t h e o ut c o m e.
8.8 Participant Remuneration Form (Clinical)
Monetary payment of $50 AUD for participation in the study conducted by PhD
candidate, George Kalatzis, of the Neuroscience Research Unit (NRU), School of Life
Sciences (SoLS), University of Technology, Sydney (UTS). This research project is being
supervised by Associate Professor Sara Lal.
Date of Participation
Address
Account Name
BSB
Account Number
Financial Institution
Payment will be deposited into the account above via UTS Financial Services Unit
fortnightly.
Signed: __________________________
(Participant)
_________________________________
Researcher (George Kalatzis)
_________________________________
Financial Delegate (Deanne Koelmeyer) (School Manager, School of Life Sciences)
INTERNAL USE
275
8.9 Breakdown of glucose-lowering therapies (n = 30)
Metformin
DPP4i
SGLT2i
Insulin
SU
Key:
DPP4i – Dipeptidyl peptidase-4 inhibitor

SGLT2i – Sodium Glucose Co-Transporter 2 inhibitor
SU – Sulfonylurea
276
8.10 Breakdown of anti-hypertensive medication (n = 15)
ACE Inhibitor
ARB
CCB
Key:
ACE – Angiotensin Converting Enzyme Inhibitor

ARB – Angiotensin Receptor Blocker
CCB – Calcium Channel Blocker
277
9. References
Abbott, N.J., Ronnback, L., & Hansson, E. (2006). Astrocyte-endothelial interactions at

the blood-brain barrier. Nature Reviews Neuroscience, 7, 41-53.
Abraham, T.M., & Fox, C.S. (2013). Implications of rising prediabetes prevalence.
Diabetes Care, 36, 2139-41.
Adleman, N.E., Menon, V., Blasey, C.M., White, C.D., Warsofsky, I.S., Glover, G.H., et
al. (2002). A Development fMRI Study of the Stroop Colour-Word Task. NeuroImage,
16, 61-75.
Adrian, E.D., & Matthews, B.H. (1934). The Berger rhythm: potential changes from the
occipital lobes in man. Brain, 57, 355-85.
Akisaki, T., Sakurai, T., Takata, T., Umegaki, H., Araki, A., Mizuno, S., et al. (2006).
Cognitive dysfunction associates with white matter hyperintensities and subcortical
atrophy on magnetic resonance imaging of the elderly diabetes mellitus Japanese elderly
diabetes intervention trial (J-EDIT). Diabetes Metabolism & Research Reviews, 22, 376-
84.
Alexander, D.M., Arns, M.W., Paul, R.H., Rowe, D.L., Cooper, N., Esser, A.H., et al.
(2006). EEG markers for cognitive decline in elderly subjects with subjective memory
complaints. Journal of Integrative Neuroscience, 5, 49-74.
Alvarez, J.A., & Emory, E. (2006). Executive Function and the Frontal Lobes: A Meta-
Analytic Review. Neuropsychology Review, 16, 17-42.
Alzheimer’s Australia (2017). Economic cost of Dementia in Australia 2016-2056.

https://treasury.gov.au/sites/default/files/2019-03/C2016-052_NATSEM-for-
Alzheimers-Australia.pdf. Accessed 16/04/2018.
278
American Diabetes Association. (2020). Introduction: Standards of Medical Care in
Diabetes: 2020. Diabetes Care, 43 (Supplement 1), S152-63.
Amiel, S.A., Pottinger, R.C., Archibald, H.R., et al. (1991). Effect of antecedent glucose
control on cerebral function during hypoglycaemia. Diabetes Care, 14, 109-18.
An, Y.J., Jung, K.Y., Kim, S.M., Lee, C., & Kim, D.W. (2015). Effects of blood glucose
levels on resting-stage EEG and attention in healthy volunteers. Journal of Clinical
Neurophysiology, 32, 51-56.
Apostolova, L.G., & Thompson, P.M. (2007). Brain mapping as a tool to study
neurodegeneration. Neurotherapeutics, 4, 387-400.
Areosa Sastre, A., Vernooij, R.W., Gonzalez-Colaco Harmand, M., & Martinez, G.
(2017). Effect of the treatment of Type 2 diabetes mellitus on the development of
cognitive impairment and dementia. The Cochrane Database of Systematic Reviews, 15,
CD003804.
Armulik, A., Genove, G., & Betsholtz, C. (2011). Pericytes: developmental,

physiological, and pathophysiological perspectives, problems, and promises.
Developmental Cell, 21, 193-215.
Arnold, S.E., Arvanitakis, Z., Macauley-Rambach, S.L., Koenig, A.M., Wang, H.Y.,
Ahima, R.S., et al. (2018). Brain insulin resistance in type 2 diabetes and Alzheimer
disease: concepts and conundrums. Nature Reviews Neurology, 14, 168-81.
Arnsten, A.F. (2009). Stress signalling pathways that impair prefrontal cortex structure
and function. Nature Reviews Neuroscience, 10, 410-22.
Ashcroft, F.M., & Rorsman, P. (2012). Diabetes mellitus and the b cell: the last ten years.
Cell, 148, 1160-7.
279
Asvold, B.O., Sand, T., Hestad, K., & Bjørgas, M.R. (2010). Cognitive function in Type
1 diabetic adults with early exposure to severe hypoglycaemia: A 16-year follow-up
study. Diabetes Care, 33, 1945-47.
Atkinson, M.A., Eisenbarth, G.S., Michels, A.W. (2014). Type 1 diabetes. Lancet, 383,
69-82.
Australian Government Department of Health. (2020). Body mass index (BMI).

http://healthyweight.health.gov.au/wps/portal/Home/get-started/are-you-a-healthy-
weight/bmi/. Accessed 25/06/2020.
Australian Institute of Health & Welfare (2018). Australia’s Health.
Babiloni, C., Triggiani, A.L., Lizio, R., Cordone, S., Tattoli, G., Bevilacqua, V., et al.
(2016). Classification of single normal and Alzheimer’s disease individuals from crotical
sources of resting EEG rhythms. Frontiers in Neuroscience, 10, 47.
Babiloni, C., Carducci, F., Lizio, R., Vecchio, F., Baglieri, A., Bernardini, S., et al.
(2013). Resting state cortical electroencephalographic rhythms are related to gray matter
volume in subjects with mild cognitive impairment and Alzheimer’s disease. Human
Brain Mapping, 34, 1427-46.
Babiloni, C., Frison, G.B., Vecchio, F., Lizio, R., Pievani, M., Cristina, G., et al. (2011a).
Stability of clinical condition in mild cognitive impairment is related to cortical sources
of alpha rhythms: an electroencephalography study. Human Brain Mapping, 32, 1916-
31.
280
Babiloni, C., Lizio, R., Vecchio, F., Frison, G.B., Pievani, M., Geroldi, C., et al. (2010).
Reactivity of cortical alpha rhythms to eye opening in mild cognitive impairment and
Alzheimer’s disease: an EEG study. Journal of Alzheimer’s Disease, 22, 1047-64.
Babiloni, C., Visser, P.J., Frisoni, G., De Deyn, P.P., Bresciani, L., Jelic, et al. (2010).
Cortical sources of resting EEG rhythms in mild cognitive impairment and subjective
memory. Neurobiology of Aging, 31, 1787-98.
Back, S.H., & Kaufman, R.J. (2012). Endoplasmic reticulum stress and type 2 diabetes.
Annual Review of Biochemistry, 81, 767-93.
Baglietto-Vargas, D., Shi, J., Yaeger, D.M., Ager, R., & LaFerla, F.M. (2016). Diabetes
and Alzheimer’s disease crosstalk. Neuroscience & Biobehavioural Reviews, 64, 272-87.
Bartoloni, E., Alunno, A., & Gerli, R. (2018). Hypertension as a cardiovascular risk factor
in autoimmune rheumatic diseases. Nature Reviews Cardiology, 15, 33-44.
Baumgart, M., Snyder, H.M., Carrillo, M.C., Fazio, S., Kim, H., & Johns, H. (2015).
Summary of the evidence on modifiable risk factors for cognitive decline and dementia:
A population-based perspective. Alzheimer’s & Dementia: The journal of the Alzheimer’s
Association, 11, 718-26.
Bazanova, O.M., & Vernon, D. (2014). Interpreting EEG alpha activity. Neuroscience &
Biobehavioural Reviews, 44, 94-110.
Beagley, J., Guariguata, L., Weil, C., & Motala, A.A. (2014). Global estimates of
undiagnosed diabetes mellitus in adults. Diabetes Research & Clinical Practice, 103,
150-60.
Beall, C., Ashford, M.L., McCrimmon, R.J. (2012). The physiology and pathophysiology
of the neural control of the counterregulatory response. American Journal of Physiology:
Regulatory, Integrative and Comparative Physiology, 302, r215-23.
281
Beratis, I.N., Rabavilas, A., Papadimitriou, G.N., & Papageorgiou, C. (2010). Effect of
handedness on the Stroop Colour Word Task. Laterality: Asymmetries of Body, Brain &
Cognition, 15, 597-609.
Berger, H. (1929). Über das Elektrenkephalogramm des Menschen. Archiv für

Psychiatrie und Nervenkrankheiten, 87, 527-80.
Biau, D.J., Kerneis, S., & Porcher, R. (2008). Statistics in brief: The importance of sample
size in the planning and interpretation of medical research. Clinical Orthopaedics &
Related Research, 466, 2282-88.
Biessels, G.J., Nobili, F., Teunissen, C.E., Simo, R., & Scheltens, P. (2020).
Understanding multifactorial brain changes in type 2 diabetes: a biomarker perspective.
The Lancet Neurology, 20, 1-12.
Biessels, G.J., & Whitmer, A. (2019). Cognitive dysfunction in diabetes: how to

implement emerging guidelines. Diabetologia, 63, 3-9.
Biessels, G.J., & Despa, F. (2018). Cognitive decline and dementia in diabetes mellitus:
mechanisms and clinical implications. Nature Reviews Endocrinology, 14, 591-604.
Biessels, G.J., & Reagan, L.P. (2015). Hippocampal insulin resistance and cognitive
dysfunction. Nature Reviews Neuroscience, 16, 660-71.
Biessels, G.J., & Reijmer, Y.D. (2014). Brain changes underlying cognitive dysfunction
in diabetes: What can we learn from MRI? Diabetes, 63, 2244-52.
Biessels, G.J., Strachan, M.W., Visseren, F.L., Kappelle, L.J., & Whitmer, R.A. (2014).
Dementia and cognitive decline in Type 2 diabetes and prediabetic stages: towards
targeted interventions. Lancet Diabetes Endocrinology, 2, 246-55.
Biessels, G.J., Deary, I.J., & Ryan, C.M. (2008). Cognition and diabetes: a lifespan
perspective. The Lancet Neurology, 7, 184-90.
282
Biessels, G.J., Staekenborg, S., Brunner, E., Brayne, C., & Scheltens, P. (2006). Risk of
dementia in diabetes mellitus: a systematic review. Lancet Neurology, 5, 64-74.
Bjorgaas, M., Sand, T., Vik, T., Jorde, R. (1998). Quantitative EEG during controlled
hypoglycaemia in diabetic and non-diabetic children. Diabetic Medicine, 15, 30-7.
Birns, J., & Kalra, L. (2009). Cognitive function and hypertension. Journal of Human
Hypertension, 23, 86-96.
Braak, H., Thal, D.R., Ghebremedhin, E., & Del Tredici, K. (2011). Stages of the
pathologic process in Alzheimer disease: age categories from 1 to 100 years. Journal of
Neuropathology and Experimental Neurology, 70, 960-9.
Brands, A.M., Van den Berg, E., Manschot, S.M., Biessels, G.J., Kappelle, L.J., De Haan,
E.H., et al. (2007). A detailed profile of cognitive dysfunction and its relation to
psychological distress in patients with type 2 diabetes mellitus. Journal of the
International Neuropsychological Society, 13, 288-97.
Brands, A.M., Kessels, R.P., Hoogma, R.P., Henselmans, J.P., Boter, J.W., Kappelle, L.J,
et al. (2006). Cognitive performance, psychological well-being, and brain magnetic
resonance imaging in older patients with type 1 diabetes. Diabetes, 55, 1800-6.
Brands, A.M., Biessels, G.J., Haan, E.H., Kappelle, L.J., & Kessels, R.P. (2005). The
effects of Type 1 Diabetes on cognitive performance. Diabetes Care, 28, 726-35.
Brismar, T., Hyllienmark, L., Ekberg, K., & Johansson, B.L. (2002). Loss of temporal
lobe beta power in young adults with type 1 diabetes mellitus. Neuroreport, 13, 2469-73.
Brown, T., Mapleston, J., & Nairn, A. (2011). Convergent validity of the Occupational
Therapy Adult Perceptual Screening Test (OT-APST) with two other cognitive-
perceptual tests. British Journal of Occupational Therapy, 74, 562-72.
Brownlee, M. (2001). Biochemistry and molecular cell biology of diabetic complications.

Nature, 414, 813-20.
283
Bruce, D.G., Davis, W.A., Casey, G.P., Clarnette, R.M., Brown, S.G., Jacos, I.G., et al.
(2009). Severe hypoglycaemia and cognitive impairment in older patients with diabetes:
the Fremantle diabetes study. Diabetologia, 52, 1808-15.
Bubu, O.M., Brannick, M., Mortimer, J., Umasabor-Bubu, O., Sebastiao, Y.V., Wen, Y.,
et al. (2017). Sleep, cognitive impairment, and Alzheimer’s disease: A systematic review
and meta-analysis. Sleep, 40, 1.
Buckner, R.L. (2013). The cerebellum and cognitive function: 25 years of insight from
anatomy and neuroimaging. Neuron, 80, 807-15.
Burdakov, D., Jensen, L.T., Alexopoulos, H., Williams, R.H., Fearon, I.M., O’Kelly, I.,
et al. (2006). Tandem-pore K+ channels mediate inhibition of orexin neurons by glucose.
Neuron, 50, 711-22.
Burle, B., Spieser, L., Roger, C., Casini, L., Hasbroucq, T., & Vidal, F. (2015). Spatial
and temporal resolutions of EEG: Is it really black and white? A scalp current density
view. International Journal of Psychophysiology, 97, 210-20.
Button, K.S., Ioannidis, J.P., Mokrysz, C., Nosek, B.A., Flint, J., Robinson, E.S., et al.
(2013). Power failure: why small sample size undermines the reliability of neuroscience.
Nature Reviews Neuroscience, 14, 365-76.
Buzaki, G., Anastassiou, C.A., & Koch, C. (2012). The origins of extracellular fields and
currents – EEG, ECoG, LFP, and spikes. Nature Reviews Neuroscience, 13, 407-20.
Buzsaki, G., & Draguhn, A. (2004). Neuronal oscillations in cortical networks. Science,
304, 1926-9.
Cabeza, R.L., & Nyberg, L. (1997). Imaging cognition: an empirical review of PET
studies with normal subjects. Journal of Cognitive Neuroscience, 9, 1-26.
Campisi, P., & La Rocca, D. (2014). Brain waves for automatic biometric-based user
recognition. IEEE Transactions on Information Forensics and Security, 9, 782-800.
284
Chalmers, T., Maharaj, S., Lees, T., Lin, C.T., Newton, P., Clifton-Bligh, R., et al. (2020).
Impact of acute stress on cortical electrical activity and cardiac autonomic coupling.
Journal of Integrative Neuroscience, 19, 239-48.
Chalmers, J., Risk, M.T., Kean, D.M., Grant, R., Ashworth, B., & Campbell, I.W. (1991).
Severe amnesia after hypoglycaemia. Clinical, psychometric, and magnetic resonance
imaging correlations. Diabetes Care, 14, 922-5.
Chen, K.H., Henderson, V.W., Stolwyk, R.J., Dennerstein, L., & Szoeke, C. (2015).
Prehypertension in midlife is associated with worse cognition a decade litter in middle-
aged and older women. Age and ageing, 44, 439-45.
Cheng, C., Lin, C.H., Tsai, Y.W., Tsai, C.J., Chou, P.H., & Lan, T.H. (2014). Type 2
diabetes and antidiabetic medications in relation to dementia diagnosis. Journal of
Gerontology, Series A, Biological Sciences and Medical Sciences, 69, 1299-1305.
Cholerton, B., Baker, L.D., & Craft, S. (2013). Insulin, Cognition, and dementia.
European Journal of Pharmacology, 719, 170-9.
Christman, A.L., Matushita, K., Gottesman, R.F., Mosley, T., Alonso, A., Coresh, J., et
al. (2011). Glycated haemoglobin and cognitive decline: The Atherosclerosis Risk in
Communities (ARIC) study. Diabetologia, 54, 1645-52.
Citron, M. (2010). Alzheimer’s disease: strategies for disease modification. Nature

Reviews Drug Discovery, 9, 387-98.
Cobb, C.A., & Cole, M.P. (2015). Oxidative and nitrative stress in neurodegeneration.
Neurobiology of Disease, 84, 4-21.
Cobb, S.R., Buhl, E.H., Halasy, K., Paulsen, O., & Somogyi, P. (1995). Synchronisation
of neuronal activity in hippocampus by individual GABAergic interneurons. Nature, 378,
75-8.
285
Coben, L.A., Danziger, W., & Storandt, M. (1985). A longitudinal EEG study of mild
senile dementia of Alzheimer type: changes at 1 year and at 2.5 years.
Electroencephalography & Clinical Neurophysiology, 61, 101-12.
Cohen, D.L., & Townsend, R.R. (2017). Blood pressure goals, variability, and SGLT2
blockade in CKD. Nature Reviews Nephrology, 13, 75-6.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155-59.
Cooray, G., Hyllienmark, L., & Brismar, T. (2011). Decreased cortical connectivity and
information flow in type 1 diabetes. Clinical Neurophysiology, 122, 1943-50.
Cox, D.J., Kovatchev, B.P., Gonder-Frederick, L.A., Summers, K.H., McCall, A.,
Grimm, K.J., et al. (2005). Relationships between Hyperglycemia and Cognitive
Performance Among Adults with Type 1 and Type 2 Diabetes. Diabetes Care, 28, 71-77.
Crane, P.K., Walker, R., Hubbard, R.A., Li, G., Nathan, D.M., Zheng, H. (2013). Glucose
Levels and Risk of Dementia. The New England Journal of Medicine, 369, 540-48.
Crum, R.M., Anthony, J.C., Bassett, S.S., & Folstein, M.F. (1993). Population-based
norms for the Mini-Mental State Examination by age and education level. The Journal of
the American Medical Association, 269, 2386-91.
Cryer, P.E. (2006). Mechanisms of sympathoadrenal failure and hypoglycaemia in

diabetes. Journal of Clinical Investigation, 116, 1470-3.
Cukierman-Yaffe, T., Gerstein, H.C., Williamson, J.D., Lazar, R.M., Lovato, L., Miller,
M.E., et al. (2009). Relationship between baseline glycaemic control and cognitive
function in Individuals with Type 2 diabetes and other cardiovascular risk factors: The
Action to Control Cardiovascular Risk in Diabetes-Memory in Diabetes (ACCORD-
MIND) trial. Diabetes Care, 32, 221-26.
286
Da Silva, F.L. (1991). Neural mechanisms underlying brain waves: from neural
membranes to networks. Electroencephalography and Clinical Neurophysiology, 79, 81-
93.
DeCarli, C. (2003). Mild cognitive impairment: prevalence, prognosis, aetiology, and

treatment. Lancet Neurology, 2, 15-21.
DeFronzo, R.A., Norton, L., Abdul-Ghani, M. (2017). Renal, metabolic and

cardiovascular considerations of SGLT2 inhibition. Nature Reviews Nephrology, 13, 11-
26.
DeFronzo, R.A., Ferrannini, E., Groop, L., Henry, R.R., Herman, W.H., Holst, J.J., et al.
(2015). Type 2 diabetes mellitus. Nature Reviews Disease Primers, 1, 1-22.
De Galan, B., Zoungas, S., Chalmers, J., Anderson, C., Dofouil, C, Pillai, A., et al. (2009).
Cognitive function and risk of cardiovascular disease and hypoglycaemia in patients with
type 2 diabetes: the Action in Diabetes and Vascular Disease: Preterax and Diamicron
Modified Release Controlled Evaluation (ADVANCE) trial. Diabetologia, 52, 2328-36.
Deiber, M.P., Missionnier, P., Bertrand, O., Gold, G., Fazio-Costa, L., Ibanez, V., et al.
(2007). Distinction between perceptual and attentional processing in working memory
tasks: a study of phase-locked and induced oscillatory brain dynamics. Journal of
Cognitive Neuroscience, 19, 158-72.
de la Monte, S.M., & Wands, J.R. (2008). Alzheimer’s disease is type 3 diabetes-evidence
reviewed. Journal of Diabetes, Science and Technology, 2, 1103-13.
Dementia Australia. (2020). Vascular Dementia. https://www.dementia.org.au/about-

dementia/types-of-dementia/vascular-dementia accessed 13/02/2020.
Dementia Australia. (2018). Dementia Prevalence Data 2018-2058. The National Centre
for Social and Economic Modelling (NATSEM): The University of Canberra.
287
Deshpande, A.D., Harris-Hayes, M., & Schootman, M. (2008). Epidemiology of Diabetes
and Diabetes-Related Complications. Physical Therapy, 88, 1254-64.
de Quesada-Martinez, M.E., Blanco-García, M., & Díaz-de Quesada, L. (2005).

Cognitive evoked potentials in uncomplicated hypertensive patients. Revista de
neurologia, 40, 523-31.
Ding, J., Davis-Plourde, K.L., Sedaghat, S., Tully, P.T., Wang, W., Phillips, C., et al.
(2020). Antihypertensive medications and risk for incident dementia and Alzheimer’s
disease: a meta-analysis of individual participant data from prospective cohort studies.
The Lancet Neurology, 19, 61-70.
Dong, Y.H., Sharma, V.K., Chan, B.P., Venketasubramanian, N., Teoh, H.K., Seet, R.C.,
et al. (2010). The Montreal Cognitive Assessment (MoCA) is superior to the Mini-Mental
State Examination (MMSE) for the detection of vascular cognitive impairment after
stroke. Journal of the Neurological Sciences, 299, 15-18.
Drane, D.L., Yuspeh, R.L., Huthwaite, J.S., Klingler, L.K., Foster, L.M., Mrazik, M., et
al. (2003). Healthy Older Adult Performance on a Modified Version of the Cognistat
(NCSE): Demographic Issues and Preliminary Normative Data. Journal of Clinical and
Experimental Neuropsychology, 25, 133-44.
Drane, D.L. & Ossato, S.S. (1997). Using the Neurobehavioral Cognitive Status
Examination as a Screening Measure for Older Adults. Archives of Clinical
Neuropsychology, 12, 139-43.
Drummond, G.R., Vinh, A., Guzik, T.J., & Sobey, C.G. (2019). Immune mechanisms of
hypertension. Nature Reviews Immunology, 19, 517-32.
Duschek, S., & Schandry, R. (2007). Reduced brain perfusion and cognitive performance
due to constitutional hypotension. Clinical Autonomic Research, 17, 69-76.
Duschek, S., Meinhardt, J., & Schandry, R. (2006). Reduced cortical activity due to
chronic low blood pressure: an EEG study. Biological Psychology, 72, 241-50.
288
Eeg-Olofsson, O. (1977). Hypoglycaemia and neurological disturbances in children with
diabetes mellitus. Acta Paediatrica, 66, 91-96.
Eeg-Olofsson, O., Petersen, I., & Sellden, U. (1971). The development of the
electroencephalogram in normal children from the age of 1 through 15 years. Paroxysmal
activity. Neuropadiatrie, 2, 375-404.
Eeg-Olofsson, O., & Petersen, I. (1966). Childhood diabetic neuropathy: A childhood and
neurophysiological study. Acta Paediatrica, 55, 163-76.
Eisenstein, N., Engelhart, C.I., Johnson, V., Wolf, J., Williamson, J., & Losonczy, M.B.
(2002). Normative Data for Healthy Elderly Persons with the Neurobehavioral Cognitive
Status Exam (Cognistat). Applied Neuropsychology, 9, 110-13.
Elahi, F.M., & Miller, B.L. (2017). A clinicopathological approach to the diagnosis of
dementia. Nature Reviews Neurology, 13, 457-76.
Elias P.K., Elias M.F., Robbins M.A., & Budge M.M. (2004). Blood pressure-related
cognitive decline: does age make a difference? Hypertension, 44, 631-66.
Elias, M.F., Wolf, P.A., D’Agostino, R.B., Cobb, J., & White, L.R. (1993). Untreated
blood pressure level is inversely related to cognitive functioning: the Framingham Study.
American Journal of Epidemiology, 138, 353-64.
Elsborg, R., Remvig, L., Beck-Nielsen, H., & Juhl, C. (2011). Detecting hypoglycaemia
by using the brain as a biosensor. In P.A. Sierra (ed.) Biosensors for Health, Environment
and Biosecurity, ISBN: 978-953-307-443-6, InTechOpen, Available from:
https://www.intechopen.com/books/biosensors-for-health-environment-and-
biosecurity/detecting-hypoglycemia-by-using-the-brain-as-a-biosensor
Engel, A.K., Fries, P. & Singer, W. (2000). Dynamic predictions: oscillations and
synchrony in top-down processing. Nature Reviews Neuroscience, 2, 704-16.
289
Engelhart, C., Eisenstein, N., & Meininger, J. (1994). Psychometric Properties of the
Neurobehavioral Cognitive Status Exam. The Clinical Neuropsychologist, 8, 405-15.
Exalto, L.G., Biessels, G.J., Karter, A.J., Huang, E.S., Katon, W.J., Minkoff, J.R., et al.
(2013). Risk score for prediction of 10-year dementia risk in individuals with type 2
diabetes: a cohort study. The Lancet Diabetes & Endocrinology, 1, 183-90.
Faigle, R., Sutter, R., & Kaplan, P.W. (2013). The electroencephalography of
encephalopathy in patients with endocrine and metabolic disorders. Journal of Clinical
Farmer, M.E., White, L.R., Abbott, R.D., Kittner, S.J., Kaplan, E., Wolz, M.M., et al.
(1987). Blood pressure and cognitive performance: the Framingham Study. American
Journal of Epidemiology, 126, 1103-10.
Farrer, L.A., Cupples, L.A., Haines, J.L., Hyman, B., Kukull, W.A., Mayeux, R., et al.
(1997). Effects of age, sex, and ethnicity on the association between apolipoprotein E
genotype and Alzheimer’s disease: a meta-analysis. APOE and Alzheimer Disease meta-
analysis consortium. The Journal of the American Medical Association, 278, 1349-56.
Fazli, S., Mehnert, J., Steinbrink, J., Curio, G., Villringer, A., Mueller, K.R, et al. (2012).
Enhanced performance by a hybrid NIRS-EEG brain computer interface. NeuroImage,
59, 519-29.
Feinkohl, I., Price, J.F., Strachan, M.W., & Frier, B.M. (2015). The impact of diabetes on
cognitive decline: potential vascular, metabolic, and psychosocial risk factors.
Alzheimer’s Research Therapy, 7, 46.
Feinkohl, I., Aung, P.P., Keller, M., Robertson, C.M., Morling, J.R., McLachlan, S., et
al. (2014). Severe hypoglycaemia and cognitive decline in older people with type 2
diabetes: the Edinburgh type 2 diabetes study. Diabetes Care, 37, 507-15.
Ferguson, S.C., Blane, A., Wardlaw, J.M., (2005). Influence of an early-onset age of
T1DM on cerebral structure and cognitive function. Diabetes Care, 28, 1431-37.
290
Ferrannini, E., & Cushman, W.C. (2012). Diabetes and hypertension: the bad
companions. Lancet, 380, 601-10.
Folstein, M.F., McHugh, P.R., & Folstein, S.E. (1975). “Mini-mental state”. A practical
method for grading the cognitive status of patients for the clinician. Journal of Psychiatric
Research, 12, 189-98.
Forbes, J.M., & Cooper, M.E. (2013). Mechanisms of diabetic complications.

Physiological Review, 93, 137-88.
Foreman, B., & Claassen, J. (2012). Quantitative EEG for the detection of brain
ischaemia. Critical Care, 16, 216.
Forouzanfar, M.H., Liu, P., Roth, G.A., Ng, M., Biryukov, S., Marczak, L., et al. (2017).
Global burden of hypertension and systolic blood pressure of at least 110 to 115 mmHg,
1990-2015. The Journal of the American Medical Association, 317, 165-82.
Franklin, S.S., Thijs, L., Hansen, T.W., O’Brien, E.O., & Staessen, J.A. (2013). White-
Coat hypertension: New insights from recent studies. Hypertension, 62, 982-87.
Frederick, S. (2005). Cognitive Reflection and Decision Making. The Journal of

Economic Perspectives, 19, 25-42.
Frier, B.M. (2014). Hypoglycaemia in diabetes mellitus: epidemiology and clinical

implications. Nature Reviews Endocrinology, 10, 711-22.
Fries, P. (2009). Neuronal Gamma-Band Synchronization as a Fundamental Process in

Cortical Computation. Annual Review of Neuroscience, 32, 209-24.
Galasko, D., Klauber, M.R., Hofstetter, C.R., Salmon, D.P., Lasker, B., & Thai, L.J.
(1996). The Mini Mental State Examination in the early diagnosis of Alzheimer’s disease.
Archives of Neurology, 47, 49-52.
291
Garrett, K.D., Browndyke, J.N., Whelihan, W., Paul, R.H., DiCarlo, M., Moser, D.J., et
al. (2004). The neuropsychological profile of vascular cognitive impairment – no
dementia: comparisons to patients at risk for cerebrovascular disease and vascular
dementia. Archives of Clinical Neuropsychology, 19, 745-57.
Gaudieri, P.A., Chen, R., Greer, T.F., & Holmes, C.S. (2008). Cognitive function in
children with Type 1 Diabetes; a meta-analysis. Diabetes Care, 31, 1892-97.
Gauthier, S, Resiberg, B., Zaudig, M., Petersen, RC., Ritchie, K., Broich, K., et al. (2006).
Mild cognitive impairment. Lancet, 367, 1262-70.
Geijselaers, S.L., Sep, S.J., Stehouwer, C.D., & Biessels, G.J. (2017). Glucose regulation,
cognition, and brain MRI in type 2 diabetes: a systematic review. Lancet Diabetes
Endocrinology, 3, 75-89.
George, P.M., & Steinberg, G.K. (2015). Novel stroke therapeutics: unravelling stroke
pathophysiology and its impact on clinical treatments. Neuron, 87, 297-309.
Giacino, J.T., Fins, J.J., Laureys, S., Schiff, N.D. (2014). Disorders of consciousness after
acquired brain injury: the state of the science. Nature Reviews Neurology, 10, 99-114.
Giaquinto, S., & Nolfe, G. (1986). The EEG in the normal elderly: a contribution to the
interpretation of aging and dementia. Electroencephalography & Clinical
Gispen, W.H., & Biessels, G.J. (2000). Cognition and synaptic plasticity in diabetes
mellitus. Trends in Neurosciences, 23, 542-9.
Global Burden of Disease Study. (2018). Global, regional, and national comparative risk
assessment of 84 behavioural, environmental, and occupational, and metabolic risks or
clusters of risks for 195 countries and territories, 1990-2017: a systematic analysis for the
global burden of disease study 2017. Lancet, 392, 1923-94.
292
Goldstein, F.C., Levey, A.I., & Steenland, N.K. (2013). High blood pressure and
cognitive decline in mild cognitive impairment. Journal of the American Geriatric
Society, 61, 67-73.
Goldstein, M.A., & Silverman, M.E. (2006). Autonomics and Cognition. Clinical
Autonomic Research, 16, 86-9.
Gomèz, C.M., Vaquero, E., & Vasquèz-Marrufo, M. (2004). A neurocognitive model of

short-term sensory and motor preparatory activity in humans. Psicologica, 25, 229-71.
Graveling, A.J., Deary, I.J., Frier, B.M. (2013). Acute hypoglycemia impairs executive
cognitive function in adults with and without type 1 diabetes. Diabetes Care, 36, 3240-6.
Grayson, B.E., Seeley, R.J., & Sandoval, D.A. (2013). Wired on sugar: the role of the
CNS in the regulation of glucose homeostasis. Nature Reviews Neuroscience, 14, 24-37.
Greenblatt, M., Murray, J., & Root, H.F. (1946). Electroencephalographic studies in
diabetes mellitus. New England Journal of Medicine, 234, 119-21.
Gudala, K., Bansal, D., Schifano, F., & Bhansali, A. (2013). Diabetes mellitus and risk
of dementia: A meta-analysis of prospective observational studies. Journal of Diabetes
Investigation, 4, 640-50.
Gunning-Dixon, F.M., & Raz, N. (2000). The cognitive correlates of white matter
abnormalities in normal aging: a quantitative review. Neuropsychology, 14, 224-32.
Guo, Z., Fratiglioni, L., Winblad, B., Viitanen, M. (1997). Blood pressure and
performance on the Mini-Mental State Examination in the very old: cross-sectional and
longitudinal data from the Kungsholmen Project. American Journal of Epidemiology,
145, 1106-13.
Hamm, V., Heraud, C., Cassel, J.C., Mathis, C., & Goutagny, R. (2015). Precocious
alterations of brain oscillatory activity in Alzheimer’s disease: a window of opportunity
for early diagnosis and treatment. Frontiers in Cellular Neuroscience, 9, 491.
293
Hampel, H., & Lista, S. (2016). The rising global tide of cognitive impairment. Nature
Reviews Neurology, 12, 131-32.
Harcourt, B.E., Penfold, S.A., & Forbes, J.M. (2013). Coming full circle in diabetes
mellitus: from complications to initiation. Nature Reviews Endocrinology, 9, 113-23.
Harrington, F., Saxby, B.K., McKeith, I.G., Wesness, K., & Ford, G.A. (2000). Cognitive
performance in hypertensive and normotensive older subjects. Hypertension, 36, 1079-
82.
Hauser, E., Strohmayer, C., Seidl, R., Birnbacher, R., Lischka, A., & Schoer, E.J. (1995).
Quantitative EEG in young diabetics. Child Neurology, 10, 330-34.
Hayward, R.A., Reaven, P.D., Wiitala, W.L., Bahn, G.D., Reda, D.J., Ge, L., et al. (2015).
Follow-up of glycaemic control and cardiovascular outcomes in type 2 diabetes. The New
England Journal of Medicine, 372, 2197-2206.
Hedden, T., & Gabrieli, J.D. (2004). Insights into the ageing mind: a view from cognitive
neuroscience. Nature Reviews Neuroscience, 5, 87-96.
Herculano-Houzel, S. (2009). The human brain in numbers: a linearly scaled-up primate

brain. Frontiers in Human Neuroscience, 3, 31.
Herzog, R.I, & Sherwin, R.S. (2012). Can tight glycaemic control in diabetes benefit
cognition. Nature Reviews Neurology, 8, 124-26.
Hinkle, D.E., Wiersma, W., & Jurs, S.G. (2003). Applied Statistics for the Behavioural
Sciences. (5th Edition). Boston: Houghton Mifflin.
Holden, R.J. (2011). Cognitive performance-altering effects of electronic medical

records: An application of the human factors paradigm for patient safety. Cognition,
Technology, and Work, 13, 11-29.
294
Holman, R.R., Paul, S.K., Bethel, M.A., Matthews, D.R., & Neil, A.W. (2008). 10-year
follow-up of intensive glucose control in type 2 diabetes. The New England Journal of
Medicine, 359, 1577-89.
Homack, S., & Riccio, C.A. (2004). A meta-analysis of the sensitivity and specificity of
the Stroop Colour and Word Test with children. Archives of Clinical Neuropsychology,
19, 725-43.
Horani, M.H., & Mooradian, A.D. (2003). Effect of diabetes on the blood brain barrier.
Current Pharmaceutical Design, 9, 833-40.
Hossmann, K.A. (1994). Viability thresholds and the penumbra of focal ischemia. Annals
of Neurology, 36, 557-65.
Houmani, N., Vialatte, F., Gallego-Jutgla, E., Dreyfus, G., Nguygen-Michel, G., Mariani,
V, et al. (2018). Diagnosis of Alzheimer’s disease with electroencephalography in a
differential framework. PLoS One, 13, e0193607.
Howe, A.S., Bani-Fatemi, A., De Luca, V. (2014). The clinical utility of the auditory
P300 latency subcomponent event-related potential in preclinical diagnosis of patients
with mild cognitive impairment and Alzheimer’s disease. Brain and Cognition, 86, 64-
74.
Huang, C., Wahlund, L., Dierks, T., Julin, P., Winblad, B., & Jelic, V. (2000).
Discrimination of Alzheimer’s disease and mild cognitive impairment by equivalent EEG
sources: a cross-sectional and longitudinal study. Clinical Neurophysiology, 111, 1961-
7.
Huisa, B.N., Caprihan, A., Thompson, J., Prestopnik, J., Qualls, C.R., & Rosenberg, G.A.
(2015). Long-term blood-brain permeability changes in Binswanger disease. Stroke, 46,
2413-18.
295
Huster, R.J., Enriquez-Geppert, S., Lavallee, C.F., Falkenstein, M., & Hermann, C.S.
(2013). Electroencephalography of response inhibition tasks: Functional networks and
cognitive contributions. International Journal of Psychophysiology, 87, 217-233.
Hwang, J.J., Jiang, L., Rangel, E.S., Fan, X., Ding, Y., Lam, W., et al. (2019). Glycaemic
variability and brain glucose levels in type 1 diabetes. Diabetes Care, 68, 163-71.
Hyllienmark, L., Maltez, J., Dandenell, A., Ludvigsson, J., & Brismar, T. (2005). EEG
abnormalities with and without relation to severe hypoglycaemia in adolescents with type
1 diabetes. Diabetologia, 48, 412-19.
Iadecola, C., Gottesman, R.F. (2019). Neurovascular and cognitive dysfunction in

hypertension. Circulation Research, 124, 1025-44.
Iadecola, C., Duering, M., Hachinski, V., Joutel, A., Pendlebury, S.T., Schneider, J.A., et
al. (2019). Vascular cognitive impairment and dementia. Journal of American College of
Cardiology. 73, 3326-44.
Iadecola, C., Yaffe, K., Biller, J., Bratzke, L., Faraci, F.M., Gorelick, P.B., et al. (2016).
Impact of hypertension on cognitive function: A scientific statement from the American
Heart Association. Hypertension, 68, e67-94.
Iadecola, C. (2014). Hypertension and Dementia. Hypertension, 64, 3-5.
Iadecola, C. (2004). Neurovascular regulation in the normal brain and in Alzheimer’s

disease. Nature Reviews Neuroscience, 5, 347-60.
Inkster, B., & Frier, B.M. (2012). The effects of acute hypoglycaemia on cognitive
function in Type 1 diabetes. British Journal of Diabetes Vascular Disease, 12, 221-26.
International Diabetes Federation (IDF) Diabetes Atlas 9th Edition (2019), 2.
Izzo, J.L., Schuster, D.B., & Engel, G.L. (1953). The electroencephalogram of patients
with diabetes mellitus. Diabetes, 2, 93-9.
296
Jacobson, A.M., Musen, G., Ryan, C.M., Silvers, N., Cleary, P., Waberski, B., et al.
(2007). Long-term effects of diabetes and its treatment on cognitive function. New
Jasper, H.H. (1958). Report of the committee on methods of clinical examination in

electroencephalography. Electroencephalography and Clinical Neurophysiology, 10,
370-5.
Jelic, V., Johansson, S.E., Almkvist, O., Shigeta, M., Julin, A., Nordberg, B., et al. (2000).
Quantitative electroencephalography in mild cognitive impairment: longitudinal changes
and possible prediction of Alzheimer’s disease. Neurobiology of Aging, 21, 533-40.
Jelic, V., Shigeta, M., Julin, P., Almkvist, O., Winblad, B., & Wahlund, L.O. (1996).
Quantitative electroencephalography power and coherence in Alzheimer’s disease and
mild cognitive impairment. Dementia, 7, 314-23.
Jennings, J.R., Mendelson, D.N., Muldoon, M.F., Ryan, C.M., Gianaros, P.J., Raz, N, et
al. (2012). Regional grey matter shrinks in hypertensive individuals despite successful
lowering of blood pressure. Journal of Human Hypertension, 26, 295-305.
Jennings, J.R., Muldoon, M.F., Ryan, C., Price, J.C., Greer, P., Sutton-Tyrrell, K., et al.
(2005). Reduced cerebral blood flow response and compensation among patients with
untreated hypertension. Neurology, 64, 1358-65.
Jeong, J. (2004). EEG dynamics in patients with Alzheimer’s disease. Clinical

Neurophyshiology, 115, 1490-1505.
Jia, X., & Kohn, A. (2011). Gamma Rhythms in the Brain. PLos Biology, 9, e1001045.
Jordan, K.G. (2004). Emergency EEG and continuous EEG monitoring in acute
ischaemic stroke. Journal of Clinical Neurophysiology, 21, 341-52.
297
Juhl, C.B., Hojlund, K., Elsborg, R., et al. (2010). Automated detection of
hypoglycaemia-induced EEG changes recorded by subcutaneous electrodes in subjects
with type 1 diabetes – the brain as a biosensor. Diabetes Research Clinical Practice, 88,
22-8.
Kähönen-Väre, M., Brunni-Hakala, S., Lindroos, M., Pitkala, K., Strandberg, T., &
Tilvis, R. (2004). Left ventricular hypertrophy and blood pressure as predictors of
cognitive decline in old age. Aging Clinical & Experimental Research, 16, 147-52.
Kaiboriboon, K., Luders, H.O., Hamaneh, M., Turnbull, J., & Lhatoo, S.D. (2012). EEG
source imaging in epilepsy – practicalities and pitfalls. Nature Reviews Neurology, 8,
498-507.
Kalmijn, S., Feskens, E.J., Launer, L.J., Stijnen, T., & Kromhout, D. (1995). Glucose
intolerance, hyperinsulinaemia and cognitive function in a general population of elderly
men. Diabetologia, 38, 1096-102.
Kandimalla, R., Thirumala, V., & Reddy, P.H. (2017). Is Alzheimer’s disease a Type 3
Diabetes? A critical appraisal. Biochimica et biophysica Acta Molecular Basis of Disease,
1863, 1078-89.
Karama, S., Ducharme, S., Corley, J., Chouinard-Decorte, F., Starr, J.M., Wardlaw, J.M.,
et al. (2015). Cigarette smoking and thinning of the brain’s cortex. Molecular Psychiatry,
20, 778-785.
Keil, A., Debener, S., Gratton, G., Junghofer, M., Kappenman, E.S., Luck, S.J., et al.
(2014). Committee report: publication guidelines and recommendations for studies using
electroencephalography and magnetoencephalography. Psychophysiology, 51, 1-21.
Kesse-Guyot, E., Julia, C., Andreeva, V., Fezeu, L., Hercberg, S., & Galan P. (2015).
Evidence of a cumulative effect of cardiometabolic disorders at midlife and subsequent
cognitive function. Age Ageing, 44, 648–54.
298
Khanna, A., Pascual-Leone, A., Michel, C.M., & Farzan, F. (2015). Microstates in
resting-state EEG: Current status and future directions. Neuroscience and Biobehavioural
Reviews, 49, 105-13.
Kiernan, R.J., Mueller, J., Langston, J.W., & Van Dyke, C. (1987). The Neurobehavioral
Cognitive Status Examination: A Brief but Differentiated Approach to Cognitive
Assessment. Annals of Internal Medicine, 107, 481-5.
Kilander, L., Nyman, H., Boberg, M., & Lithell, H. (2000). The association between low
diastolic blood pressure in middle age and cognitive function in old age: a population-
based study. Age Ageing, 29, 243-48.
Kilander, L., Nyman, H., Bober, M., Hansson, L, & Lithell, H. (1998). Hypertension is
related to cognitive impairment: a 20-year follow-up of 999 men. Hypertension, 31, 780-
86.
Kisler, K., Nelson, A.R., Montagne, A., & Zlokovic., B.V. (2017). Cerebral blood flow
regulation and neurovascular dysfunction in Alzheimer’s disease. Nature Reviews
Neuroscience, 18, 419-434.
Kivipelto, M., Mangialasche, F., & Ngandu, T. (2018). Lifestyle interventions to prevent
cognitive impairment, dementia and Alzheimer disease. Nature Reviews Neurology, 14,
653-66.
Klimesch, W. (2012). Alpha-band oscillations, attention, and controlled access to stored

information. Trends in Cognitive Sciences, 16, 606-17.
Klimesch, W., Sauseng, P., & Hanslmayr, S. (2007). EEG alpha oscillations: the
inhibition-timing hypothesis. Brain Research Review, 53, 63-88.
Klimesch, W. (1997). EEG-alpha rhythms and memory processes. International Journal

of Psychophysiology, 26, 319-40.
299
Knyazev, G.G. (2007). Motivation, emotion, and their inhibitory control mirrored in brain
oscillations. Neuroscience & Biobehavioural Reviews, 31, 377-95.
Kodl, C.T., & Seaquist, E.R. (2008). Cognitive dysfunction and diabetes mellitus.
Endocrine Reviews, 29, 494-511.
Koekkoek, P.S., Kappelle, L.J., Berg, E., Rutten, G.E., & Biessels, G.J. (2015). Cognitive
function in patients with diabetes mellitus: guidance for daily care. The Lancet Neurology,
14, 329-40.
Kopf, S.R., & Baratti, C.M. (1999). Effects of post-training administration of insulin on
retention of a habituation response in mice: participation of a central cholinergic
mechanism. Neurobiology of learning and memory, 71, 50-61.
Kovatchev, B.P. (2017). Metrics for glycaemic control – from HbA1C to continuous
glucose monitoring. Nature Reviews Endocrinology, 13, 425-36.
Kuehl, M., & Stevens, M.J. (2012). Cardiovascular autonomic neuropathies as

complications of diabetes mellitus. Nature Reviews Endocrinology, 8, 405-16.
Lal, S.K., & Craig, A. (2002). Driver fatigue: electroencephalography and psychological
assessment. Psychophysiology, 39, 313-21.
Lal, S.K. & Craig, A. (2001). A critical review of the psychophysiology of driver fatigue.
Biological Psychology, 55, 173-94.
Lancu, I., & Olmer, A. (2006). The minimental state examination--an up-to-date review.
Harefuah, 145, 687-90.
Lande, M.B., Kaczorowski, J.M., Auinger, P., Schwartz, G.J., & Weitzman, M. (2003).
Elevated blood pressure and decreased cognitive function among school-age children and
adolescents in the United States. Journal of Paediatrics, 143, 720–24.
300
Languren, G., Montiel, T., Ramirez-Lugo, L., Balderas, I., Sanchez-Chavez, G., Sotres-
Bayon, F., Bermudez-Rattoni, R, et al. (2017). Recurrent moderate hypoglycaemia
exacerbates oxidative damage and neuronal death leading to cognitive dysfunction after
the hypoglycaemic coma. Journal of Cerebral Blood Flow & Metabolism, 39, 808-21.
Languren, G., Montiel, T., Julio-Amilipas, A., & Massieu, L. (2013). Neuronal damage
and cognitive impairment associated with hypoglycaemia: An integrated view.
Neurochemistry International, 63, 331-43.
Launer, L.J., Masaki, K., Petrovitch, H., Foley, D., Haylik, R.J. (1995). The association
between midlife blood pressure levels and late-life cognitive function: The Honolulu-
Asia Aging Study. The Journal of the American Medical Association, 1995, 274, 1846-
51.
Lawes, C. M., Vander Hoorn, S. & Rodgers, A. (2008). Global burden of blood-pressure-
related disease, 2001. Lancet, 371, 1513–18.
Lawson, J.S., Erdahl, D.L., Monga, T.N., Bird, C.E., Donald, M.W., Surridge, D.H., et
al. (1984). Neuropsychological function in diabetic patients with neuropathy. The British
Journal of Psychiatry, 145, 263-8.
Le Berre, A.P., Rauchs, A., La Joie, R., Mézenge, F., Boudehent, C., Vabret, F., et al.
(2014). Impaired decision-making and brain shrinkage in alcoholism. European
Psychiatry, 29, 125-33.
Lewington, S., Clarke, R., Qizilbash, N., Peto, R., & Collins, R. (2002). Age-specific
relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data
for one million adults in 61 prospective studies. Lancet, 360, 1903-13.
Lewis, R., Rennick, P.M. (1979). Manual for the Repeatable Cognitive-Perceptual-Motor
Battery. Axon, Grosse Point, MI.
301
Lewis, L.D., Ljunggren, B., Ratcheson, R.A., & Siesjo, B.K. (1974). Cerebral energy
state in insulin-induced hypoglycaemia, related to blood glucose and to EEG. Journal of
Neurochemistry, 23, 673-9.
Liotti, M., Woldorff, M.G., Perez, R., & Mayberg, H.S. (2000). An ERP study of the
temporal course of the Stroop colour-word interference effect. Neuropsychologia, 38,
701-11.
Logue, P.E., Tupler, L.A., D’Amico, C., & Schmidt, F.A. (1993). The Neurobehavioral
Cognitive Status Examination: Psychometric properties in use with psychiatric inpatients.
Journal of Clinical Psychology, 49, 80-89.
Lord, C., Brugha, T.S., Charman, T., Cusack, J., Dumas, G., Frazier, T., et al. (2020).
Autism spectrum disorder. Nature Reviews Disease Primer, 6, 1-23.
Lorenzi, M., Healy, D.P., Hawkins, R., Printz, J.M., & Printz, M.P. (1986). Studies on
the permeability of the blood-brain barrier in experimental diabetes. Diabetologia, 29,
58-62.
Luchsinger, J.A., Reitz, C., Patel, B., Tang, M.X., Manly, J.J., & Mayeux, R. (2007).
Relation of diabetes to mild cognitive impairment. Archives of Neurology, 64, 570-5.
Lund Research Ltd. (2019). Laerd Statistics (Online). Available:

(https://statistics.laerd.com).
Lyngdoh, T., Viswanathan, B., Kobrosly, R., van Wjingaarden, E., Huber, B., Davidson,
P.W., et al. (2013). Blood pressure and cognitive function: a prospective analysis among
adolescents in the Seychelles. Journal of Hypertension, 31, 1-15.
Macaulay, C., Battista, M., Lebby, P.C., & Mueller, J. (2003). Geriatric performance on
the Neurobehavioural Cognitive Status Examination (Cognistat): What is normal?
Archives of Clinical Neuropsychology, 18, 463-71.
302
MacLeod, C.M., & MacDonald, P.A. (2000). Interdimensional interference in the Stroop
effect: uncovering the cognitive and neural anatomy of attention. Trends in Cognitive
Sciences, 4, 383-91.
Maharaj, S., Lees, T., & Lal, S. (2019). Negative mental states and their association to
the cognitive function of nurses. Journal of Psychophysiology, 33, 207-18.
Mani, K.S., & Townsend, H.R. (1964). The EEG in benign intracranial hypertension.
Electroencephalography & Clinical Neurophysiology, 16, 604-10.
Marco-Pallares, J., Munte, T.F., & Rodriguez-Fornells, A. (2015). The role of high-
frequency oscillatory activity in reward processing and learning. Neuroscience &
Marioni, R.E., Chatfield, M., Brayne, C., & Matthews, F.E. (2011). The reliability of
assigning individuals to cognitive states using the Mini Mental-State Examination: a
population-based prospective cohort study. BMC Medical Research Methodology, 11, 1-
6.
Martini, F.H., Ober, W.C., & Nath, J.L. (2011). Visual Anatomy & Physiology (3rd
Edition).
Massar, S.A., Kenemans, J.L., & Schutter, D.J. (2014). Resting-state EEG theta activity
and risk learning: sensitivity to reward or punishment. International Journal of
Psychophysiology, 91, 172-7.
Masters, C.L., Bateman, R., Blennow, K., Rowe, C.C., Sperling, R.A., & Cummings, J.L.
(2015). Alzheimer’s disease. Nature Reviews Disease Primers, 1, 15056.
McBride, J.C., Zhao, X., Munro, N.B., Smith, C.D., Jicha, G.A., Hively, L., et al. (2014).
Spectral and complexity analysis of scalp EEG characteristics for mild cognitive
impairment and early Alzheimer’s disease. Computer Methods and Programs in
Biomedicine, 114, 153-63.
303
McCarley, R.W., Salisbury, D.F., Hirayasu, Y., Yurgelun-Todd, D.A., Tohen, M., Zarate,
C., et al. (2002). Association between smaller left posterior superior temporal gyrus
volume on magnetic resonance imaging and smaller left temporal P300 amplitude in first-
episode schizophrenia. Archives of General Psychiatry, 59, 321-31.
McCrimmon, R.J., Ryan, C.M., & Frier, B.M. (2012). Diabetes and cognitive
dysfunction. The Lancet, 379, 2291-99.
McNay, E.C., & Cotero, V.E. (2010). Mini-review: impact of recurrent hypoglycaemia
on cognitive and brain function. Physiology & Behaviour, 100, 234-8.
McNay, E.C., & Gold, P.E. (2001). Age-related differences in hippocampal extracellular
fluid glucose concentration during behavioural testing and following systemic glucose
administration. Journals of Gerontology Series A – Biological Sciences & Medical
Sciences, 56, B66-71.
McNay, E.C., Fries, T.M., & Gold, P.E. (2000). Decreases in rat extracellular
hippocampal glucose concentration associated with cognitive demand during a spatial
task. Proceedings of the National Academy of Science in the United States of America,
97, 2881-5.
Mert, A., & Akan, A. (2018). Emotion recognition from EEG signals by using
multivariate empirical mode decomposition. Pattern Analysis & Applications, 21, 81-89.
Messier, C. (2005). Impact of glucose tolerance and type 2 diabetes on cognitive aging.
Neurobiology of Aging, 26, 26-30.
Mezuk, B., Eaton, W.W., Albrecht, S., & Golden, S.H. (2008). Depression and type 2
diabetes over the lifespan: a meta-analysis. Diabetes Care, 31, 2383-90.
Michel, C.M., & Murray, M.M. (2012). Towards the utilization of EEG as a brain imaging
tool. NeuroImage, 61, 371-85.
304
Miles, W.R., & Root, H.F. (1922). Psychologic tests applied to diabetic patients. Archives
of Internal Medicine, 30, 767-77.
Mills, K.T., Stefanescu, A., & He, J. (2020). The global epidemiology of hypertension.
Nature Reviews Nephrology, 16, 223-37.
Misra, U.K., & Kalita, J. (2018). Electroencephalography in metabolic disorders. In

Clinical Electroencephalography, Second Edition (pp. 240-50), ISBN: 978-81-312-5295-
6, Elsevier, 2018.
Modi, M.E., & Sahin, M. (2017). Translational use of event-related potentials to assess
circuit integrity in ASD. Nature Reviews Neurology, 13, 160-70.
Mehvari, J., Gholami Motlagh, F., Najafi, M., Ghazvini, M., Naeini, A., & Zare, M.
(2016). Effects of vitamin E on seizure frequency, electroencephalogram findings, and
oxidative stress status of refractory epileptic patients. Advanced Biomedical Research, 5,
36.
Monette, M.C., Baird, A., & Jackson, D.L. (2014). A meta-analysis of cognitive
functioning in nondemented adults with type 2 diabetes mellitus. Canadian Journal of
Diabetes, 38, 401-8.
Mooradian, A.D., Perryman, K., Fitten, J., Kavonian, G.D., & Morley, J.E. (1988).
Cortical function in elderly non-insulin dependent diabetic patients: behavioural and
electrophysiologic studies. Archives of Internal Medicine, 148, 2369-72.
Morley, J.E. (2017). The diabetic brain. Nature Reviews Endocrinology, 13, 570-71.
Mormino, E.C., Betensky, R.A., Hedden, T., Schultz, A.P., Amariglio, R.E., Rentz, D.M.,
et al. (2014). Synergistic effect of b-amyloid and neurodegeneration on cognitive decline
in clinically normal individuals. The Journal of the American Medical Association
Neurology, 71, 1379-85.
305
Mulert, C., Pogarell, O., Juckel, G., Rujescu, D., Giegling, I., Rupp, D., et al. (2004). The
neural basis of the P300 potential: focus on the time-course of the underlying cortical
generators. European Archives of Psychiatry and Clinical Neuroscience, 254, 190-98.
Munshi, M., Grande, L., Hayes, M., Ayres, D., Suhl, E., Capelson, R., et al. (2006).
Cognitive dysfunction is associated with poor diabetes control in older adults. Diabetes
Care, 29, 1794-9.
Musen, G., Lyoo, I.K., Sparks, C.R., Weinger, K., Hwang, J., Ryan, C.M., et al. (2006).
Effects of type 1 diabetes on gray matter density as measured by voxel-based
morphometry. Diabetes, 55, 326-33.
Nasreddine, Z.S., Phillips, N.A., Bedirian, V., Charbonneau, S., Whitehead, V., Collin,
I., et al. (2005). The Montreal Cognitive Assessment, MoCA: A brief screening tool for
mild cognitive impairment. Journal of the American Geriatrics Society, 53, 695-9.
Nathan, D.M. (1993). Long-term complications of diabetes mellitus. New England

Journal of Medicine, 328, 1676-85.
National Centre for Social and Economic Modelling (NATSEM). (2017). Economic cost
of dementia in Australia: 2016-2056. University of Canberra.
Nilsson, P., Gullberg, G., Ekesbo, R., Von Schenck, H., & Gustafson, L. (1998). No
impaired cognitive function in treated patients with mild-moderate hypertension
compared to normotensive controls. Blood Pressure, 7, 209-13.
Ninomiya, T. (2014). Diabetes mellitus and dementia. Current Diabetes Reports, 14, 487.
Nøkleby, K., Boland, E., Bergersen, H., Schanke, A.K., Farner, L., Wagle, J., et al.
(2008). Screening for cognitive deficits after stroke: a comparison of three screening
tools. Clinical Rehabilitation, 22, 1095-1104.
Nordberg, A. (2015). Dementia in 2014: Towards early diagnosis in Alzheimer disease.

Nature Reviews Neurology, 11, 69-70.
306
Norton, S., Matthews, F.E., Barnes, D.E., Yaffe, K., & Brayne, C. (2014). Potential for
primary prevention of Alzheimer’s disease: an analysis of population-based data. The
Lancet Neurology, 13, 788-94.
Novak, V., & Hajjar, I. (2010). The relationship between blood pressure and cognitive
function. Nature Reviews Cardiology, 7, 686-98.
Obisesan, T.O. (2009). Hypertension and Cognitive Function. Clinics in Geriatric

Medicine, 25, 259-88.
Obisesan, T.O., Obisesan, O.A., Martins, S., Alamgir, L., Bond, V., Maxwell, C., et al.
(2008). High blood pressure, hypertension, and high pulse pressure are associated with
poorer cognitive function in persons aged 60 and older: the third national health and
nutrition examination survey. Journal of American Geriatric Society, 56, 501-09.
O’Bryant, S.E., Humphreys, J.D., Smith, G.E., Ivnik, R.J., Graff-Radford, N.R., Petersen,
R.C., et al. (2008). Detecting Dementia with the Mini-Mental State Examination
(MMSE) in Highly Educated Individuals. Archives of Neurology, 65, 963-67.
O’Donnell, M.J., Xavier, D., Lui, L., Zhang, H., Chin, S.L., Rao-Melacini, P., et al.
(2010). Risk factors for ischaemic and intracerebral haemorrhagic stroke in 22 countries
(the INTERSTROKE study): a case-control study. Lancet, 376, 112-23.
Oehlert, M.E., Hass, S.D., Freeman, M.R., Williams, M.D., Ryan, J.J., & Sumerall, S.W.
(1997). The Neurobehavioral Cognitive Status Examination: Accuracy of the “Screen-
Metric” Approach in a Clinical Sample. Journal of Clinical Psychology, 53, 733-37.
Oparil, S., Acelajado, M.C., Bakris, G.L., Berlowitz, D.R., Cifkova, R., Dominiczak.,
A.F., et al. (2018). Hypertension. Nature Reviews Disease Primers, 4, 18014.
Orkaby, A.R., Cho, K., Cormack, J., Gagnon, D.R., & Driver, J.A. (2017). Metfromin
versus sulfonylurea use and risk of dementia in US veterans aged >65 years with diabetes.
Neurology, 89, 1877-85.
307
Ostenson, C.G., Geelhoed-Dujivestijn, P., Lahtela, J., Weitgasser, R., Markert-Jensen,
M., & Pedersen-Bjergaard, U. (2014). Self-reported non-severe hypoglycaemic events in
Europe. Diabetic Medicine: a journal of the British Diabetic Association, 31, 92-101.
Palta, P., Schneider, A.L., Biessels, G.J., Touradji, P., Hill-Briggs, F. (2014). Magnitude
of cognitive dysfunction in adults with type 2 diabetes: a meta-analysis of six cognitive
domains and the most frequently reported neuropsychological tests within domains.
Journal of International Neuropsychological Society, 20, 278-91.
Patrone, C., Eriksson, O., & Lindholm, D. (2014). Diabetes drugs and neurological
disorders: new views and therapeutic possibilities. The Lancet: Diabetes &
Endocrinology, 2, 256-62.
Pendlebury, S.T., Cuthbertson, F.C., Welch, S.J., Mehta, Z., & Rothwell, P.M. (2010).
Underestimation of cognitive impairment by Mini-Mental State Examination verses the
Montreal Cognitive Assessment in patients with transient ischaemic attack and stroke.
Stroke, 41, 1290-3.
Pickering, T.G., Hall, J.E., Appel, L.J., Falkner, B.E, Graves, J., Hill, M.N., et al. (2005).
Recommendations for Blood Pressure Measurement in Humans and Experimental
Animals. Circulation, 111, 697-716.
Pijnenburg, Y.A, Made, Y., van Cappellen, A.M., Knol, D.L., Scheltens, P., & Stam, C.J.
(2004). EEG synchronisation likelihood in mild cognitive impairment and Alzheimer’s
disease during a working memory task. Clinical Neurophysiology, 115, 1332-9.
Pilli, R., Naidu, M.U., Pingali, U.R., Shobha, J.C., & Reddy, A.P. (2013). A computerized
Stroop Test for the Evaluation of Psychotropic Drugs in Healthy Participants. Indian
Journal of Psychological Medicine, 35, 180-89.
Pramming, S., Thorsteinsson, B., Stigsby, B., & Binder, C. (1988). Glycaemic threshold
for changes in electroencephalograms during hypoglycaemia in patients with insulin
dependent diabetes. British Medical Journal, 296, 665-7.
308
Prasad, S., Sajja, R.K., Naik, P, & Cucullo, L. (2014). Diabetes mellitus and blood-brain
barrier dysfunction: an overview. Journal of Pharmacovigilance, 2, 125.
Prince, M., Bryce, R., Albanese, E., Wimo, A., Rieiro, W., & Ferri, C.P. (2013). The
global prevalence of dementia: a systematic review and meta-analysis. Alzheimers &
Dementia: The journal of the Alzheimer’s Association, 9, 63-75.
Prins, N.D., & Scheltens, P. (2015). White matter hyperintensities, cognitive impairment
and dementia: an update. Nature Reviews Neurology, 11, 157-65.
Puente, E.C., Silverstein, J., Bree, A.J., Musikantow, D.R., Wozniak, D.F., Maloney, S.,
et al. (2010). Recurrent moderate hypoglycaemia ameliorates bbrain damage and
cognitive dysfunction induced by severe hypoglycaemia. Diabetes, 59, 1055-62.
Rachmiel, M., Cohen, M., Heymen, E., Lezinger, M., Inbar, D., Gilat, S, et al. (2016).
Hypergylcemia is associated with simultaneous alterations in electrical brain activity in
youths with type 1 diabetes mellitus. Clinical Neurophysiology, 127, 1188-95.
Rajagopalan, S., & Brook, R.D. (2012). Air pollution and type 2 diabetes: mechanistic
insights. Diabetes, 61, 3037-45.
Rawlings, A.M., Sharrett, A.R., Mosley, T.H., Ballew, S.H., Deal, G.A., & Selvin, E.
(2017). Glucose peaks and the risk of dementia and 20-year cognitive decline. Diabetes
Care, 40, 879- 86.
Reijmer, Y.D., van den Berg, E., Ruis, C., Kappelle, L.J., & Biessels, G.J. (2010).
Cognitive dysfunction in patients with type 2 diabetes. Diabetes/Metabolism Research &
Review, 26, 507-19.
Resiberg, B., Ferris, S.H., deLeon, M.J., & Crook, T. (1982). The global deterioration
scale for assessment of primary degenerative dementia. American Journal of Psychiatry,
139, 1136-9.
309
Reitz, C., Brayne, C., & Mayeux, R. (2011). Epidemiology of Alzheimer disease. Nature
Reviews Neurology, 7, 137-52.
Riboli, E., Hunt, K.J., Slimani, N., Ferrari, P., Norat, T., Fahey, M., et al. (2002).
European prospective investigation into cancer and nutrition (EPIC): study populations
and data collection. Public Health Nutrition, 5, 1113-24.
Richard, E., & Brayne, C. (2014). Dementia: Mild cognitive impairment—not always
what it seems. Nature Reviews Neurology, 10, 130-1.
Richmond, R., Jenaleen, L., Kay-Lambkin, F. (2011). Higher blood pressure associated
with higher cognition and functionality among centenarians in Australia. American
Journal of Hypertension, 24, 299-303.
Ricks, J., Molnar, M.Z., Kovesdy, C.P., Shah, A., Nissenson, A.R., Williams, M., et al.
(2012). Glycemic control and cardiovascular mortality in haemodialysis patients with
diabetes: a 6-year cohort study. Diabetes Care, 61, 708-15.
Ritter, P., & Villringer, A. (2006). Simultaneous EEG-fMRI. Neuroscience &

Roberts, R.O., Knopman, D.S., Mielke, M.M., Cha, R.H., Pankratz, V.S., Christianson,
T.J., et al. (2014). Higher risk of progression to dementia in mild cognitive impairment
cases who revert to normal. Neurology, 82, 317-25.
Roberts, R.O., Knopman, D.S., Przybelski, S.A., Mielke, M.M., Kantarci, K., Preboske,
G.M., et al. (2014). Association of type 2 diabetes with brain atrophy and cognitive
impairment. Neurology. 82, 1132-41.
Roberts, R.O., Geda, Y.E., Knopman, D.S., Teresa, J.H., Christianson, B.S., Pankratz,
V.S., et al. (2008). Duration and severity of diabetes are associated with mild cognitive
impairment. Archives of Neurology, 65, 1066-73.
310
Roh, E., Song, D.K., & Kim, M.S. (2016). Emerging role of the brain in the homeostatic
regulation of energy and glucose metabolism. Experimental Molecular Medicine, 11,
e216.
Roriz-Filho, J., Sa-Roriz, T.M., Rosset, I., Camozzato, A.L., Santos, A.C., Chaves, M.L.,
et al. (2009). (Pre)diabetes, brain aging, and cognition. Biochimica et Biophysica Acta,
1792, 432-43.
Rosenberg, G.A., Wallin, A., Wardlaw, J.M., Markus, H.S., Montaner, J., Wolfson, L., et
al. (2016). Consensus statement for diagnosis of subcortical small vessel disease. Journal
of Cerebral Blood Flow & Metabolism, 36, 6-25.
Rosenberg, G.A. (2012). Neurological diseases in relation to the blood-brain barrier.

Journal of Cerebral Blood Flow & Metabolism, 32, 1139-51.
Ross, I.S., Loeser, L.H. (1951). Electroencephalographic findings in essential

hypoglycaemia. Electroencephalography & Clinical Neurophysiology, 3, 141-8.
Rothberg, L.J., Lees, T., Clifton-Bligh, R., & Lal, S. (2016). Associations between heart
rate variability measures and blood glucose levels: implications for non-invasive glucose
monitoring for diabetes. Diabetes, Technology & Therapeutics, 18, 366-76.
Ruis, C., Biessels, G.J., Gorter, K.J., van den Donk, M., Kappelle, L.J., & Rutten, G.E.
(2009). Cognition in the early stage of type 2 diabetes. Diabetes Care, 32, 1261-5.
Ryan, C.M., vanDuinkerken, E., & Rosano, C. (2016). Neurocognitive consequences of

diabetes. The American Psychologist, 71, 563-76.
Ryan, C.M., Geckle, M.O., & Orchard, T.J. (2003). Cognitive efficiency declines over
time in adults with Type 1 diabetes: effects of micro- and macrovascular complications.
Diabetologia, 46, 940-48.
311
Ryan, C.M., Williams, T.M., Finegold, D.N., & Orchard, T.J. (1993). Cognitive
dysfunction in adults with type 1 (insulin-dependent) diabetes mellitus of long duration:
effects of recurrent hypoglycaemia and other complications. Diabetologia, 36, 329-34.
Sachdev, P.S., Blacker, D., Blazer, D.G., Ganguli, M., Jeste, D.V., Paulsen, J.S., et al.
(2014). Classifying neurocognitive disorders: the DSM-5 approach. Nature Reviews
Neurology, 10, 634-42.
Sakurai, T. (2014). The role of orexin in motivated behaviours. Nature Reviews

Sarter, M., & Bruno, J.P. (2000). Cortical cholinergic inputs mediating arousal,
attentional processing and dreaming: differential afferent regulation of the basal forebrain
by telencephalic and brainstem afferents. Neuroscience, 95, 933-52.
Sauseng, P., Griesmayr, B., Freunberger, R., & Klimesch, W. (2010). Control
mechanisms in working memory: a possible function of EEG theta oscillations.
Neuroscience & Biobehavioural Reviews, 34, 1015-22.
Sauseng, P., & Klimesch, W. (2008). What does phase information of oscillatory brain
activity tell us about cognitive processes? Neuroscience & Biobehavioural Reviews, 32,
1001-13.
Schroeder, C.E., & Lakatos, P. (2009). Low-frequency neuronal oscillations as

instruments of sensory election. Trends in Neuroscience, 32, 9-18.
Schulze, M.B., Kroke, A., Bergmann, M.M., & Boeing, H. (2000). Differences of blood
pressure estimates between consecutive measurements on one occasion: Implications for
inter-study comparability of epidemiologic studies. European Journal of Epidemiology,
16, 891-98.
Secnik, J., Cermakova, P., Fereshtehnejad, S.M., Dannberg, P., Johnell, K., Fastbom, J.,
et al. (2017). Diabetes in a large dementia cohort: clinical characteristics and treatment
from the Swedish dementia registry. Diabetes Care, 40, 1159-66.
312
Sierra, C., Domenech, M., Camafort, M., & Coca, A. (2012). Hypertension and mild
cognitive impairment. Current Hypertension Reports, 14, 548-55.
Sims-Robinson, C., Kim, B., Rosko, A., & Feldman, E.L. (2010). How does diabetes
accelerate Alzheimer disease pathology? Nature Reviews Neurology, 6, 551-9.
Smith, G.E., Pankratz, V.S., Negash, S., Machulda, M.M., Petersen, R.C., Boeve, B.F.,
et al. (2007). A plateau in pre-Alzheimer memory decline: evidence for compensatory
mechanisms? Neurology, 69, 133-9.
Smith, S.J. (2005). EEG in the diagnosis, classification, and management of patients with
epilepsy. Journal of Neurology, Neurosurgery & Psychiatry, 76, ii2-ii7.
Smolina, K., Wotton, C.J., & Goldacre, M.J. (2015). Risk of dementia in patients
hospitalised with type 1 and type 2 diabetes in England, 1998-2011: a retrospective
national record linkage cohort study. Diabetologia, 58, 942-50.
Solomon, E.A., Kragel, J.E., Sperling, M.R., Sharan, A., Worrell, G., Kucewicz, M, et al.
(2017). Widespread theta synchrony and high-frequency desynchronisation underlies
enhanced cognition. Nature Communications, 8, 1704.
Soltész, G., & Acsádi, G. (1989). Association between diabetes, severe hypoglycaemia,
and electroencephalographic abnormalities. Archives in Disease in Childhood, 64, 992-
6.
Sommerfield, A.J., Deary, I.J., & Frier, B.M. (2004). Acute hyperglycaemia alters mood
state and impairs cognitive performance in people with type 2 diabetes. Diabetes Care,
27, 2335-40.
Song, Z., & Routh, V.H. (2006). Recurrent hypoglycaemia reduces the glucose sensitivity
of glucose-inhibited neurons in the ventromedial hypothalamus nucleus. American
Journal of Physiology, Regulatory, Integrative, and Comparative Physiology, 291,
R1283-7.
313
Sörös, P., Whitehead, S., Spence, J.D., & Hachinski, V. (2013). Antihypertensive therapy
treatment can prevent stroke and cognitive decline. Nature Reviews Neurology, 9, 174-8.
Sprague, J.E., & Arbelaez, A.M. (2011). Glucose counterregulatory responses to

hypoglycaemia. Paediatric Endocrinology Reviews, 9, 463-73.
Spieth, W. (1964). Cardiovascular health status, age, and psychological performance.

Journal of Gerontology, 19, 277-84.
Spyridaki, E.C., Avgoustinaki, P.D., & Margioris, A.N. (2016). Obesity, inflammation
and cognition. Current Opinion in Behavioural Sciences, 9, 169-75.
Srikanth, V., Sinclair, A.J., Hill-Briggs, F., Moran, C., & Biessels, G.J. (2020). Type 2
diabetes and cognitive dysfunction – towards effective management of both
comorbidities. The Lancet Diabetes & Endocrinology, 8, 535-45.
Srinivasan, N. (2007). Cognitive neuroscience of creativity: EEG based approaches.

Methods, 42, 109-16.
Starr, J.M., Wardlaw, J., Ferguson, K., MacLullich, A., Deary, I.J., & Marshall, I. (2003).
Increased blood-brain barrier permeability in type II diabetes demonstrated by
gadolinium magnetic resonance imaging. Journal of Neurology, Neurosurgery and
Psychiatry, 74, 70-6.
Starr, J.M., Whalley, L.J., Inch, S., Shering, P.A. (1993). Blood pressure and cognitive
function in healthy old people. Journal of American Geriatric Society, 41, 753-56.
Straaten, E.C., Scheltens, P., Gouw, A., & Stam, C.J. (2014). Eyes-closed task-free
electroencephalography in clinical trials for Alzheimer’s disease: an emerging method
based upon brain dynamics. Alzheimer’s Research & Therapy, 6, 1-10.
Strachan, M.W., Reynolds, R.M., Marioni, R.E., & Price, J.F. (2011). Cognitive function,
dementia, and type 2 diabetes in the elderly. Nature Reviews Endocrinology, 7, 108-14.
314
Stroop, J.R. (1935). Studies of interference in serial verbal reactions. Journal of
Experimental Psychology, 18, 643-62.
Swan, G.E., Carmelli, D., & Larue, A. (1998). Systolic blood pressure tracking over 25
to 30 years and cognitive performance in older adults. Stroke, 29, 2334-40.
Sweeney, M.D., Sagare, A.P., & Zlokovic, B.V. (2018). Blood-brain barrier breakdown
in Alzheimer disease and other neurodegenerative disorders. Nature Reviews Neurology,
14, 133-50.
Sweeney, M.D., Ayyadurai, S., & Zlokovic, B.V. (2016). Pericytes of the neurovascular
unit: key functions and signalling pathways. Nature Reviews Neuroscience, 19, 771-83.
Szendroedi, J., Phielix, E., & Roden, M. (2011). The role of mitochondria in insulin
resistance and type 2 diabetes mellitus. Nature Reviews Endocrinology, 8, 92-103.
Takechi, R., Lam, V., Brook, E., Giles, C., Fimognari, N., Mooranian, A., et al. (2017).
Blood-brain barrier dysfunction precedes cognitive decline and neurodegeneration in
diabetic insulin resistant mouse model: an implication for causal link. Frontiers in aging
Neuroscience, 9, 399.
Tallroth, G., Ryding, E., & Agardh, C.D. (1992). Regional cerebral blood flow in normal
man during insulin-induced hypoglycaemia and in the recovery period following glucose
infusion. Metabolism, 41, 717-21.
Tallroth, G., Lindgren, M., Stenberg, G., Rosen, I., & Agardh, C.D. (1990).
Neurophysiological changes during insulin-induced hypoglycaemia and in the recovery
period following glucose infusion in Type 1 (insulin-dependent) diabetes mellitus and in
normal man. Diabetologia, 33, 319-23.
Tau, G.Z., & Peterson, B.S. (2010). Normal Development of Brain Circuits.
Neuropsychopharamacology, 35, 147-68.
315
Taylor, B.C., Wilt, T.J., & Welch, H.G. (2011). Impact of diastolic and systolic blood
pressure on mortality: implications for the definition of “normal”. Journal of General
Internal Medicine, 26, 685-90.
The Diabetes Control and Complications Trial Research Group. (1993). The effect of
intensive treatment of diabetes on the development and progression of long-term
complications in insulin-dependent diabetes mellitus. The New England Journal of
Medicine, 329, 977-86.
The UK Prospective Diabetes Study (UKPDS) Group. (1998). Intensive blood-glucose

control with sulphonylureas or insulin compared with conventional treatment and risk of
complications with type 2 diabetes. The Lancet, 352, 837-53.
Thomas, M.C., Brownlee, M., Susztak, K., Jandeleit-Dahm, K.A., Zoungas, S., Rossing,
P., et al. (2015). Diabetic kidney disease. Nature Reviews Disease Primer, 1, 1-19.
Tombaugh, T.N., & McIntyre, N.J. (1992). The Mini-Mental State Examination: A
Comprehensive Review. Journal of the American Geriatrics Society, 40, 922-35.
Tomkins, O., Shelef, I., Kaizerman, I., Eliushin, A., Afawi, Z., Misk, A., et al. (2008).
Blood-brain barrier disruption in post-traumatic epilepsy. Journal of Neurology,
Neurosurgery, and Psychiatry, 79, 774-7.
Tomlinson, D.R., & Gardiner, N.J. (2008). Glucose neurotoxicity. Nature Reviews
Tsalikian, E., Becker, D.J., Crumrine, P.K., Daneman, D., & Drash, A.L. (1981).
Electroencephalographic changes in diabetic ketosis in children with newly and
previously diagnosed insulin-dependent diabetes mellitus. The Journal of Paediatrics, 99,
355-9.
Tsivgoulis, G., Alexandrov, A.V., Wadley, V.G., Unverzagt, F.W., Go, R.C., Moy, S., et
al. (2009). Association of higher diastolic blood pressure levels with cognitive
impairment. American Journal of Neurology, 73, 589-95.
316
Uhlhaas, P.J., & Singer, W. (2010). Abnormal neural oscillations and synchrony in
schizophrenia. Nature Reviews Neurology, 11, 100-13.
Unger, T., Borghi, C., Charchar, F., Khan, N.A., Poulter, N.R., Prabhakaran, D., et al.
(2020). 2020 International society of hypertension global hypertension practice
guidelines. Hypertension, 75, 1334-57.
Valdez, P. (2019). Circadian rhythms in attention. Yale Journal of Biology and Medicine,
92, 81-92.
Valdez, P., Ramírez, C., & García, A. (2012). Circadian rhythms in cognitive
performance: implications for neuropsychological assessment. ChronoPhysiology and
Therapy, 2, 81-92.
Van Cauwenberghe, C., Van Broeckhoven, C., & Sleegers, K. (2016). The genetic
landscape of Alzheimer disease: clinical implications and perspectives. Genetics in
Medicine, 18, 421-30.
Van der Elst, W., Van Boxtel, M.P., Van Breukelen, G.J., & Jolles, J. (2006). The Stroop-
Colour Word Test Influence of Age, Sex, and Education; and Normative Data for a Large
Sample Across the Adult Age Range. Assessment, 13, 62-79.
Van der Flier, W., Skoog, I., Schneider, J.A., Pantoni, L., Mok, V., Chen, C.L., et al.
(2018). Vascular cognitive impairment. Nature Reviews Disease Primers, 4, 1-16.
Van der Hiele, K., Vein, A.A., Reijntjes,R.H., Westendorp, R.J., Bollen, E.L., van
Buchem, V.A., et al. (2011). EEG correlates in the spectrum of cognitive decline. Clinical
van Gorp, W., Marcotte, T.D., Sultzer, D., Hinkin, C., Mahler, M., & Cummings, J.L.
(1999). Screening for Dementia: Comparison of Three Commonly Used Instruments.
Journal of Clinical and Experimental Neuropsychology, 21, 29-38.
317
Viggiano, D., Wagner, C.A., Martino, G., Nedergaard, M., Zoccali, C., Unwin, R., et al.
(2020). Mechanisms of cognitive dysfunction in CKD. Nature Reviews Neurology, 16,
452-69.
Vithian, K., & Hurel, S. (2010). Microvascular complications: pathophysiology and

management. Clinical Medicine (London, England), 10, 505-9.
Vlahou, E., Iris-Tatjana Kolassa, F., & Schlee, W. (2014). Resting-state slow wave
power, healthy aging, and cognitive performance. Scientific Reports, 4, 5101.
Waldstein, S.R., Giggey, P.P., Thayer, J.F., Zonderman, A.B. (2005). Nonlinear relations
of blood pressure to cognitive function: the Baltimore Longitudinal Study of Aging.
Hypertension, 45, 374-79.
Waldstein, S.R., Brown, J.R., Maier, K.J., & Katzel, L.I. (2005). Diagnosis of
hypertension and high blood pressure levels negatively affect cognitive function in older
adults. Annals of Behavioural Medicine, 29, 174-80.
Warren, R.E., & Frier, B.M. (2005). Hypoglycaemia and cognitive function. Diabetes,
Obesity, & Metabolism, 7, 493-505.
Weber, M.A., Schiffrin, E.L., White, W.B., Mann, S., Lindholm, L.H., Kenerson, J.G., et
al. (2014). Clinical practice guidelines for the management of hypertension in the
community: a statement by the American Society of Hypertension and the International
society of Hypertension. Journal of Clinical Hypertension, 16, 14-26.
Weinger, K., Jacobson, A.M., Musen, G., Lyoo, I.K., Ryan, C.M., Jimerson, D.C., et al.
(2008). The effects of type 1 diabetes on cerebral white matter. Diabetologia, 51, 417-25.
Weiss, S., & Mueller, H.M. (2012). “Too many betas do not spoil the broth”: The role of
beta brain oscillations in language processing. Frontiers in Psychology, 3, 201.
318
Weisz, N., Schandry, R., Jacobs, A.M., Mialet, J.P., & Duschek, S. (2002). Early
contingent negative variation of the EEG and attention flexibility are reduced in
hypotension. International Journal of Psychophysiology, 45, 253-60.
Wechsler, D. (1955). Manual for the Wechler Adult Intelligence Scale. Psychological
Corp, APA Psyc Net.
Wessels, A.M., Scheltens, P., Barkhof, F., & Heine, R.J. (2008). Hyperglycaemia as a
determinant of cognitive decline in patients with Type 1 diabetes. European Journal of
Pharmacology, 585, 88-96.
Wessels, A.M., Rombouts, S.A., Remijnse, P.L., Boom, Y., Scheltens, P., Barkhof, F., et
al. (2007). Cognitive performance in type 1 diabetes patients is associated with cerebral
white matter volume. Diabetologia, 50, 1763-69.
Wessels, A.M., Simsek, S., Remijnse, P.L., Veltman, D.J., Biessels, G.J., Barkhof, F., et
al. (2006). Voxel-based morphometry demonstrates reduced grey matter density on brain
MRI in patients with diabetic retinopathy. Diabetologia, 49, 2474-80.
Whiteside, D.M., Padula, M.A., Jeffrey, L.K., & Zetterman, R. (1996). Cognitive
Screening with The Neurobehavioral Cognitive Status Examination in a Chronic Liver
Disease Population. The Clinical Neuropsychologist, 10, 459-63.
Wood, J.N. & Grafman, J. (2003). Human Prefrontal Cortex: Processing and
Representational Perspectives. Nature Reviews Neuroscience, 4, 139-47.
World Health Organisation (2020). Raised blood pressure: situation and trends.
https://www.who.int/gho/ncd/risk_factors/blood_pressure_prevalence_text/en/.
Accessed 23/01/2020.
World Alzheimer Report (2015). The global impact of Dementia: an analysis of

prevalence, incidence, cost, and trends.
https://www.alz.co.uk/research/worldalzheimerreport2015summary.pdf. Accessed
05/05/2016.
319
Wright, K.P., Lowry, C.A., & LeBourgeois, M.K. (2012). Circadian and wakefulness-
sleep modulation of cognition in humans. Frontiers in Molecular Neuroscience, 5, 1-12.
Wu, L., He, Y., Jiang, B., Liu, M., Wang, J., Yang, S., et al. (2016). The association
between the prevalence, treatment and control of hypertension and the risk of mild
cognitive impairment in an elderly urban population in China. Hypertension Research,
39, 367-75.
Wyke, B. (1959). Electroencephalographic studies in the syndrome of relative cerebral

hypoglycaemia. Electroencephalography & Clinical Neurophysiology, 11, 602.
Yaffe, K., Vittinghoff, E., Pletcher, M.J., Hoang, T.D., Launer, L.J, Whitmer, R., et al.
(2014). Early adult to midlife cardiovascular risk factors and cognitive function.
Circulation, 129, 1560-67.
Yaffe, K., Falvey, C.M., Hamilton, N., Harris, T.B., Simonsick, E.M., Strotmeyer, E.S.,
et al. (2013). Associations between hypoglycaemia and dementia in a biracial cohort of
older adults with diabetes mellitus. The Journal of the American Medical Association
Internal Medicine, 173, 1300-6.
Yusuf, S., Hawken, S., Ounpuu, S., Dans, T., Avezum, A., Lanas, F., et al. (2004). Effect
of potentially modifiable risk factors associated with myocardial infarction in 52
countries (the INTERHEART study): case-control study. Lancet, 364, 937-52.
Zammitt, N., Warren, R.E., Deary, I.J., & Frier, B.M. (2008). Delayed recovery of
cognitive function following hypoglycaemia in adults with type 1 diabetes: effect of
impaired awareness of hypoglycaemia. Diabetes, 57, 732-6.
Zhang, J., Chen, C., Hua, S., Liao, H., Wang, M., Xiong, Y., et al. (2017). An updated
meta-analysis of cohort studies: diabetes and risk of Alzheimer’s disease. Diabetes
Research &. Clinical Practice, 124, 41-47 (2017).
Zheng, Y., Ley, S.H., & Hu, F.B. (2018). Global aetiology and epidemiology of type 2
diabetes mellitus and its complications. Nature Reviews Endocrinology, 14, 88-98.
320
Zinman, B., Wanner, C., Lachin, J.M., Fitchett, D., Bluhmki, E., Hantel, S., et al. (2015).
Empagliflozin, cardiovascular outcomes, and mortality in type 2 diabetes. The New
Zoefel, B., Huster, R.J., & Herrmann, C.S. (2011). Neurofeedback training of the upper
alpha frequency band in EEG improves cognitive performance. Neuroimage, 54, 1427-
31.
321
ProQuest Number: 30611943
INFORMATION TO ALL USERS

The quality and completeness of this reproduction is dependent on the quality
and completeness of the copy made available to ProQuest.
Distributed by ProQuest LLC ( 2023 ).

Copyright of the Dissertation is held by the Author unless otherwise noted.
This work may be used in accordance with the terms of the Creative Commons license
or other rights statement, as indicated in the copyright statement or in the metadata
associated with this work. Unless otherwise specified in the copyright statement
or the metadata, all rights are reserved by the copyright holder.
This work is protected against unauthorized copying under Title 17,

United States Code and other applicable copyright laws.
Microform Edition where available © ProQuest LLC. No reproduction or digitization

of the Microform Edition is authorized without permission of ProQuest LLC.
ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346 USA

Cog Ad Thesis

Uploaded by

Copyright:

Available Formats

Cog Ad Thesis

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cog Ad Thesis

Uploaded by

Copyright:

Available Formats

Investigating Cognitive Function in Clinical and

Non-clinical Samples using Electroencephalography

Principal supervisor: Associate Professor Sara Lal (UTS)

Co-supervisors: Dr Najah Nassif (UTS)

This research is supported by the Australian Government Research Training Program.

Published Abstracts/Conference Presentations:

Figure 1.1 – Life expectancy by birth and sex (1886-2016) ................................................... 1

Figure 1.2 – Estimated number of Australians categorised by age group in 2018.................. 3

group (2016-2096) .............................................................................................. 3

Figure 1.4 – Lateral view of the cerebral cortex ..................................................................... 6

Figure 1.5 – Diagnostic criteria used to diagnose mild cognitive impairment........................ 8

Figure 1.6 – Hypothetical trajectories in normal brain ageing .............................................. 10

Figure 1.7 – Projected prevalence of dementia in Australians (2016-2056) ......................... 12

Figure 1.8 – Diagnostic criteria used to diagnose major neurocognitive disorder/dementia..14

Figure 1.12 – Patterns in the accumulation of neuropathologies in Alzheimer’s disease ....... 19

Figure 1.15 – Homeostatic regulation of blood glucose concentration ................................... 25

Figure 1.18 – Trends in leading cause of disease in Australia ................................................ 32

Figure 1.22 – The blood-brain barrier ..................................................................................... 49

Figure 1.23 – Risk factors linked to the development of hypertension ................................... 52

Figure 1.25 – A neurovascular unit ......................................................................................... 61

Figure 1.26 – Cerebral arteries damaged by hypertension ...................................................... 63

Figure 1.27 – The International 10-20 system of electroencephalography ............................. 65

Figure 2.1 – Non-invasive digital blood pressure monitor .................................................... 89

Figure 2.5 – 32-channel, non-invasive elastic EEG cap........................................................ 97

Figure 2.7 – Electrode gel, sterile syringe, and blunted needle............................................. 99

Figure 2.8 – Unprocessed EEG tracing ............................................................................... 100

Figure 2.10 – The Mini-Mental State Examination (MMSE) ............................................... 104

Figure 2.11 – Stimulus sheet (Part 2) of the MMSE ............................................................. 105

Figure 2.12 – Participant completing Construction sub-test of the Cognistat....................... 107

Figure 2.13 – The Cognistat cognitive status profile ............................................................ 108

Figure 2.14 – Protocol for the present study ......................................................................... 112

performance in the Cognistat for Type 1 diabetes mellitus ............................ 140

performance in the Cognistat for Type 2 diabetes mellitus ............................ 142

performance in the Cognistat for Type 2 diabetes mellitus ............................ 143

Cognistat score for Type 2 diabetes mellitus .................................................. 144

domain of the Cognistat for Type 2 diabetes mellitus .................................... 145

domain of the Cognistat for Type 2 diabetes mellitus .................................... 146

the same individual ......................................................................................... 187

active phase for Type 2 diabetes mellitus ....................................................... 217

the active phase for Type 2 diabetes mellitus ................................................. 228

at PZ during the baseline phase for hypertension ............................................ 230

at TP7 during the baseline phase for hypertension .......................................... 235

diabetes mellitus.............................................................................................. 262

Table 1.2 – Brain waves commonly observed in an electroencephalogram and their

corresponding psychophysiological state(s) ..................................................... 68

or each cognitive domain of the Cognistat ..................................................... 111

Table 3.1 – Key demographic characteristics for all groups.............................................. 121

occurred for each group .................................................................................. 124

well as total Cognistat score ........................................................................... 131

mismatched stimuli of the Stroop Colour Word Test ..................................... 134

Colour Word Test ........................................................................................... 135

for the non-clinical group................................................................................ 161

for the non-clinical group................................................................................ 165

for the non-clinical group................................................................................ 169

for the non-clinical group................................................................................ 171

phase for the non-clinical group ..................................................................... 173

for the non-clinical group................................................................................ 174

for the non-clinical group................................................................................ 176

for the non-clinical group................................................................................ 178

for the non-clinical group................................................................................ 180