J Rehabil Med 2016; 48: 48–55

ORIGINAL REPORT

COMPARISON OF CRANIO-CERVICAL FLEXION TRAINING VERSUS

CERVICAL PROPRIOCEPTION TRAINING IN PATIENTS WITH CHRONIC
NECK PAIN: A RANDOMIZED CONTROLLED CLINICAL TRIAL

Tomás Gallego Izquierdo, PT, PhD1, Daniel Pecos-Martin, PT, PhD1, Enrique Lluch Girbés,
PT2,3,4,5, Gustavo Plaza-Manzano, PT, PhD6, Ricardo Rodríguez Caldentey, PT1, Rodrigo Mayor
Melús, PT1, Diego Blanco Mariscal, PT1 and Deborah Falla, PT, PhD7,8
From the 1Department of Physical Therapy, University of Alcalá, Alcalá de Henares, Madrid, 2Department of
Physical Therapy, Faculty of Physiotherapy, University of Valencia, Spain, 3Departments of Human Physiology and
Physiotherapy, Faculty of Physical Education & Physiotherapy, Vrije Universiteit Brussel, 4Department of Physical
Medicine and Physiotherapy, University Hospital Brussels, Belgium, 5Pain in Motion Research Group,
www.paininmotion.be, 6Department of Medicine, University Complutense of Madrid, Madrid, Spain, and 7Pain Clinic,
Center for Anesthesiology, Emergency and Intensive Care Medicine, University Hospital Göttingen, Göttingen,
and 8Institute for Neurorehabilitation Systems, Bernstein Center for Computational Neuroscience, University Medical
Center Göttingen, Georg-August University, Göttingen, Germany

Objective: To compare the effects of cranio-cervical flexion INTRODUCTION

vs cervical proprioception training on neuromuscular con-
trol, pressure pain sensitivity and perceived pain and dis- A myriad of sensorimotor disturbances have been identified in
ability in patients with chronic neck pain. people with chronic neck pain (1, 2). Consequently, a pragmatic
Methods: Twenty-eight volunteers with chronic non-specific multimodal intervention approach has been recommended to
neck pain were randomly assigned to 1 of 2 interventions address changes in sensorimotor control (2, 3), since no ap-
and undertook 6 physiotherapist-supervised sessions over proach in isolation has been shown to address all aspects of
a period of 2 months. Both groups performed daily home sensorimotor function. As an example, impaired proprioception
exercise. Performance on the cranio-cervical flexion test, and postural control is a feature of some patients with neck
pressure pain thresholds and reported levels of pain and dis- pain, and several treatment approaches have been proposed to
ability were measured before and immediately after the first directly address these impairments, such as head-neck aware-
treatment session, 1 month after starting treatment and 2 ness exercises (i.e. training of joint position sense with head
months after starting treatment (at completion of the inter- relocation), oculomotor exercises (i.e. gaze stability, eye-neck
vention). coordination) or balance training (4). These programmes lead
Results: At 2 months, both groups improved their perfor- to enhanced sensorimotor control and reduced neck pain and
mance on the cranio-cervical flexion test (p < 0.05), but this disability (5–7). However, it is not known if this training ap-
did not differ between groups (p > 0.05). Both groups showed proach also has an effect on other aspects of neuromuscular
a reduction in their pain at rest and disability at 2 months,
function in people with neck pain. For instance, there is
but this was also not different between groups (p > 0.05).
evidence for a high prevalence of increased activation of the
Pressure pain sensitivity did not change for either group.
sternocleidomastoid muscle in people with neck pain (8, 9)
Conclusion: Both specific cranio-cervical flexion training
and reduced activation of deeper cervical flexor muscles has
and proprioception training had a comparable effect on
performance on the cranio-cervical flexion test, a test of the commonly been associated with this augmented activity (8,
neuromuscular control of the deep cervical flexors. These re- 10). Altered coordination between the deep and superficial
sults indicate that proprioception training may have positive flexor muscles is reflected as poor performance on the cranio-
effects on the function of the deep cervical flexors. cervical flexion test (CCFT), a clinical test used to evaluate
neuromuscular control of the deep cervical flexors (11).
Key words: neck pain; proprioception; exercise; intervention.
Typically specific proprioceptive training regimes are de-
J Rehab Med 2016; 48: 48–55 signed to target the deep muscles, especially the suboccipital
Correspondence address: Deborah Falla, Pain Clinic, Center muscles (7), since they have the highest density of receptors
for Anesthesiology, Emergency and Intensive Care Medicine (12) and are known to have a specific role in reflex and cen-
University Hospital Göttingen, Robert-Koch-Str. 40, DE- tral connections to the vestibular, visual and postural control
37075 Göttingen, Germany. E-mail: deborah.falla@bccn. systems (13). It is not known whether this training approach
uni-goettingen.de has a beneficial effect on the control of other muscles, such
Accepted Aug 27, 2015; Epub ahead of print Nov 26, 2015 as the coordination between the deep and superficial cervical

J Rehabil Med 48 © 2016 The Authors. doi: 10.2340/16501977-2034

Journal Compilation © 2016 Foundation of Rehabilitation Information. ISSN 1650-1977
 Evaluation of specific training for neck pain 49

flexors. Especially considering that the deep cervical flexor the cervical spine observed during prescribed active movements of
muscles (e.g. longus colli) also have a relatively high density the neck and/or upper limb (25).
Subjects were excluded if they had vascular, neoplastic or vestibular
of muscle spindles (12).
disease, a diagnosis of fibromyalgia or rheumatoid arthritis, or any
One form of exercise that has been advocated for address- medical condition that prevented exercise.
ing impaired neuromuscular control of the cervical flexor The study was approved by the ethics committee of the University of
muscles is cranio-cervical flexion (CCF) training, which aims Alcalá, Spain and procedures were conducted according to the Declara-
to enhance activation of the deep cervical flexor muscles and tion of Helsinki. The study was registered in Clinical Trials database
(NCT 022258730). Patients were informed orally and in writing about the
restore coordination between the deep and superficial cervi-
procedures. Written informed consent was provided before participation.
cal flexors (14). Clinical trials examining the effectiveness of
this exercise regime have demonstrated positive outcomes in Interventions
terms of reduction in neck pain and disability, improvement This study was a randomized controlled clinical trial, with parallel
in sitting posture and enhanced neuromuscular control of the groups and a blinded assessor. It was guided according to the CON-
cervical flexor muscles (7, 14–21). Interestingly, CCF training SORT Statement. Patients were randomized into 2 exercise groups:
also improved proprioceptive acuity of the neck (7), indicating CCF training and proprioception training (Fig. 1). Randomization was
performed using computer-generated random numbers (Epidat® ver-
that proprioception can be enhanced with specific exercise. sion 3.1). Concealment of allocation was ensured using sequentially
The primary aim of this study was to compare the effects of numbered opaque, sealed envelopes. Patients in each group received 6
conventional cervical proprioception training vs CCF training physiotherapist-supervised sessions distributed within a 2 month-period
on cervical flexor muscle function, measured through perfor- by the same physical therapist with 5 years of experience. The first ses-
sion was delivered immediately after the baseline assessment, the second
mance on the CCFT, in patients with chronic non-specific
and third sessions were 1 and 2 weeks after the baseline assessment,
neck pain, in order to evaluate whether proprioception training respectively, and the following 3 sessions were given once every 2 weeks
can achieve a beneficial effect on neuromuscular control of for the remainder of the training duration (i.e. 4th, 6th and 8th week).
the cervical flexor muscles. Besides effects on neck muscle Exercise sessions lasted a maximum of 45 min. All the subjects
function, the impact of the 2 exercise programmes on patient received an exercise diary and were requested to practice twice per
day for the 2 month-period of the study. Participants were instructed
self-reports of pain and disability and neck muscle hyperalge-
to perform their exercise in the same way that they were instructed
sia was explored. In addition to evaluating the effects after 2 during their supervised sessions. The exercise session at home took no
months of training, we were interested in understanding how longer than 20 min per day and no provocation of pain was permitted.
rapidly improvements can be seen. Thus, we included meas- Participants were advised not to receive any other specific treatments
ures immediately after the first session of exercise, after the for neck pain, although their usual medication was not withdrawn.
Performance on the CCFT, pressure pain threshold (PPT) over
first month of exercise and after 2 months when the exercise cervical muscle sites and levels of pain and disability, were measured
programme was completed. This was based on earlier work before and immediately after the first treatment session, 1 month after
showing immediate positive effects of specific neck training starting treatment and 2 months after starting treatment, when the
on performance on the CCFT (22). intervention was completed.

METHODS Assessed for eligibility (n=34)

Participants
Twenty-eight subjects (18 women) with a history of Excluded (n=6)
Enrollment

chronic non-specific neck pain, greater than 3 months’ >15/50 NDI (n=5)
Fibromyalgia (n=1)
duration, participated in the study. The sample size was
calculated using G*Power 3.0.18 Software (23) based
on the CCFT score as the primary outcome measure. Randomized (n=28)
The effect size for the CCFT was determined as 0.25
and the correlation between repeated measurements as-
sumed at 0.5. Assuming 4 measurements (pre-session,
post-session 1, 1 month after starting treatment, and 2
Allocation

months after starting treatment (i.e. after completion of Allocated to CCF exercise (n=14) Allocated to propiocepcion exercise
treatment)) in the 2 groups, the correction of spheric- Received allocated intervention (n=14)
(n=14) Received allocated intervention
ity was determined as 0.5. With a statistical power of
(n=14)
0.80 and an alpha level of 0.05, a total sample size of
24 patients was estimated. An allowance was made for
Follow-up

a 15% drop-out rate, increasing the sample size to 28

Lost to follow-up (n=0) Lost to follow-up (n=0)
patients (14 per group). Discontinued intervention (n=0) Discontinued intervention (n=0)
Participants were recruited via advertisements during
the period January to June 2014. Inclusion criteria were:
age between 18 and 55 years, score ≤ 15/50 on the Neck
Analysis

Disability Index (NDI) (24), showing signs of cervi-

cal movement control dysfunction (25), and manual Analysed (n=14) Analysed (n=14)
physical examination revealing muscle tenderness. A
cervical movement control dysfunction was defined as
the presence of aberrant or uncontrolled movements of Fig. 1. Participant flow and retention. CCF: cranio-cervical flexion.

J Rehabil Med 48
50 T. Gallego Izquierdo et al.

Cranio-cervical flexion training. Low load training of the cranio- 5 s using the correct CCF action. Secondly, the assessor evaluated the
cervical flexor muscles followed the protocol described by Jull et al. number of repetitions of a 10 s hold that the patient could perform at
(3). This exercise specifically targets the deep flexor muscles (longus their target level. Performance on the CCFT was scored as the pressure
capitis and longus colli), whilst aiming to minimize the activation of the level that the patient was able to achieve (activation score) multiplied
superficial flexor muscles (sternocleidomastoid and anterior scalene). by the number of repetitions they could perform. For example, if a
Initially, patients were taught to perform the CCF movement slowly patient could achieve the second level of the test (24 mmHg) and per-
and in a controlled manner in a supine position, with the head and form 6 10-s holds with the correct action of CCF, then the performance
neck in a neutral position. Once the correct CCF motion was achieved, index was 4 × 6 = 24. The performance index ranged from 0 to 100.
subjects began to hold progressively increasing ranges of CCF using
feedback from an air-filled pressure sensor (Stabilizer™, Chattanooga Pressure pain thresholds. PPT were recorded bilaterally over the leva-
Group Inc., Tennessee, USA) placed behind the neck. The patient tor scapulae, upper trapezius, splenius capitis, sternocleidomastoid
initially performed CCF to sequentially reach 5 pressure targets in 2 and anterior scalene muscles at the points of greatest hyperalgesia.
mmHg increments from a baseline of 20 mmHg to the final level of 30 The hyperalgesic points were considered for assessment if a nodule
mmHg. The physiotherapist identified the target level that the patient within a taut band of skeletal muscle was palpated, pressure over the
could hold steadily for 5 s without resorting to retraction, without muscle fully or partially reproduced the usual pain experienced by the
dominant use of the superficial neck flexor muscles, and without a subject and full muscle stretching was restricted (27).
quick, jerky cranio-cervical flexion movement. Training commenced To ensure repeatability of location for the subsequent assessment
at this target level. For each target level, the contraction duration was sessions, the distance to the hyperalgesic points from the spinous
increased to 10 s, and the subject trained to perform 10 repetitions with process of C7 (for the upper trapezius), the mastoid process (for the
brief rest periods between each contraction (~3–5 s). Once one set of sternocleidomastoid, splenius capitis and levator scapulae) and the
10 repetitions of 10 s was achieved at one target level, the exercise middle third of the clavicle (for the anterior scalene), was measured.
was progressed to train at the next target level up to the final target of Testing sites were also marked and photographed as this has been
10 repetitions of 10 s at 30 mmHg. shown to improve the reliability of PPT measurements (28).
The exercise load prescribed to each patient was based on their as- The PPT was measured using an analogue algometer (Force Dial
sessment performance. Patients were not given a pressure biofeedback model FDK 20, Wagner Instruments, Connecticut, USA), with a surface
unit for home training, but they were taught in the clinic to replicate area at the round tip of 1cm2. The algometer probe tip was applied
their training pressure level by performing several repetitions to reach perpendicularly to the skin at a rate of 1kg/cm2/s. PPT was measured
the target without visual feedback from the dial. The training load of at each site with a 30 s rest period between each measurement. A
each patient was adjusted at each session. familiarization phase preceded the formal measurements where par-
ticipants were instructed on the procedure and the examiner practiced
Proprioceptive training. Patients trained cervical proprioception with them at a remote site (forearm). Subjects were instructed to in-
following the protocol described by Revel et al. (5, 26). This regime dicate the moment when pressure changed to pain, which correspond
consisted of exercises of head relocation, eye-follow, gaze stability to the definition of the PPT. The same researcher performed the PPT
and eye-head coordination. For head relocation exercises, subjects measurements in all subjects and was blinded to purpose of the study.
started in a sitting position, with a laser attached to a helmet at the Four consecutive PPT measures were performed at each location
apex of their head, and a target located at eye level on a wall 90 cm with 30 s of rest between measurements. The first PPT measure was
away. This was established as the natural head posture. Subjects then discarded and the mean of the subsequent 3 PPT measures was used
practiced relocating their head to the natural head posture after active for further analysis.
neck movements, first with eyes open using feedback from the laser Pressure algometry is a valid and reliable method to measure PPT
attached to their head, then with pupillary glasses preventing pupillary (29), with studies showing good repeatability of measurements on the
excursion, and finally with their eyes closed. All active movements of neck muscles (intracorrelation coefficient 0.78–0.93) (30).
the cervical spine (flexion, extension, rotation, lateral flexion) were
performed. Measures of pain and disability. Neck pain intensity was measured on
Oculomotor exercises were progressed through several stages. First, a 10-cm visual analogue scale (VAS), anchored with “no pain” and “the
eye movement following a target located at a comfortable distance worst pain imaginable”. Patients were requested on each assessment
was practiced with the head stationary, progressing to movements of to complete a triple VAS: their worst or maximal pain experienced
the head with visual fixation on a target (i.e. gaze stability). Pupillary during the last 2 weeks (maximum VAS), their least or minimum pain
glasses were used in the clinic to ensure a steady gaze during this experienced during the last 2 weeks (minimum VAS), and current pain
exercise. Eye-head coordination exercises started with rotation of the at rest (current VAS). The Spanish version of the Neck Disability Index
eyes and head to the same side, both left and right. After that, patients (NDI) questionnaire was used to assess perceived pain and disability
practiced following a target with the eyes first, followed by the head, related to neck pain (31, 32).
ensuring that they maintained focus on the target. As a further progres-
sion, the eyes moved first, and then the head, to look between 2 targets Data analysis
positioned horizontally or vertically, and finally, the eyes and head Data were analysed using the statistical package SPSS 21.00 for Win-
rotated in opposite directions, both left and right. All these exercises dows. Statistical significance was set at p < 0.05. In order to analyse
were progressed by increasing the speed and range of motion of the the effectiveness of the 2 interventions, an intention-to-treat analysis
target and with patients in a standing position. was performed. The existence of differences between groups in the
dependent variables was determined. Four measurements were per-
Outcome measures formed: baseline (pre), immediately after the first treatment session
The main outcome measure for this study was performance on the (post session 1), 1 month after commencement of treatment (post
CCFT. The secondary outcome measures were patient-reported lev- month 1), and 2 months after commencement of treatment when the
els of pain and disability and PPT. The assessor, who was blinded to intervention was completed (post month 2). For comparison between
subject group for the outcome assessments, was a physical therapist groups, a variable defined as “difference” was calculated for each
with 3 years of clinical experience in outpatient orthopaedic practice dependent variable, by subtracting pre-test and post-test.
with a Master’s degree in physical therapy. Prior to statistical comparisons, all data were tested for normal dis-
tribution by the Shapiro-Wilk test. Then, a descriptive analysis of the
Performance on the cranio-cervical flexion test. The assessor identified data was obtained for the dependent variables in the pre, post session
the target level (22–30 mmHg) that the subject could hold steadily for 1, post month 1 and post month 2 measurements, and for the variable

J Rehabil Med 48
 Evaluation of specific training for neck pain 51

Table I. Baseline characteristics of the variables for the cranio-cervical flexion and proprioceptive training groups
CCF training Proprioception training
(n = 14) (n = 14) p-value
Age, years, mean (SD) 28.43 (6.16) 29.93 (7.34) 0.56
Weight, kg, mean (SD) 65.87 (12.80) 62.50 (9.51) 0.43
Height, cm, mean (SD) 167 (0.08) 168 (0.60) 0.73
BMI, kg/m2, mean (SD) 23.68 (3.97) 22.04 (0.76) 0.22
CCFT, median (IQR) 10 (2.0, 25.0) 11 (2.0, 20.8) 0.35
NDI (0–50) mean (SD) 7.71 (2.78) 7.42 (2.87) 0.79
VAS at rest (0–10 cm), median (IQR) 3.45 (2.95, 4.35) 4.00 (2.66, 4.62) 0.48
VAS maximum (0–10 cm), median (IQR) 5.90 (4.40, 7.00) 5.00 (3.26, 6.40) 0.22
VAS minimum (0–10 cm), median (IQR) 0.22 (0.00, 0.77) 1.00 (0.00, 2.10) 0.12
IQR: interquartile range; SD: standard deviation; CCF: cranio-cervical flexion; BMI: body mass index; CCFT: cranio-cervical flexion test; NDI: Neck
Disability Index; VAS: visual analogue scale.

“difference”. In those analyses, mean and standard deviation (SD) they failed to fulfil the necessary criteria for inclusion. A final
were shown for dependent variables normally distributed. For those total of 28 (18 women, 10 men) participants were included.
variables that did not conform to normal, the data were expressed as
Participants were randomized into 2 groups: CCF training
median and first and third quartiles.
Subsequently, homogeneity of the 2 intervention groups was studied (n = 14) and proprioception training (n = 14). There were no
using the Student’s t-test for independent samples. For values that did drop-outs during the study period. All participants completed
not conform to normal, the Mann-Whitney U test was applied. For the the 6 physiotherapist-supervised sessions and verbally con-
variable sex, homogeneity was studied through Pearson’s χ2 test or the firmed that they complied with the home-exercise programme.
Fisher’s exact test, in case the former could not be used.
To determine whether there were differences in each group, con-
Baseline characteristics of the CCF and proprioception train-
sidering each group in isolation, between the 4 assessments (pre, post ing groups are presented in Table I. There were no significant
session 1, post month 1 and post month 2) in each of the variables, differences in baseline variables between groups (all p > 0.05).
repeated measures analysis of variance (ANOVA) was used. ANOVA
was complemented with Simple and Helmert contrasts in those vari-
ables where the 4 assessments presented a data distribution similar to
Cranio-cervical flexion test performance
normal. Friedman ANOVA together with multiple comparisons tests Fig. 2 shows the performance index on the CCFT for each group
were used when variables were not normally distributed. at each time-point. For the CCF training group, significant
To determine whether there were differences in the dependent
differences were observed between post month 2 and baseline
variables between the 2 intervention groups, the Student’s t-test for
independent samples was used. In those following normal, the effect (p < 0.01). The proprioception training group showed significant
size was calculated according to the formula d = 2t/√g. For the variables differences in the performance index on the CCFT at both post
that did not conform to normal, this analysis was performed using month 1 and post month 2 compared with baseline (p < 0.01).
the Mann-Whitney U test and the estimated effect size according to
Grissom’s procedures.
Pressure pain thresholds
Fig. 3 presents the PPT for each group at each time-point. Both
RESULTS
groups showed a significant increase in the PPT over the right
The study flow diagram is shown in Fig. 1. A total of 34 patients anterior scalene muscle from baseline to post month 2. In addi-
were screened for eligibility. Six participants were excluded as tion, the proprioceptive training group showed an increase in

A C-CF Training B Proprioception Training

40 40
Performance Index (0-100)

Performance Index (0-100)

30 30

20 20

10 10

0 0
Pre Session 1 Month 1 Month 2 Pre Session 1 Month 1 Month 2

Fig. 2. Performance index on the cranio-cervical flexion (CCF) test for CCF and proprioception training groups at each measurement time-point.

J Rehabil Med 48
52 T. Gallego Izquierdo et al.

A Proprioception Training Pain and disability

2,00 Table II presents the pain and disability scores
across sessions and Table III presents the within-
group differences obtained with each variable at
Pressure Pain Threshold (Kg/cm2)

1,60
each measurement time. Both groups showed a
significant reduction in NDI score and current
1,20 VAS immediately after the first treatment session
(post session 1) (all p < 0.05).
One month after starting treatment, both groups
0,80
showed a significant reduction in current VAS
compared with baseline (both p < 0.001). In addi-
0,40 tion, the proprioception training group showed a
significant reduction in NDI score (p < 0.01) and
in maximum VAS (p < 0.05) between these time-
0,00
R UT L UT R SCap L SCap R LS L LS R SCM L SCM R AS L AS points. When evaluating the change from post
session 1 to post month 1, only the propriocep-
B C-CF Training Pre
tion training group showed a significant change
2,00 Post Session 1
Post Month 1 in NDI score (p < 0.05), current VAS (p < 0.001)
Post Month 2 and maximum VAS (p < 0.05).
1,60 Between post session 1 and post month 2 both
Pressure Pain Threshold (Kg/cm2)

groups showed a significant reduction in NDI

1,20
score (both p < 0.01), current VAS (both p < 0.001)
and maximum VAS (both p < 0.001). Moreover,
the proprioception training group showed a signif-
0,80 icant change in minimum VAS (p < 0.05) between
these time-points. Between post month 1 and post
0,40
month 2, only the CCF training group showed a
significant change in NDI score (p < 0.001) and
in current VAS (p < 0.001). Both groups showed
0,00 a significant reduction in maximum VAS score
R UT L UT R SCap L SCap R LS L LS R SCM L SCM R AS L AS
(both p < 0.001) between these time-points.
Fig. 3. Pressure pain thresholds for cranio-cervical flexion (CCF) and proprioception Two months after initiating the intervention,
training groups at each measurement time-point. R: right; L: left; AS: anterior scalene; LS: when treatment was completed, both groups
elevator scapulae; SCap: splenius capitis; SCM: sternocleidomastoid; UT: upper trapezius.
showed a significant reduction in NDI score
(both p < 0.001), current VAS (both p < 0.001)
the PPT over the right upper trapezius (p < 0.05), right splenius and maximum VAS (both p < 0.001) compared with base­line.
(p < 0.05) and right elevator scapulae (p < 0.05) at post month Moreover, the proprioception training group showed a sig-
2 compared with baseline. No within-group differences were nificant change in minimum VAS (p < 0.05) between these
found for the rest of the variables. time-points.

Table II. Pain and disability scores obtained at each measurement session (n = 14)
Group Pre Post session 1 Post month 1 Post month 2
NDI, mean (SD)
CCF 7.71 (2.78) 6.64 (2.61) 7.60 (3.19) 4.46 (2.02)
Proprioception 7.42 (2.87) 6.21 (2.35) 4.78 (2.60) 4.14 (2.62)
VAS maximum, median (IQR)
CCF 5.90 (4.40, 7.00) 2.75 (1.00, 4.00) 2.25 (1.40, 3.17) 0.20 (0.00, 1.01)
Proprioception 5.00 (3.26, 6.40) 3.45 (2.30, 4.30) 1.70 (0.95, 2.30) 1.25 (0.00, 2.02)
VAS minimum, median (IQR)
CCF 0.22 (0.00, 0.77) 6.10(4.43,7.00) 4.60 (3.43, 6.35) 2.17 (1.00, 340)
Proprioception 1.00 (0.00, 2.10) 4.20 (3.97, 5.77) 3.45 (1.97, 4.67) 2.05 (0.65, 3.10)
VAS rest, median (IQR)
CCF 3.45 (2.95, 4.35) 0.30 (0.00, 0.70) 0.35 (0.00, 0.50) 0.00 (0.00, 0.26)
Proprioception 4.00 (2.66, 4.62) 1.10 (0.00, 2.92) 0.55 (0.00; 0.92) 0.00 (0.00, 0.70)
IQR: interquartile range; SD: standard deviation; CCF: cranio-cervical flexion; NDI: Neck Disability Index; VAS: visual analogue scale.

J Rehabil Med 48
 Evaluation of specific training for neck pain 53

Table III. Results of data analysis and within-group differences obtained with each variable at each measurement time
Pre – post Pre – post Pre – post Post session Post session Post month
Group p-value session 1 month 1 month 2 1-post month 1 1-post month 2 1-post month 2
CCFT
CCF 0.002a 0.986 0.112 0.002 > 0.999 0.169 > 0.999
Proprioception < 0.001a > 0.999 0.005 < 0.001 0.094 0.08 > 0.999
VAS minimum
CCF 0.359a – – – – – –
Proprioception < 0.001a > 0.999 0.746 < 0.05 0.342 0.016 > 0.999
VAS rest
CCF < 0.001a 0.746 0.50 < 0.001 > 0.999 0.02 0.62
Proprioception < 0.001a > 0.999 < 0.001 < 0.001 0.41 < 0.001 > 0.999
VAS maximum
CCF < 0.001a > 0.999 > 0.999 < 0.001 > 0.999 < 0.001 0.013
Proprioception < 0.001a > 0.999 0.94 < 0.001 0.32 < 0.001 0.554
NDI
CCF 0.020b 0.006 0.747 < 0.001 0.418 0.003 < 0.001
Proprioception < 0.001b 0.022 0.01 < 0.001 0.002 < 0.001 0.082
a
Friedman test; banalysis of variance.
–: In those measurements for which the Friedman analysis of variance (ANOVA) test showed no significant differences, tests for multiple comparisons
were not performed; IQR: interquartile range; SD: standard deviation; CCF: cranio-cervical flexion; CCFT: cranio-cervical flexion test; NDI: Neck
Disability Index; VAS: visual analogue scale.

Between-group differences (d = 0.44). Moderate effect sizes were observed for the right
Differences between the 2 interventions at the end of treatment sternocleidomastoid (d = 0.37), right splenius capitis (d = 0.38)
were studied using the Student’s t-test and Mann-Whitney U and right upper trapezius (d = 0.38) PPTs, but no between-group
test, in addition to the effect size (Table IV). The comparison differences were observed.
between the CCF and the proprioceptive training groups at
the end of treatment (post month 2) did not show significant
DISCUSSION
differences for the score on the CCFT (Fig. 2). The effect size
for the CCFT was small (d = 0.05) at the end of both interven- This study is the first to evaluate whether proprioception train-
tions. There was no significant difference between groups ing is capable of inducing improvements in neck muscle func-
for any of the outcome measures except for the change in the tion, in the same manner that CCF exercise has demonstrated a
minimum VAS, which was significantly different favouring significant benefit on impaired proprioception (7). Our results
the CCF training group (p < 0.05) with a moderate effect size showed that cervical proprioception training not only improves

Table IV. Between-group variable differences between baseline (pre) and assessment at 2 months after starting treatment. Results are expressed as
mean and standard deviation (SD) for normally distributed variables and median and first and third quartiles for non-normal variables
Effect size CCF training Proprioception training
Variables Student t-test Cohen’s d Mean (SD) Mean (SD)
PPT right SCM 0.347 0.37 –0.013 (0.161) 0.060 (0.237)
PPT right AS 0.781 0.11 –0.033 (0.134) –0.016 (0.172)
PPT left LE 0.538 0.23 0.044 (0.290) 0.140 (0.494)
PPT left SC 0.715 0.13 –0.028 (0.299) 0.018 (0.353)
PPT right SC 0.313 0.38 –0.012 (0.239) –0.145 (0.420)
NDI 0.964 0.01 3.250 (2.026) 3.285 (2.163)
Mann-Whitney U test Grissom Median (IQR) Median (IQR)
CCFT 0.804 0.05 –20 (–28, –3.75) –21.25 (–29, –5)
PPT right UT 0.085 0.38 0.120 (–0.247, 0.312) 0.405 (–0.015, 0.792)
PPT left SCM 0.635 0.10 0.000 (–0.155, 0.107) 0.000 (–0.125, 0.185)
PPT left AS 0.603 0.12 –0.040 (–0.147, 0.240) –0.035 (–0.221, 0.007)
PPT right LE 0.701 0.24 0.130 (–0.272, 0.390) 0.000 (–0.186, 0.257)
PPT left UT 0.306 0.23 0.100 (0.000, 0.270) 0.015 (–0.177, 0.275)
VAS minimum 0.044 0.44 0.000 (–0.025, 0.500) 0.550 (0.000, 1,075)
VAS rest 0.769 0.08 2.92 (2.37, 3.62) 2.87 (1.57, 4.12)
VAS maximum 0.810 0.04 3.40 (2.22, 4.05) 3.05 (1.35, 4.52)
IQR: interquartile range; PPT: pressure pain threshold; AS: anterior scalene; LE: elevator scapulae; SC: splenius capitis; SCM: sternocleidomastoid;
UT: upper trapezius; SD: standard deviation; CCF: cranio-cervical flexion; CCFT: cranio-cervical flexion test; NDI: Neck Disability Index; VAS:
visual analogue scale.

J Rehabil Med 48
54 T. Gallego Izquierdo et al.

patient perceived pain and disability, but also has an effect mobilizations (33, 34), although several others have shown
on other aspects of neuromuscular function, specifically the no effect (e.g. 15, 22).
coordination between the deep and superficial cervical flexors Both interventions resulted in a reduction in pain and dis-
measured through performance on the CCFT. Perhaps this is ability, which confirms the pain-modulation properties of
not surprising, considering that the proprioception exercises active neck exercises (35) and highlights the importance of
required fine control of neck movement. In addition, a signifi- exercise as a component of treatment for the management of
cant improvement in neck flexor function, as measured via the patients with chronic neck pain. However, the change in the
CCFT, was noted with proprioception training prior to CCF subjective report of pain and disability was not accompanied
training (1 month vs 2 months after starting treatment). At 2 by a consistent objective improvement on PPTs evaluated with
months, when treatment was completed, the improvements in algometry. This observation is in line with a finding that 6
neck flexor function were comparable between groups. weeks of CCF training did not change pressure pain sensitivity
Patients with neck pain can present with a myriad of senso- over myofascial trigger points located in superficial cervical
rimotor impairments and 1 mode of exercise may not address muscles (20). In the current study, we measured PPT not only
all potential motor impairments adequately (21). Moreover, a in the upper trapezius, elevator scapulae and splenius capitis
single exercise is unlikely to affect both the functional as well as in Lluch et al. (20), but also extended PPT measurements to
as structural changes that may be present in the neck muscles in include the sternocleidomastoid and anterior scalene, which are
people with neck pain (1). Exercise-induced changes in motor often excessively activated in people with chronic neck pain
performance are usually specific to the mode of the exercise (1). However, changes in pressure pain sensitivity were not
protocol applied (21). Yet, studies investigating exercise- observed for these muscles. A too low training intensity and/or
induced changes in motor performance show somewhat con- insufficient intervention duration may have accounted for the
flicting results. For instance, training of the deep neck flexors lack of significant changes in muscle tissue sensitivity (36). In
using CCF exercises has been shown to produce significant addition, a longer follow-up period may inform whether there
improvements in neuromuscular coordination between the deep are eventual changes in pressure pain sensitivity.
and superficial neck flexors, but negligible effects on flexor
muscle strength (16, 21) or neck flexor muscle activation during Study limitations
a functional upper limb activity (18). Furthermore, endurance
training of the deep cervical flexors showed significant gains in Some limitations of this study should be considered. We only
endurance performance and some carryover effects in strength, included people with neck pain with a NDI score less than
but no gains in coordination performance (21). However, other 15/50, in order to avoid potential aggravation of neck symp-
studies have demonstrated a potential carryover effect of neck toms due to exercise. Thus, generalization of the results to
exercises to domains of motor function different from the patients with higher levels of pain and disability is not feasible,
specific mode of the exercise intervention (7, 14, 17, 19). In and future studies should assess the effectiveness of the 2 pro-
accordance, the current study demonstrates that improvements tocols on individuals with varying clinical severity. Moreover,
in domains of motor performance (i.e. neuromuscular coordi- we did not include a control group that did not receive a form
nation) other than those aligned with the primary behavioural of specific training, thus it is difficult to appreciate the small,
demand of the exercise protocol (i.e. proprioception training), yet significant, reduction in pain and disability.
may be acquired. Unlike previous clinical trials (19, 21), patients in the CCF
An improvement in the activation of the deep cervical exercise group were not given a pressure biofeedback unit for
flexors, as assessed via the CCFT, was obtained with the 2 home training. Although they practiced several repetitions of
interventions applied in this study. Enhanced deep cervical the CCF exercise before training at home, we cannot be sure
flexor muscle performance with CCF exercise was expected that the level or movement they practiced at home was the
and is in agreement with previous observations (7, 14–21). same as under physiotherapist supervision. This fact may
However, improvement in performance on the CCFT following have influenced the results obtained with the protocol of CCF
proprioception training is a novel finding. The high density of training. Similarly, pupillary glasses were only used in the
muscle spindles in the deep cervical flexor muscles (12) could clinic for cervical proprioception training. Both oculomotor
justify the effect of proprioception training on the activation control and cervical joint position sense are often altered in
and control of these muscles and not just the sub-occipital people with neck pain (4), although they were not formally
muscles and cervical multifidus as classically proposed (5). evaluated prior to the study commencement in the way some
Improved performance on the CCFT following a period of authors have recently suggested (37).
cervical proprioception exercise may be also related to other
peripheral and central mechanisms, although the mechanisms Conclusion
underlying the positive effect of the studied interventions on Training protocols of CCF and proprioception training pro-
motor performance were not addressed directly in this study. duced an improvement in activation and endurance of the deep
Other studies have also reported positive effects on perfor- cervical flexors, as assessed via the CCFT, on pain measured
mance of the CCFT with interventions not directly addressing by triple VAS and on the level of disability evaluated with
the activation of the deep cervical flexors, such as passive NDI, with similar results in both groups. However, pressure

J Rehabil Med 48
 Evaluation of specific training for neck pain 55

pain sensitivity was not affected in either group. Propriocep- exercise on sitting posture in patients with chronic neck pain. Phys
tion training may provide an additional benefit of facilitating Ther 2007; 87: 408–417.
18. Falla D, Jull G, Hodges P. Training the cervical muscles with
the deep cervical flexor muscles. Care should be taken in
prescribed motor tasks does not change muscle activation during
extrapolating these results to patients with higher levels of a functional activity. Man Ther 2008; 13: 507–512.
pain and disability. 19. Falla D, Lindstrøm R, Rechter L, Boudreau S, Petzke F. Effec-
tiveness of an 8-week exercise programme on pain and specific-
ity of neck muscle activity in patients with chronic neck pain: a
REFERENCES randomized controlled study. Eur J Pain 2013; 17: 1517–1528.
20. Lluch E, Arguisuelas MD, Coloma PS, Palma F, Rey A, Falla D.
1. Falla D, Farina D. Neuromuscular adaptation in experimental and Effects of deep cervical flexor training on pressure pain thresholds
clinical neck pain. J Electromyogr Kinesiol 2008; 18: 255–261. over myofascial trigger points in patients with chronic neck pain.
2. O’Leary S, Falla D, Elliott JM, Jull G. Muscle dysfunction in J Manipulative Physiol Ther 2013; 36: 604–611.
cervical spine pain: implications for assessment and management. 21. O’Leary S, Jull G, Kim M, Uthaikhup S, Vicenzino B. Training
J Orthop Sports Phys Ther 2009; 39: 324–333. mode-dependent changes in motor performance in neck pain. Arch
3. Jull G, Sterling M, Falla D, Treleaven J, O’Leary S. Whiplash, Phys Med Rehabil 2012; 93: 1225–1233.
headache, and neck pain: research-based directions for physical 22. Lluch E, Schomacher J, Gizzi L, Petzke F, Seegar D, Falla D. Imme-
therapies: research-based directions for physical therapies Edin- diate effects of active cranio-cervical flexion exercise versus passive
burgh: Churchill Livingstone (Elsevier); 2008. mobilisation of the upper cervical spine on pain and performance on
4. Kristjansson E, Treleaven J. Sensorimotor function and dizziness in the cranio-cervical flexion test. Man Ther 2014; 19: 25–31.
neck pain: implications for assessment and management. J Orthop 23. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible
Sports Phys Ther 2009; 39: 364–377. statistical power analysis program for the social, behavioral, and
5. Revel M, Minguet M, Gregoy P, Vaillant J, Manuel JL. Changes biomedical sciences. Behav Res Methods 2007; 39: 175–191.
in cervicocephalic kinesthesia after a proprioceptive rehabilitation 24. Vernon H, Mior S. The Neck Disability Index: a study of reliability
program in patients with neck pain: a randomized controlled study. and validity. J Manipulative Physiol Ther 1991; 14: 409–415.
Arch Phys Med Rehabil 1994; 75: 895–899. 25. Segarra V, Dueñas L, Torres-Cueco R, Falla D, Jull G, Lluch E.
6. Ekvall Hansson E, Månsson NO, Ringsberg KA, Håkansson A. Inter-and intra-tester reliability of a battery of cervical movement
Dizziness among patients with whiplash-associated disorder: a control dysfunction tests. Man Ther 2015; 20: 570–579.
randomized controlled trial. J Rehabil Med 2006; 38: 387–390. 26. Revel M, Andre-Deshays C, Minguet M. Cervicocephalic kines-
7. Jull G, Falla D, Treleaven J, Hodges P, Vicenzino B. Retraining thetic sensibility in patients with cervical pain. Arch Phys Med
cervical joint position sense: the effect of two exercise regimes. J Rehabil 1991; 72: 288–291.
Orthop Res 2007; 25: 404–412. 27. Lucas N, Macaskill P, Irwig L, Moran R, Bogduk N. Reliability of
8. Falla D, Jull G, Hodges PW. Patients with neck pain demon- physical examination for diagnosis of myofascial trigger points: a
strate reduced activation of the deep neck flexor muscles during systematic review of the literature. Clin J Pain 2009; 25: 80–89.
performance of the craniocervical flexion test. Spine 2004; 29: 28. Vatine JJ, Shapira SC, Magora F, Adler D, Magora A. Electronic
2108–2114. pressure algometry of deep pain in healthy volunteers. Arch Phys
9. O’Leary S, Falla D, Jull G. The relationship between superficial mus- Med Rehabil 1993; 74: 526–530.
cle activity during the cranio-cervical flexion test and clinical features 29. Vanderweeën L, Oostendorp RA, Vaes P, Duquet W. Pressure
in patients with chronic neck pain. Man Ther 2011; 16: 452–455. algometry in manual therapy. Man Ther 1996; 1: 258–265.
10. Falla D, O’Leary S, Farina D, Jull G. Association between inten- 30. Ylinen J, Nykänen M, Kautiainen H, Häkkinen A. Evaluation of
sity of pain and impairment in onset and activation of the deep repeatability of pressure algometry on the neck muscles for clinical
cervical flexors in patients with persistent neck pain. Clin J Pain use. Man Ther 2007; 12: 192–197.
2011; 27: 309–314. 31. Vernon H. The Neck Disability Index: state-of-the-art, 1991–2008.
11. Jull GA, O’Leary SP, Falla DL. Clinical assessment of the deep J Manipulative Physiol Ther 2008; 31: 491–502.
cervical flexor muscles: the craniocervical flexion test. J Manipula- 32. Andrade Ortega JA, Delgado Martínez AD, Almécija Ruiz R.
tive Physiol Ther 2008; 31: 525–533. Validation of the Spanish version of the Neck Disability Index.
12. Boyd-Clark LC, Briggs CA, Galea MP. Muscle spindle distribution, Spine (Phila Pa 1976) 2010; 35: E114–E118.
morphology, and density in longus colli and multifidus muscles 33. Sterling M, Jull G, Wright A. Cervical mobilisation: concurrent
of the cervical spine. Spine (Phila Pa 1976) 2002; 27: 694–701. effects on pain, sympathetic nervous system activity and motor
13. Bolton PS. The somatosensory system of the neck and its effects activity. Man Ther 2001; 6: 72–81.
on the central nervous system. J Manipulative Physiol Ther 1998; 34. Jesus-Moraleida FR, Ferreira PH, Pereira LSM, Vasconcelos CM,
21: 553–563. Ferreira ML. Ultrasonographic analysis of the neck flexor muscles
14. Jull GA, Falla D, Vicenzino B, Hodges PW. The effect of thera- in patients with chronic neck pain and changes after cervical spine
peutic exercise on activation of the deep cervical flexor muscles mobilization. J Manipulative Physiol Ther 2011; 34: 514–524.
in people with chronic neck pain. Man Ther 2009; 14: 696–701. 35. O’Leary S, Falla D, Hodges PW, Jull G, Vicenzino B. Specific
15. Jull G, Trott P, Potter H, Zito G, Niere K, Shirley D, et al. A therapeutic exercise of the neck induces immediate local hypoal-
randomized controlled trial of exercise and manipulative therapy gesia. J Pain 2007; 8: 832–839.
for cervicogenic headache. Spine (Phila Pa 1976) 2002; 27: 36. Ylinen J. Physical exercises and functional rehabilitation for the
1835–1843. management of chronic neck pain. Eura Medicophys 2007; 43:
16. Falla D, Jull G, Hodges P, Vicenzino B. An endurance-strength 119–132.
training regime is effective in reducing myoelectric manifestations 37. Della Casa E, Affolter Helbling J, Meichtry A, Luomajoki H, Kool
of cervical flexor muscle fatigue in females with chronic neck pain. J. Head-eye movement control tests in patients with chronic neck
Clin Neurophysiol 2006; 117: 828–837. pain; inter-observer reliability and discriminative validity. BMC
17. Falla D, Jull G, Russell T, Vicenzino B, Hodges P. Effect of neck Musculoskelet Disord 2014; 15: 16.

J Rehabil Med 48