Occupational Medicine 2017;67:52–57

Advance Access publication 7 November 2016 doi:10.1093/occmed/kqw161

Predicting absenteeism: screening for work ability

or burnout
R. Schouteten
Institute for Management Research, Radboud University, 6500 HK Nijmegen, The Netherlands.
Correspondence to: R. Schouteten, Institute for Management Research, Radboud University, PO Box 9108, 6500 HK Nijmegen,
The Netherlands. E-mail: [email protected]

Downloaded from https://academic.oup.com/occmed/article/67/1/52/2445871 by guest on 25 February 2024

Background In determining the predictors of occupational health problems, two factors can be distinguished:
personal (work ability) factors and work-related factors (burnout, job characteristics). However,
these risk factors are hardly ever combined and it is not clear whether burnout or work ability best
predicts absenteeism.
Aims To relate measures of work ability, burnout and job characteristics to absenteeism as the indicators
of occupational health problems.
Methods Survey data on work ability, burnout and job characteristics from a Dutch university were related to
the absenteeism data from the university’s occupational health and safety database in the year fol-
lowing the survey study. The survey contained the Work Ability Index (WAI), Utrecht Burnout Scale
(UBOS) and seven job characteristics from the Questionnaire on Experience and Evaluation of Work
(QEEW).
Results There were 242 employees in the study group. Logistic regression analyses revealed that job char-
acteristics did not predict absenteeism. Exceptional absenteeism was most consistently predicted by
the WAI dimensions ‘employees’ own prognosis of work ability in two years from now’ and ‘mental
resources/vitality’ and the burnout dimension ‘emotional exhaustion’. Other significant predictors of
exceptional absenteeism frequency included estimated work impairment due to diseases (WAI) and
feelings of depersonalization or emotional distance from the work (burnout).
Conclusions Absenteeism among university personnel was best predicted by a combination of work ability and
burnout. As a result, measures to prevent absenteeism and health problems may best be aimed at
improving an individual’s work ability and/or preventing the occurrence of burnout.
Key words Sickness absence; stress; work ability index.

Introduction Work ability is a personal trait of each worker, reflect-

ing his or her means, at the present moment and in
Faced with ageing workforces, many nations seek effect­ the near future, to perform his or her job, meet work
ive methods to determine key antecedents of health demands and apply health and mental resources [8].
problems and thereby minimize the threats of work dis- A systematic review of 20 empirical studies of determi-
ability and premature retirement to ensure a sufficient nants of work ability [9] revealed that work ability may
labour force, as well as a well-funded social security decrease due to individual characteristics (lack of leisure
system [1]. Two main research streams suggest ways to time, vigorous physical activity, older age), lifestyle (obes
identify these antecedents, focused on either personal ity), work demands (high physical or psychosocial work
or work-related factors. First, in extensive research from demands) and physical condition (poor musculoskeletal
the Finnish Institute of Occupational Health [2,3], capacity). Work engagement functions as a determinant
work ability, as a personal characteristic, emerges as of work ability too [10,11]. In addition to these findings,
an affirmative means to mitigate health-related prob- the Work Ability Index (WAI) provides a well-accepted
lems. Second, several other researchers propose that job technique to assess individual work ability. Evidence
demands and other work characteristics are the most based on test–retest reliability confirms the suitability of
important antecedents of health problems in the labour the WAI questionnaire for occupational health research
force [4–7]. and occupational health care [12]. Although the WAI

© The Author 2016. Published by Oxford University Press on behalf of the Society of Occupational Medicine.
All rights reserved. For Permissions, please email: [email protected]
 R. SCHOUTETEN: PREDICTING ABSENTEEISM AMONG UNIVERSITY STAFF 53

was designed explicitly for general applicability, most First, we constructed a survey containing the WAI, a
research using this inventory focuses on occupations that burnout scale and seven work characteristics scales. The
impose physical demands at work, with just a few excep- survey questions and scales came from previously vali-
tions [9,11,13–16]. dated instruments [24–27]. The WAI, which was devel-
Studies focusing on work-related psychosocial factors oped in Finland [2], already has been translated into 25
instead highlight psychological job demands as a central languages, so we used the official Dutch translation [25].
factor [4]. Both industrial and organizational psycho- This questionnaire consists of seven dimensions [24,25]:
logical theories propose that the quality of a person’s
1. Current work ability compared with the lifetime best
working life depends on the balance (or lack thereof)
2. Work ability in relation to the demands of the job
between job demands and something else, such as deci-
3. Number of current diseases as diagnosed by a physician
sion latitude [5], job resources (e.g. social support)
(reverse coding)
[6] or rewards [7]. Any imbalance results in stress and
4. Estimated work impairment due to diseases (reverse
burnout, which then lead to production losses, absent­

Downloaded from https://academic.oup.com/occmed/article/67/1/52/2445871 by guest on 25 February 2024

coding)
eeism [17–19] and health problems for workers [20].
5. Sickness absence during the past year (reverse coding)
However, consideration of job demands needs to dif-
6. Own prognosis of work ability 2 years from now
ferentiate across types of jobs, such as blue- and white-
7. Mental resources/vitality
collar work. The job demands for blue-collar workers
tend to be associated with the pace of work; those for The cumulativeWAI, which adds the scores across the seven
white-collar workers often relate more to long working dimensions, thus ranges from 7 to 49 points. However,
hours or overtime [4]. because prior research indicates that the WAI is not a one-
Across these different approaches, studies concur that dimensional construct [28], we used the seven dimensions
the relative importance of personal and work-related separately, as explanatory variables in the analyses. The
aspects probably varies for different kinds of jobs. The burnout scale includes 16 items from the Utrecht Burnout
work-related factors also differ for different jobs, such Scale [26], which comprises three dimensions: emotional
that they likely exert distinct effects on pertinent out- exhaustion (Cronbach’s α = 0.897), depersonalization
comes, such as absenteeism, work disability or early (measured as emotional distance or cynicism; α = 0.733)
retirement. To assess these various elements, the WAI and feelings of personal accomplishment (α = 0.797).The
questionnaire can identify workers at risk by taking indi- dimensions can be added to create a cumulative score,
vidual measures of occupational health; at the group but prior research recommends studying them sepa-
level, pertinent questionnaires instead focus on burnout, rately [26]. The work characteristics scales come from the
job demands and job resources. Yet only a few studies Questionnaire on Experience and Evaluation of Work [27]
combine measures of burnout and work ability [16,21– and refer to job demands (work load α = 0.829, role ambi-
23], and those that do indicate the relation between these guity α = 0.808, task changes α = 0.748) and job resources
factors. We know of no studies that confirm the ability of (autonomy α = 0.889, voice α = 0.875, relationship with
these measures to function as screening instruments that superior α = 0.898, career opportunities α = 0.765). An
might predict health problems among a labour force. imbalance between job demands and job resources results
Therefore, this study aimed to relate the personal and in work pressure and, eventually, work stress [5]. For these
work-related measures of work ability and burnout to scales, all the predictor variables are treated as continuous
various indicators of health problems. variables.
Second, to assess health problems, we used the uni-
Methods versity’s occupational health and safety (OHS) data-
base, which provided information about absenteeism
The employees (academic and non-academic) of three (frequency and duration) in the year following the sur-
departments of a Dutch university were invited to par- veys. The data included both the start and end dates of
ticipate in a 2009 study, designed to test the usability of all absenteeism events throughout the university, such
WAI. The dean of each department issued the e-mailed that we could derive measures of absenteeism frequency
invitations to these employees; a reminder followed a and duration per employee. This objective data source
week later. Participation was voluntary, but the deans enabled us to test whether the results of the survey
­
also agreed to act on the findings, as necessary. The study measures predicted actual absenteeism in the subse-
was approved in advance by the executive board and quent year. This quasi-longitudinal treatment of the data
works council of the university. We were granted permis- ensures that the measurement of the explanatory varia-
sion to survey employees who were at least 35 years of bles precedes the occurrence of the effect (absenteeism).
age, reflecting the initial goal for this research, which we Finally, as control variables, we included age, gender
conducted in conjunction with the university’s human and work hours per week, due to their potential cor-
resources department. For the current study, we used relations with work ability [2] and burnout [4,6]. We
data from two sources. also differentiated between academic (e.g. professors,
54 OCCUPATIONAL MEDICINE

researchers) and non-academic (e.g. administrative, sec- non-academic staff to be exceptionally absent, which
retarial, student support) staff, with the recognition that might reflect their job functions. That is, academic staff
their distinct job characteristics and job requirements may be less strict about reporting their absences than
might influence the absenteeism behaviours of these two non-academic staff, because they can often work more
groups differently. flexibly, whereas non-academic (support) staff members
generally need to be present in the workplace to perform
their jobs. If they are ill, academic staff can choose to
Results work from home (and not report themselves absent),
In total, 242 employees out of 575 completed the survey whereas non-academic staff who cannot work due to
(42% response rate). An analysis of the absenteeism data their illness must be reported absent.
indicated no significant differences in terms of frequency Regarding exceptional absenteeism, the results in
or duration of absenteeism between respondents and Table 2 indicate that two WAI dimensions and one burn-
non-respondents. The sample consisted of 157 men and out dimension were significant predictors. The better

Downloaded from https://academic.oup.com/occmed/article/67/1/52/2445871 by guest on 25 February 2024

85 women, with a mean age of 49.3 years (SD = 7.49) and employees’ own prognosis of their work ability 2 years
average work experience in the current job of 12.0 years hence (B = −0.542, P < 0.05), the less likely they were
(SD = 9.79). The average length of the working week to be exceptionally absent in the next year (odds ratio
was 32.9 h (SD = 9.61). Most respondents were well [OR] = 0.582). The mental resources and vitality dimen-
educated, such that 79% completed at least some higher sion in the WAI also provided a significant predictor
education. Non-academic staff members were slightly (B = −0.813, P < 0.05). The more respondents enjoyed
overrepresented, with 129 respondents compared with their work, felt fit and had faith in the future, the lower
113 academic staff. their chance of exceptional absenteeism (OR = 0.444).
The absenteeism data were highly skewed, such that The emotional exhaustion dimension of burnout exerted
more than half of the respondents did not report any an influence too (B = 0.560, P < 0.05); the more respond-
absences in 2009. The mean absenteeism frequency was ents experienced emotional exhaustion, the higher their
1.13 (SD = 0.12), with an average duration of 11.3 days chances of exceptional absenteeism (OR = 1.750).
(SD = 2.1). Due to the extremely skewed nature of these For exceptional duration, the same WAI and burnout
absenteeism data, a binary variable indicated ‘exceptional’ dimensions emerged as important predictors. The better
absenteeism. Specifically, absenteeism was considered employees’ own prognoses of their work ability in 2 years
exceptional when the frequency exceeded once per year (B = −0.588, P < 0.01); the more they enjoyed their work,
and/or the duration exceeded 7 days per year (Table 1). felt fit and had faith in the future (B = −0.751, P < 0.05)
These values coincide with the highest quartiles in the and the less they experienced emotional exhaustion
frequency and duration measures. Moreover, a frequency (B = 0.852, P < 0.05), the less likely those employees
of one episode or a duration of 7 days per year could be were to report exceptionally long absenteeism durations.
considered normal and probably due to common illnesses, Finally, exceptional absenteeism frequency stemmed
such as colds, which are unavoidable and not subject to from slightly different WAI and burnout dimensions.
reduction efforts. The focus in this study instead was on The more respondents thought that their workability was
predicting and preventing exceptional absenteeism. impaired by their diseases (B = 0.544, P < 0.05), the
To test whether work ability, burnout and/or work more likely they were to exhibit exceptionally frequent
characteristics predicted exceptional absenteeism, three absences (OR = 1.722). In addition, the more respond-
separate logistic regression analyses featured exceptional ents enjoyed their work, felt fit and had faith in the future
absenteeism, exceptional frequency or exceptional dura- (B = −1.280, P < 0.01), the lower their chance of excep-
tion as the dependent variable (see Table 2). All the ana­ tionally frequent absences (OR = 0.278). The burnout
lyses confirmed that academic staff were less likely than dimension of depersonalization also revealed an unex-
pected relation: feeling distant from work generally is
Table 1. Prevalence of exceptional absenteeism in 2009 (number regarded as a negative work outcome, but in the current
of respondents) analysis, it implied a positive effect, in that when employ-
ees sensed greater distance from their work (B = −1.008,
Exceptional Total P < 0.01), they were less likely to report exceptionally
frequency (>1 time)
frequent absenteeism (OR = 0.365).
In none of the models did work characteristics sig-
No Yes
nificantly predict absenteeism. Even if job demands and
job resources relate to stress and burnout, they did not
Exceptional duration (>7 days)
directly affect absenteeism in this study. In unreported
No 170 10 180
Yes 17 45 62
robustness analyses with work characteristics as explana-
Total 187 55 242 tory variables and burnout and work ability as depend-
ent variables, we further determined that job demands
 R. SCHOUTETEN: PREDICTING ABSENTEEISM AMONG UNIVERSITY STAFF 55

Table 2. ORs for work ability, burnout and work characteristics on exceptional absenteeism, exceptional frequency and exceptional
duration

Variable Exceptional absenteeism Exceptional frequency Exceptional duration

OR 95% CI P value OR 95% CI P value OR 95% CI P value

Control variables
Age 1.00 0.95, 1.05 NS 0.97 0.91, 1.03 NS 0.99 0.94, 1.04 NS
Gender (1 = female) 2.00 0.86, 4.68 NS 2.11 0.82, 5.45 NS 1.41 0.57, 3.48 NS
Work hours/week 1.00 0.97, 1.05 NS 1.01 0.96, 1.06 NS 0.99 0.94, 1.03 NS
Staff type (1 = academic) 0.29 0.13, 0.66 <0.01 0.18 0.07, 0.50 <0.01 0.24 0.10, 0.60 <0.01
Work ability
WAI1: current WA compared with 1.08 0.75, 1.56 NS 0.76 0.50, 1.16 NS 1.06 0.73, 1.55 NS

Downloaded from https://academic.oup.com/occmed/article/67/1/52/2445871 by guest on 25 February 2024

lifetime best
WAI2: WA in relation to job demands 1.47 0.94, 2.30 NS 1.62 0.95, 2.75 NS 1.49 0.94, 2.36 NS
WAI3: current diseases 1.03 0.83, 1.29 NS 0.86 0.68, 1.10 NS 1.11 0.88, 1.41 NS
WAI4: work impairment 1.03 0.66, 1.61 NS 1.72 1.01, 2.93 <0.05 0.94 0.59, 1.49 NS
WAI5: sick leave in past year 0.74 0.44, 1.22 NS 0.59 0.33, 1.06 NS 0.71 0.42, 1.22 NS
WAI6: own WA prognosis 0.58 0.38, 0.89 <0.05 0.78 0.46, 1.32 NS 0.56 0.37, 0.84 <0.01
WAI7: mental resources 0.44 0.22, 0.91 <0.05 0.28 0.12, 0.66 <0.01 0.47 0.23, 0.99 <0.05
Burnout
Emotional exhaustion 1.75 1.02, 3.02 <0.05 1.56 0.81, 2.99 NS 1.79 1.02, 3.15 <0.05
Depersonalization 0.57 0.33, 1.01 NS 0.37 0.18, 0.73 <0.01 0.60 0.33, 1.08 NS
Feelings of personal accomplishment 0.86 0.47, 1.56 NS 1.62 0.82, 3.22 NS 0.86 0.46, 1.61 NS
Work characteristics
Work load 0.65 0.22, 1.98 NS 0.77 0.21, 2.88 NS 0.54 0.17, 1.76 NS
Role ambiguity 1.82 0.72, 4.55 NS 1.23 0.43, 3.51 NS 2.22 0.83, 5.91 NS
Task changes 0.97 0.37, 2.55 NS 0.67 0.21, 2.16 NS 1.06 0.39, 2.90 NS
Autonomy 0.74 0.23, 2.43 NS 0.72 0.18, 2.83 NS 0.64 0.19, 2.21 NS
Voice 1.51 0.52, 4.36 NS 2.10 0.61, 7.23 NS 1.80 0.58, 5.64 NS
Superior 0.63 0.23, 1.70 NS 0.46 0.14, 1.48 NS 0.61 0.22, 1.71 NS
Career opportunities 0.89 0.47, 1.69 NS 0.90 0.43, 1.87 NS 0.86 0.44, 1.69 NS

NS, not significant.

related significantly to higher burnout and lower work and work ability was somewhat unexpected, in that
ability scores. respondents with higher levels of emotional distance from
their work tended to be less frequently absent. Emotional
distance thus might function not only as an indicator of
Discussion
burnout [6] but also as a kind of coping mechanism, ena-
The results of the logistic regression analyses showed bling people to cope with stressful working conditions.
that absenteeism could best be predicted by a combina- Keeping a distance may help university staff prevent nega-
tion of work ability and burnout. The most consistent tive effects, such as health problems or sickness absence.
predictors of exceptional absenteeism were emotional From a theoretical perspective, the results of this study
exhaustion (burnout), the employee’s own prognosis of are in line with findings that suggest that job motivation
work ability 2 years hence (WAI) and mental resources/ and commitment (which are similar to the WAI dimension
vitality (WAI). People who reported high levels of emo- of mental resources/vitality) relate strongly to absenteeism
tional exhaustion (i.e. find their work tiring) tended to be frequency [19]. Prior studies differentiate between the
absent for longer or more frequently, whereas employees predictors of absenteeism frequency (i.e. low job motiva-
with strong mental resources, who enjoyed their work tion as a result of job resources [19]) and duration (health
and believed their strong work ability would continue, issues due to job demands and burnout [19]). The current
were less likely to be exceptionally absent. results instead challenge the uniqueness of the predictors
Other factors functioned as significant predictors of of either absenteeism frequency or duration [29]. That is,
exceptional absenteeism frequency, namely estimated work the burnout dimensions of emotional distance (cynicism)
impairment due to diseases (WAI) and feelings of deper- and feelings of accomplishment relate to absenteeism
sonalization or emotional distance from work (burnout). frequency but not to absenteeism duration in prior work
The more employees felt work impaired due to diseases, [19]. But these results indicate that both personal factors
the more likely they were to be absent more frequently. (measured by work ability) and work-related factors (meas-
However, the direction of the effects of depersonalization ured by burnout) simultaneously predict absenteeism
56 OCCUPATIONAL MEDICINE

frequency. We recognize that this result requires some of this research might implement a power calculation,
caution though, considering the 1-year time frame of the to determine the sample size needed to confirm that the
study. Shorter time frames might be more appropriate for non-significant results in this study truly mean that these
testing the strict distinction between the antecedents of variables failed to predict exceptional absenteeism.
absence duration and frequency [29]. Finally, there are two main managerial implications of
Furthermore, the lack of correlation between work this study. First, when screening for absenteeism, employ-
characteristics (job demands and job resources) and ers should use a combination of work ability and burnout
ab­senteeism indicates that work characteristics have at measures, because both instruments offer good predictors
best an indirect effect on absenteeism, through work ability of absenteeism. Second, efforts to prevent absenteeism
and burnout. The relationship among work characteristics, and health problems should aim at improving the employ-
work ability and burnout thus is complex. A longitudinal ee’s work ability, especially mental resources and vital-
approach (rather than the quasi-longitudinal approach we ity, such as with vitality programmes [30]. Alternatively,
adopted) could help determine specifically how these con- these efforts might function to prevent the occurrence of

Downloaded from https://academic.oup.com/occmed/article/67/1/52/2445871 by guest on 25 February 2024

structs are causally related. burnout, in the form of emotional exhaustion, such as by
We also noted that several work ability and burnout decreasing the level of job demands [5,6,30].
dimensions were important predictors of absenteeism, but By testing whether work ability (personal characteristic)
some of the central WAI dimensions were not. Perceived or burnout and work characteristics (work-related character-
current work ability, current diseases and absenteeism in istics) best predict absenteeism among university staff with
the preceding year did not predict exceptional absentee- a single study, this contribution shows that a combination
ism in the following year. This finding might reflect the of factors can best predict absenteeism. Therefore, measures
sample; work ability scores are relatively high among our to decrease absenteeism should seek to improve employees’
respondents (M = 41.61, SD = 5.12), and only a few of work ability and prevent their experience of burnout.
them indicate low current work ability or a high number
of diagnosed diseases. A more heterogeneous sample in
Key points
terms of work ability could produce different results.
Another factor might be limited information about •• This study combined personal (work ability) and
the character of the absences in the OHS database. work-related (burnout and work characteristics)
Absenteeism is not always sickness absence; situational factors to identify the best predictors of absentee-
issues, such as family needs, also might lead to absent­ ism, as a measure of health problems.
eeism [29]. On the other hand not all health problems •• The results, obtained from a sample of Dutch uni-
result in absenteeism. For example, a secretary with a versity staff, indicate that a combination of work
broken leg might not be able to get to the office and thus ability and burnout best predicted absenteeism.
would need to call in sick, whereas a professor with a •• Measures to decrease absenteeism should aim at
broken leg might work from home or teach from a sitting improving individual work ability and preventing
position and not take any absence. Similarly, a profes- the occurrence of burnout.
sor dealing with a serious health condition might rely on
alternative options, such as a sabbatical, while undergo-
ing treatment. This absenteeism alternative would not
Funding
appear in the database. A more detailed absenteeism This work was supported by SoFoKles (Dutch Social Fund for
database could offer more specific conclusions, though the Knowledge Sector, grant number 08F0313).
we are invariably hindered by privacy restrictions.
By conducting this study among university personnel Acknowledgements
(i.e. white-collar workers) older than 35 years of age, we
extend prior research that focuses mainly on blue-collar The data for this study were gathered in cooperation with the
workers [19] and offer further support for the notion ‘Werkgroep pilot werkvermogen’ (Wouter Brok, Jacqueline
Janssen, Louis Konickx, Pierre Nouws, Elise van Tiem). The
that the relative importance of various antecedents dif-
author thanks Erik Poutsma and Beatrice van der Heijden for
fers for distinct jobs [4]. In particular, the current study their valuable comments on a previous version of this article.
reveals significant differences in absenteeism rates across
different types of jobs (i.e. academic and non-academic
staff) [4]. We also found that although work ability relates
Conflicts of interest
strongly to age [2], age is not a predictor of exceptional None declared.
absenteeism. Applying these new insights to other occu-
pations and age categories would require careful test- References
ing and further research though, using larger, long-term
studies that rely on similar tools but consider other sec- 1. Gould R, Ilmarinen J, Järvisalo J, Koskinen S (eds).
tors, jobs and employees of all ages. Another extension Dimensions of Work Ability. Results of the Health 2000 Survey.
 R. SCHOUTETEN: PREDICTING ABSENTEEISM AMONG UNIVERSITY STAFF 57

Helsinki, Finland: Finish Institute of Occupational Health, 16. Salmela-Aro K, Nurmi J-E. Employees’ motivational
2008. ori­
entation and well-being at work; a person-oriented
2. Tuomi K, Ilmarinen J, Klockars M et al. Finnish research approach. J Organ Change Manage 2004;17:471–489.
project on aging workers in 1981–1992. Scand J Work 17. de Croon EM, Sluiter JK, Frings-Dresen MH. Psy­
Environ Health 1997;23(Suppl. 1):7–11. chometric properties of the need for recovery after work
3. Ilmarinen J, Tuomi K, Klockars M. Changes in the work scale: test-retest reliability and sensitivity to detect change.
ability of active employees over an 11-year period. Scand J Occup Environ Med 2006;63:202–206.
Work Environ Health 1997;23(Suppl. 1):49–57. 18. van Veldhoven M, Broersen S. Measurement quality and
4. Kristensen TS, Bjorner JB, Christensen KB, Borg V. The validity of the ‘need for recovery scale’. Occup Environ Med
distinctions between work pace and working hours in the 2003;60(Suppl. 1):i3–i9.
measurement of quantitative demands at work. Work Stress 19. Bakker AB, Demerouti E, de Boer E, Schaufeli WB. Job
2004;18:305–322. demands and job resources as predictors of absence dura-
5. Karasek RA. Job demands, job decision latitude, and impli- tion and frequency. J Vocat Behav 2003;62:341–356.
cations for job redesign. Admin Sci Q 1979;24:285–307. 20. Sluiter JK, van der Beek AJ, Frings-Dresen MH. The influ-

Downloaded from https://academic.oup.com/occmed/article/67/1/52/2445871 by guest on 25 February 2024

6. Demerouti E, Bakker AB, Nachreiner F, Schaufeli WB. The ence of work characteristics on the need for recovery and
job demands-resources model of burnout. J Appl Psychol experienced health: a study on coach drivers. Ergonomics
2001;86:499–512. 1999;42:573–583.
7. Siegrist J. Work, health and welfare: new challenges. Int J 21. McGonagle AK, Barnes-Farrell JB, Di Milia L et al.
Soc Welfare 2006;15(Suppl 1):S5–S12. Demands, resources, and work ability: a cross-national
8. van den Berg TI, Alavinia SM, Bredt FJ, Lindeboom D, examination of health care workers. Eur J Work Organ
Elders LA, Burdorf A. The influence of psychosocial Psychol 2014;23:830–846.
factors at work and life style on health and work ability 22. de Boer AG, van Beek JC, Durinck J, Verbeek JH, van
among professional workers. Int Arch Occup Environ Health Dijk FJ. An occupational health intervention programme
2008;81:1029–1036. for workers at risk for early retirement; a randomised con-
9. van den Berg TI, Elders LA, de Zwart BC, Burdorf A. trolled trial. Occup Environ Med 2004;61:924–929.
The effects of work-related and individual factors on the 23. Arandelović M, Nikolić M, Stamenković S. Relationship
Work Ability Index: a systematic review. Occup Environ Med between burnout, quality of life, and work ability index—
2009;66:211–220. directions in prevention. Sci World J 2010;10:766–777.
10. Airila A, Hakanen J, Punakallio A, Lusa S, Luukkonen R. 24. Ilmarinen J. The Work Ability Index (WAI). Occup Med
Is work engagement related to work ability beyond work- (Lond) 2007;57:160.
ing conditions and lifestyle factors? Int Arch Occup Environ 25. Blik op Werk. Work Ability Index. Werkwijzer. Utrecht, The
Health 2012;85:915–925. Netherlands: Blik op Werk, 2008.
11. Mache S, Danzer G, Klapp BF, Groneberg DA. Surgeons’ 26. Schaufeli WB,Van Dierendonck D. Utrechtse Burnout Schaal
work ability and performance in surgical care: relations (UBOS). Lisse, The Netherlands: Swets Test Publishers,
between organisational predictors, work engagement and 2000.
work ability. Langenbecks Arch Surg 2013;398:317–325. 27. Van Veldhoven M, Meijman T. Het meten van psychosociale
12. de Zwart BC, Frings-Dresen MH, van Duivenbooden JC. arbeidsbelasting met een vragenlijst. De Vragenlijst Beleving en
Test-retest reliability of the Work Ability Index question- Beoordeling van de Arbeid. Amsterdam, The Netherlands:
naire. Occup Med (Lond) 2002;52:177–181. NIA, 1994.
13. Martinez MC, Latorre Mdo R. Health and work ability 28. Martus P, Jakob O, Rose U, Seibt R, Freude G. A compara-
among office workers. Rev Saude Publica 2006;40:851–858. tive analysis of the Work Ability Index. Occup Med (Lond)
14. Sjögren-Rönkä T, Ojanen MT, Leskinen EK, Tmustalampi 2010;60:517–524.
S, Mälkiä EA. Physical and psychosocial prerequisites of 29. Ten Brummelhuis LL, Ter Hoeven CL, De Jong MDT,
functioning in relation to work ability and general subjec- Peper B. Exploring the linkage between the home domain
tive well-being among office workers. Scand J Work Environ and absence from work: health, motivation, or both? J
Health 2002;28:184–190. Organiz Behav 2013;34:273–290.
15. Marqueze EC, Voltz GP, Borges FN, Moreno CR. A 2-year 30. Schaufeli WB, Enzmann D. The burnout companion for
follow-up study of work ability among college educators. research and practice: a critical analysis of theory, assessment,
Appl Ergon 2008;39:640–645. research, and interventions. London: Taylor & Francis, 1998.