American Journal of Epidemiology
© The Author 2012. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of
Public Health. All rights reserved. For permissions, please e-mail: [email protected].
Vol. 176, No. 12
DOI: 10.1093/aje/kws337
Advance Access publication:
November 9, 2012
Invited Commentary
Invited Commentary: Job Strain and Health Behaviors—Developing a Bigger
Picture
Anthony D. LaMontagne*
* Correspondence to Dr. Anthony D. LaMontagne, McCaughey Centre: VicHealth Centre for the Promotion of Mental Health and Community
Wellbeing, Melbourne School of Population Health, University of Melbourne, Level 5, 207 Bouverie Street, Parkville, VIC 3010, Australia
(e-mail: [email protected]).
Initially submitted May 15, 2012; accepted for publication July 20, 2012.
Investigation of the association between job stressors and health behaviors has a long history that has been
marked by mixed findings. Fransson et al. (Am J Epidemiol. 2012;176(12):1078–1089) find robust prospective
and cross-sectional associations between job strain and leisure-time physical inactivity in combined data from 14
cohort studies. Further research to better understand the observed heterogeneity in the contributing cohorts and
other studies will be crucial for application to intervention design and tailoring. The population health significance
of these findings requires consideration of other job strain–health behavior ( particularly the parallel analyses
conducted for body mass index and smoking in the same data set) and job strain–health outcome associations,
as well as these same associations for other job stressors. Job strain can be seen as a “fundamental cause” of
work-related disease, in that intervention to reduce exposure to job strain could have beneficial impacts on many
outcomes, making a compelling case for intervention. The significantly strengthened evidence linking job stressors to health behaviors provided by Fransson et al. may help to further direct workplace health promotion research, policy, and practice towards an approach that better integrates intervention on working conditions and
health behaviors. The benefits to population health could be substantial.
health behaviors; job strain; job stressors; physical inactivity
Abbreviation: IPD-Work, individual-participant-data meta-analysis in working populations.
Investigation of the association between exposure to job
stressors and health behaviors has a long history that has
been marked by mixed findings (1). The evidence that exposure to job stressors or poor psychosocial working conditions predicts adverse physical and mental health outcomes
is relatively strong (2–7). The proposed mechanisms
include direct autonomic nervous system, neuroendocrine,
and self-esteem pathways, as well as indirect mediation
through health behaviors (1, 8). The relevant health behaviors include physical activity/inactivity, smoking, alcohol
consumption, and nutritional/dietary behaviors (1). Body
mass index has been investigated in a similar fashion, effectively representing an integrated proxy for low physical
activity and poor nutrition (e.g., high salt, saturated fat, and
alcohol intake). Simply stated, the theoretical argument is
that exposures to job stressors (and associated job stress or
distress) may foster adverse health behaviors as (shortterm) adaptive mechanisms for coping with distress and
may simultaneously reduce the odds of making and succeeding at positive health behavior change (e.g., lowering
the odds of making a smoking quit attempt, as well as the
odds of success when a quit attempt is made) (1, 9). Poor
psychosocial working conditions can also manifest as
passive rather than excessively demanding jobs, which is
hypothesized to result in reduced self-efficacy and passive
life styles (e.g., low physical activity, low civic engagement, and high sedentary time). The theory has both sociologic and biologic plausibility (1, 9). So why the history of
such mixed findings?
In the current issue, Fransson et al. (10) apply a powerful
meta-analytical approach to this question. The authors combined individual-level data from 14 European cohort
1090
Am J Epidemiol. 2012;176(12):1090–1094
Job Strain and Health Behaviors
studies to investigate various job stress research questions
in a sample of ∼170,000 working men and women—the
largest data set to date in this research area. Following
Karasek and Theorell’s demand-control theory (9), the
question addressed was whether leisure-time physical inactivity was more common among employees working in
high job strain (high demand/low control) jobs or passive
(low demand/low control) jobs compared with those in low
strain (low demand/high control) jobs. In addition, the
pooled individual-level data provided adequate power to investigate effect modification by gender, age, socioeconomic
status, and smoking status. Indeed, the variation in physical
inactivity prevalence across cohorts (21% in the pooled
sample, ranging from 7% to 38%) may reflect the social
nature of this behavior and the potential for contextual
influence.
Fransson et al. (10) found robust prospective as well as
cross-sectional associations between poor psychosocial
working conditions and leisure-time physical inactivity,
while acknowledging limitations such as assumptions made
in combining various measures of exposure and outcome
and varying elapsed times between baseline and final measurements. The effect size was modest, with 21%–26%
higher odds of inactivity for participants in high strain and
passive jobs. In addition, there was no cross-sectional association between inactivity and active jobs (high demand/
high control), which are hypothesized to enhance selfesteem and promote health (9), lending further theoretical
coherence to the findings. Although the active job finding
was not borne out in prospective analyses (10, Table 4),
only active jobs at baseline predicted elevated odds of
being physically active at follow-up (odds ratio = 1.10, 95%
confidence interval: 0.98, 1.22). The authors also found evidence of reverse causality (10, Table 5). For example, participants who were physically inactive and in low strain
jobs at baseline had elevated odds of moving into high
strain jobs at follow-up (odds ratio = 1.15, 95% confidence
interval: 1.07, 1.24). The dynamics of such reciprocal relations are a ripe area for future investigation, for health behaviors generally, with potentially valuable implications for
policy and practice, as discussed further below. Future exploration of change-on-change analyses might shed further
light in this regard.
With respect to the main findings, this paper represents
an exemplary use of meta-analysis: resolving a question of
whether there is an association or not, as opposed to simply
gaining effect size estimate precision. An elegant sensitivity
analysis demonstrates that the minimum sample size required to detect the association observed in the full pooled
sample is approximately 13,000 (10, Figure 2). Ten of the
14 contributing cohorts had smaller sample sizes than this,
as well as most of the other published studies on this question, suggesting that inadequate power has been a factor in
the mixed literature findings to date.
Yet, some smaller studies have shown associations
between job strain and physical inactivity, as acknowledged
by Fransson et al. (10). Taken together with the heterogeneity observed across the 14 contributing cohorts (refer to
Web Appendix II, Web Figure 1, available at http://aje.
oxfordjournals.org/), this suggests that effect modification
Am J Epidemiol. 2012;176(12):1090–1094
1091
may also have contributed to the mixed findings to date.
The finding of no significant differences in job strain–
physical inactivity associations by gender, age, socioeconomic status, and smoking in the paper by Fransson et al.
was to some extent surprising, particularly in relation to
gender and socioeconomic status. There is some empirical
evidence of effect modification in previous job stress–
health behavior associations ( particularly by gender (1), as
well as in job stressor–health outcome associations),
though it is inconsistent (11). For example, a German study
found a significantly stronger association between effortreward imbalance and depression among lower versus
higher socioeconomic status workers, but no difference by
socioeconomic status in relation to self-rated health and
angina outcomes (12).
Further research to better understand the observed heterogeneity will be crucial for applying this study’s findings to
intervention design and ultimately to workplace health
policy and practice. Contextual influences might operate in
various ways (e.g., social, cultural, economic) and at various
levels (e.g., from the labor market to the organization or
work context). A recent study of 12 European countries, for
example, found that the association of high job stress with
pronounced depressive symptoms varied according to type
of welfare regime, with the largest effect size in a “neoliberal” country, the United Kingdom, and the smallest in
Scandinavian countries. This suggests that weak social protections may magnify the health implications of poor psychosocial working conditions (13). There is also some
empirical evidence, at the organizational level, of contextual
influences on job stressor–health behavior associations. For
example, a recent US study found that high workplace social
capital buffered the association between job strain and
smoking (14). These examples suggest that contextual
factors, as well as job stressors, could be targeted for improvement in order to reduce the impacts of job stressors
on both health behaviors and health (11). Variation in coexposures might also explain some heterogeneity in job
stressor–health behavior associations. These could include
other psychosocial stressors (e.g., job insecurity, social
support at work, long working hours), as well as other physical or ergonomic working conditions (e.g., sedentary work,
physically demanding work).
The population health significance of the findings by
Fransson et al. (10) needs to be considered in relation to
other job strain–health behavior and job strain–health
associations, as well as these same relations for other job
stressors. The individual-participant-data meta-analysis in
working populations (IPD-Work) Consortium (the group
represented in the paper by Fransson et al. (10)) has also
conducted pooled analyses of job strain–smoking (15) and
job strain–body mass index (16) associations and has found
evidence of association for each of these outcomes as well
as physical inactivity. Consistent with these findings, those
of previous cross-sectional analyses in the largest of the 14
contributing cohorts found that job strain and effort–reward
imbalance were associated with the co-occurrence of 2–3
adverse health behaviors (1, 17). In short, the findings of
Fransson et al. greatly strengthen the accumulating evidence that job stressors are significantly related to health
1092 LaMontagne
Figure 1. Conceptual model of the relations among job stressors, health behaviors, and adverse health impacts.
behaviors, even if the effects taken individually may be
modest and, possibly, context specific. When considered
together with the direct effects, a large number of health
behaviors and health outcomes are associated with job
stressors, ranging from premature mortality to various
common mental disorders and cardiometabolic disease outcomes (Figure 1) (2–8, 18). In brief, there are many more
reasons to reduce job strain (and other job stressors) than
its association with physical inactivity alone.
Job stressors can thus be seen in some sense as “fundamental causes” of work-related disease, in that intervention
to reduce exposure to job strain and other stressors would
have beneficial impacts on many health outcomes
(Figure 1) (19). Indeed, job strain was used in Link and
Phelan’s classic 1995 paper as an example of a link
between social conditions and disease (19, p. 84). Job
strain and many other poor working conditions increase in
prevalence with decreasing occupational status (11, 19), as
do poor health behaviors, such as low leisure-time physical
activity, smoking, and high saturated fat consumption (1).
Setting aside the extent to which population patterns of
poor working conditions might explain patterns of poor
health behaviors, job strain and other job stressors may
account for a substantial preventable disease burden. Job
strain, alone, has been recently estimated to account for
6.5% of common mental health disorders (7), 15% of prevalent depression (20), and 6.5%–25.2% of cardiovascular
disease morbidity and mortality in working populations (7).
The job strain- and job stressor-attributable disease burden
would rise further still if all job strain-associated outcomes
were considered, as well as all other established job
stressor–outcome associations (e.g., for effort-reward imbalance, job insecurity, bullying).
This makes a compelling case for intervention to prevent
and control exposure to job stressors, with the potential to
realize substantial health and economic benefits. There is
growing evidence of the feasibility and effectiveness of
comprehensive work- and worker-directed intervention
strategies to reduce job stressors (21–23). There is also
growing evidence on the feasibility and effectiveness of integrated intervention on working conditions and health behaviors, although the evidence base is more limited (24). In
a 2007 systematic review of 90 job stress intervention
studies, 8 studies were identified that integrated job stress
with health behavioral intervention (21). Only one of these
studies targeted physical activity and included a (nonrandomized) comparison group, reporting significant improvements in cardiovascular health risks (decrease) and
exposures to job stressors (decrease in job demands and an
increase in job control) (25). Given the preventive potential
of integrated approaches, intervention research in this area
should clearly be a high priority, in particular for governments investing substantially in workplace health promotion such as in the United States.
Some government, health promotion, and other agencies
are indeed pursuing integrated approaches to improving
working conditions and health behaviors, as demonstrated
in the US Centers for Disease Control and Prevention
(CDC)/National Institute for Occupational Safety and
Health (NIOSH) “Total Worker Health” Initiative and associated research centers (http://www.cdc.gov/niosh/twh/) and
the American Heart Association’s 2010 worksite wellness
Am J Epidemiol. 2012;176(12):1090–1094
Job Strain and Health Behaviors
policy (26). Yet elsewhere in the Centers for Disease
Control and Prevention, the recently launched “Million
Hearts” Initiative (27) ignores job stressors despite the evidence that job stress-attributable fractions for cardiovascular
disease are comparable to those for physical activity and
fruit and vegetable consumption (7, 28).
The significantly strengthened evidence linking job stressors to health behaviors provided by Fransson et al. (10) and
related work from the IPD-Work Consortium may help to
stimulate intervention to reduce job stressors, as well as to
further direct workplace health promotion research, policy,
and practice towards a more integrated approach. The benefits
to population health could be substantial.
ACKNOWLEDGMENTS
Author affiliation: The McCaughey Centre: VicHealth
Centre for the Promotion of Mental Health and Community
Wellbeing, Melbourne School of Population Health, University of Melbourne, Parkville, Victoria, Australia
(Anthony D. LaMontagne).
This work was supported by the Australian National
Health and Medical Research Council (grant 375196) and
centre grant funding from the Victorian Health Promotion
Foundation (Melbourne).
Conflict of interest: none declared.
REFERENCES
1. Siegrist J, Rodel A. Work stress and health risk behavior.
Scand J Work Environ Health. 2006;32(6):473–481.
2. Belkic K, Landsbergis P, Schnall P, et al. Is job strain a major
source of cardiovascular disease risk? Scand J Work Environ
Health. 2004;30(2):85–128.
3. Stansfeld SA, Candy B. Psychosocial work environment and
mental health—a meta-analytic review. Scand J Work
Environ Health. 2006;32(6):443–462.
4. Kivimaki M, Virtanen M, Elovainio M, et al. Work stress in
the etiology of coronary heart disease—a meta-analysis.
Scand J Work Environ Health. 2006;32(6):431–442.
5. Netterstrom B, Conrad N, Bech P, et al. The relation between
work-related psychosocial factors and the development of
depression. Epidemiol Rev. 2008;30(1):118–132.
6. LaMontagne AD, Keegel T, Louie AM, et al. Job stress as a
preventable upstream determinant of common mental
disorders: a review for practitioners and policy-makers. Adv
Mental Health. 2010;9(1):17–35.
7. Sultan-Taıeb H, Lejeune C, Drummond A, et al. Fractions of
cardiovascular diseases, mental disorders, and musculoskeletal disorders attributable to job strain. Int Arch Occup
Environ Health. 2011;84(8):911–925.
8. Chandola T, Britton A, Brunner E, et al. Work stress and
coronary heart disease: what are the mechanisms? Eur Heart
J. 2008;29(5):640–648.
9. Karasek R, Theorell T. Healthy Work: Stress, Productivity,
and the Reconstruction of Working Life. New York, NY:
Basic Books, Inc; 1990.
Am J Epidemiol. 2012;176(12):1090–1094
1093
10. Fransson EI, Heikkilä K, Nyberg ST, et al. Job strain as a risk
factor for leisure-time physical inactivity: an individualparticipant meta-analysis of up to 170,000 men and women:
The IPD-Work Consortium. Am J Epidemiol. 2012;176(12):
1078–1089.
11. Landsbergis PA, Grzywacz JG, LaMontagne AD. Work
organization, job insecurity, and occupational health
disparities [ published online ahead of print October 16,
2012]. Am J Ind Med. (doi:10.1002/ajim.22126).
12. Wege N, Dragano N, Erbel R, et al. When does work stress
hurt? Testing the interaction with socioeconomic position in
the Heinz Nixdorf Recall Study. J Epidemiol Community
Health. 2008;62(4):338–341.
13. Dragano N, Siegrist J, Wahrendorf M. Welfare regimes,
labour policies and unhealthy psychosocial working
conditions: a comparative study with 9917 older employees
from 12 European countries. J Epidemiol Community Health.
2011;65(9):793–799.
14. Sapp AL, Kawachi I, Sorensen G, et al. Does workplace
social capital buffer the effects of job stress? A crosssectional, multilevel analysis of cigarette smoking among
U.S. manufacturing workers. J Occup Environ Med. 2010;
52(7):740–750.
15. Heikkila K, Nyberg ST, Fransson EI, et al. Job strain
and tobacco smoking: an individual-participant data
meta-analysis of 166,130 adults in 15 European studies.
PLoS One. 2012;7(7):e35463. (doi:10.1371/journal.
pone.0035463).
16. Nyberg ST, Heikkila K, Fransson EI, et al. Job strain in
relation to body mass index: pooled analysis of 160,000
adults from 13 cohort studies. J Intern Med. 2012;272(1):
65–73.
17. Kouvonen A, Kivimaki M, Vaananen A, et al. Job strain and
adverse health behaviors: the Finnish Public Sector Study.
J Occup Environ Med. 2007;49(1):68–74.
18. Chandola T, Brunner E, Marmot M. Chronic stress at work
and the metabolic syndrome: prospective study. BMJ. 2006;
332(7540):521–525.
19. Link BG, Phelan J. Social conditions as fundamental
causes of disease. J Health Soc Behav. 1995;35(spec no):
80–94.
20. LaMontagne AD, Keegel T, Vallance D, et al. Job strain—
attributable depression in a sample of working Australians:
assessing the contribution to health inequalities. BMC
Public Health. 2008;8(181). (doi:10.1186/1471-24588-181).
21. LaMontagne AD, Keegel T, Louie AM, et al. A systematic
review of the job stress intervention evaluation literature:
1990–2005. Int J Occup Environ Health. 2007;13(3):
268–280.
22. Bambra C, Gibson M, Sowden AJ, et al. Working for health?
Evidence from systematic reviews on the effects on health
and health inequalities of organisational changes to the
psychosocial work environment. Prev Med. 2009;48(5):
454–461.
23. Landsbergis PA. Interventions to reduce job stress and
improve work organization and worker health. In: Schnall P,
Dobson M, Rosskam E, et al., eds. Unhealthy Work: Causes,
Consequences, Cures. Amityville, NY: Baywood Publishing;
2009:193–209.
24. Sorensen G, Landsbergis P, Hammer L, et al. Preventing
chronic disease in the workplace: a workshop report and
recommendations. Am J Public Health. 2011;101(suppl 1):
S196–S207.
1094 LaMontagne
25. Maes S, Verhoeven C, Kittel F, et al. Effects of a Dutch
work-site wellness-health program: the Brabantia Project.
Am J Public Health. 1998;88(7):1037–1041.
26. Carnethon M, Whitsel LP, Franklin BA, et al. Worksite
wellness programs for cardiovascular disease prevention: a
policy statement from the American Heart Association.
Circulation. 2009;120(17):1725–1741.
27. Frieden TR, Berwick DM. The “Million Hearts” Initiative—
preventing heart attacks and strokes. N Engl J Med. 2011;
365(13):e27. (doi:10.1056/NEJMp1110421).
28. Yusuf S, Hawken S, Ounpuu S, et al. Effect of potentially
modifiable risk factors associated with myocardial infarction
in 52 countries (the INTERHEART Study): case control.
Lancet. 2004;364(9438):937–952.
Am J Epidemiol. 2012;176(12):1090–1094