You are here

Article Text

Article menu
Original articles
Workplace
Chronic workplace stress and insufficient physical activity: a cohort study
  1. Anne Kouvonen1,
  2. Jussi Vahtera2–4,
  3. Tuula Oksanen2,
  4. Jaana Pentti2,
  5. Ari K P Väänänen2,
  6. Tarja Heponiemi5,
  7. Paula Salo2,6,
  8. Marianna Virtanen2,
  9. Mika Kivimäki2,7,8
  1. 1School of Sociology, Social Policy & Social Work, Queen's University Belfast, Belfast, UK and Centre of Excellence for Public Health (NI), Queen's University Belfast, Belfast, UK
  2. 2Finnish Institute of Occupational Health, Turku and Helsinki, Finland
  3. 3Department of Public Health, University of Turku, Turku, Finland
  4. 4Turku University Hospital, Turku, Finland
  5. 5National Institute for Health and Welfare, Helsinki, Finland
  6. 6Department of Psychology, University of Turku, Turku, Finland
  7. 7Research Department of Epidemiology and Public Health, University College London, London, UK
  8. 8Institute of Behavioral Sciences, University of Helsinki, Helsinki, Finland
  1. Correspondence to Dr Anne Kouvonen, School of Sociology, Social Policy & Social Work, Queen's University Belfast, 6 College Park, Belfast BT7 1LP, UK; a.kouvonen{at}qub.ac.uk

Abstract

Objectives To examine whether exposure to workplace stressors predicts changes in physical activity and the risk of insufficient physical activity.

Methods Prospective data from the Finnish Public Sector Study. Repeated exposure to low job control, high job demands, low effort, low rewards and compositions of these (job strain and effort–reward imbalance) were assessed at Time 1 (2000–2002) and Time 2 (2004). Insufficient physical activity (<14 metabolic equivalent task hours per week) was measured at Time 1 and Time 3 (2008). The effect of change in workplace stressors on change in physical activity was examined using fixed-effects (within-subject) logistic regression models (N=6665). In addition, logistic regression analysis was applied to examine the associations between repeated exposure to workplace stressors and insufficient physical activity (N=13 976). In these analyses, coworker assessed workplace stressor scores were used in addition to individual level scores.

Results The proportion of participants with insufficient physical activity was 24% at baseline and 26% at follow-up. 19% of the participants who were sufficiently active at baseline became insufficiently active at follow-up. In the fixed-effect analysis, an increase in workplace stress was weakly related to an increase in physical inactivity within an individual. In between-subjects analysis, employees with repeated exposure to low job control and low rewards were more likely to be insufficiently active at follow-up than those with no reports of these stressors; fully adjusted ORs ranged from 1.11 (95% CI 1.00 to 1.24) to 1.21 (95% CI 1.05 to 1.39).

Conclusions Workplace stress is associated with a slightly increased risk of physical inactivity.

http://dx.doi.org/10.1136/oemed-2012-100808

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    What this paper adds

    • Many fail to achieve the recommended levels of physical activity, and workplace stress may contribute to this.

    • However, the evidence on the status of workplace stress as a risk factor for physical inactivity is mixed, and the assessment of workplace stress has typically been based on a measurement at a single time point and/or an assessment of job strain stressors only.

    • Our study shows that an increase in workplace stressors was weakly related to an increase in insufficient physical activity in within-individual analysis.

    • Moreover, our study showed a weak dose–response association between chronic exposure to workplace stressors and the likelihood of insufficient physical activity.

    • Interventions to support physical activity among stressed employees might prevent some of the adverse health effects of chronic workplace stress.

    Introduction

    Insufficient physical activity is a widespread public health problem. Globally, around a third of adults are insufficiently active.1 The current recommendation is that adults should take part in physical activities of moderate intensity for at least 30 min on at least 5 days a week or in vigorous-intensity aerobic physical activity for a minimum of 20 min on 3 days each week. Combinations of moderate- and vigorous-intensity activity can be performed to meet this recommendation.2 However, many fail to achieve the recommended levels, and workplace stress may contribute to this. Stressful working conditions can result in fatigue and incomplete recovery. In addition, they may limit the individual's ability to make positive changes to their lifestyles,3 and impede the implementation of exercise intentions.4

    To date, evidence on the status of workplace stressors as a risk factor for insufficient physical activity is mixed, with some studies supporting this association,4–11 while others reporting null findings.12 ,13 Methodological limitations including the use of cross-sectional design in many studies may have contributed to some of the inconsistencies in earlier studies. The assessment of workplace stress has typically been based on measurement at a single time point which may fail to capture the effects of change and longer-lasting exposure.14 In addition, most studies have assessed job strain stressors only whereas research on the relationship between effort–reward imbalance (ERI) and physical activity is scarce.

    To overcome these limitations, we conducted a large-scale study in Finnish employees to investigate the association between change in workplace stressors and change in physical activity among those with a change in both the exposure and outcome across the three survey phases. In addition, we examined whether repeated exposure to workplace stressors was associated with an increased risk of insufficient physical activity.

    Methods

    Sample and design

    Data were obtained from the Finnish Public Sector Study, which is an ongoing cohort study of employees in the service of 10 municipalities and 21 hospitals in Finland.15 A total of 48 598 employees responded to the first survey in 2000–2002 (Time 1) (response rate 68%). Of these respondents, 36 440 were alive and still employed by the target organisations at the time of the second survey in 2004 (Time 2), and of them, 29 180 responded (response rate 80%). At Time 3 (2008), 18 431participants responded (response rate 87% among those who were still employed by the target organisations and responded both at Time 1 and Time 2). The employers’ records were used to identify the eligible populations for surveys and the work unit code for each employee. Using unique national ID numbers, the respondents were linked to comprehensive national health registers from 1994 to 2005. To determine workplace stressors, each participant's work unit at the lowest level of organisational hierarchy, such as a kindergarten or a hospital ward, was identified from employers’ records. The study was approved by the Ethics Committee of the Finnish Institute of Occupational Health.

    Participants with missing information on any study variables (N=3695) were excluded. The final sample with repeated measures of workplace stressors therefore included 13 976 participants. The final cohort did not substantially differ from those who had participated at earlier phases but had left the organisation or did not respond to the follow-up surveys (N=10 749) in terms of mean age (44.0 years in the sample vs 46.2 years in the excluded population), the proportion of women (82% vs 81%) and socio-economic status (SES) (16% vs 20% low).

    Measures

    Workplace stressors

    Multiple workplace stressors based on two leading stress models, the job strain model (also known as the demand-control model)16 and the ERI model,17 were measured. As previously,18 workplace stressors were assessed in two ways: (a) using each individual's own assessment and (b) summing up the assessments of coworkers and linking that score to each employee in the work unit. In other words, in addition to workplace stressor scores based on self-report, every participant was linked to scores that were compiled from all coworkers’ responses in the same work unit but excluded the participant's own response. Coworker assessed scores were constructed to address potential reporting bias, that is, to eliminate artificial inflation of associations due to common methods to assess the exposure and the outcome.

    Assessment of job strain was based on the modified Job Content Questionnaire.16 Three questions addressed job demands, that is, having high workload and working at a high pace and not having enough time to complete work tasks (Cronbach's α=0.76). Job control was assessed with nine questions about the worker's ability to use and develop skills and exert decision authority (Cronbach's α=0.82). The responses were given on a Likert scale ranging from 1=‘very little’ to 5=‘very much’. To construct a job strain measure, the means of job demand scores were subtracted from the means of job control scores.19 As in previous studies,18 ,20 for both self-reported and coworker assessed job control, demands and strain, scores were further divided into tertiles for between-subjects analysis. Repeated exposure to workplace stress over Time 1 and Time 2 was measured by adding together the number of times (0, 1 or 2) the participant was in a low control, high demands or a high strain job, respectively. In within-subject analyses, job strain stressors were dichotomised using the median split (high vs low).

    Effort was measured with the following item: ‘How much do you feel you invest in your job in terms of skill and energy?’ Rewards were assessed with a scale containing three questions about feelings of getting in return from work in terms of income and job benefits, recognition and prestige, and personal satisfaction (Cronbach's α=0.64).21 Response format for all the questions was a 5-point Likert scale ranging from 1=‘very little’ to 5=‘very much’. The indicator of ERI was obtained by calculating the ratio between the response score in the effort scale and the mean response score in the reward scale. The present measure of ERI has been shown to be an independent measure of workplace stress and has been associated with health and health behavioural outcomes in earlier cross-sectional studies.11 ,21 ,22 As in previous studies, the distributions of the individual and coworker assessed effort, rewards and ERI scores were divided into tertiles for between-subjects analysis.11 ,20 The accumulation of exposure to low effort, low rewards and high ERI over the two measurement points was computed by adding together the number of times the participant was in the most unfavourable tertile. In within-subject analyses, ERI stressors were dichotomised using the median split (high vs low).

    Insufficient physical activity

    Participants reported the average amount of time spent per week on leisure and on the journey to and from work in physical activity corresponding to the activity intensity of walking, vigorous walking, jogging and running. The time spent at each activity in hours per week was multiplied by its typical energy expenditure, expressed in metabolic equivalent tasks (METs). We used the following MET values (work metabolic rate divided by resting metabolic rate): 4 (for exercise intensity corresponding to walking), 6 (vigorous walking to jogging), 10 (jogging) and 13 (running) and express the activity MET index as the sum score of MET h/week.23 Moderate-intensity physical activity for about 30 min at least five times a week is recommended by physical activity guidelines;2 approximately 14 MET h per week correspond to the energy expenditure (1000 kcal, eg, brisk walking for 2.5 h/week equals 15 MET h) needed for reducing health risks. Therefore, the respondents whose volume of activity was < 14 MET h/week were classified as being insufficiently active.24 This cut-point was further used in the fixed-effects analysis to assess change in physical activity (ie, from physically active to insufficiently active).

    Covariates

    Sex, age, employer type (municipality vs hospital) and SES were obtained from employers’ records. SES was assessed using the occupational-title classification of Statistics Finland: high (eg, physicians, teachers), intermediate (eg, technicians, registered nurses) and low (eg, cleaners, maintenance workers).25 Marital status (married or cohabiting vs other) was obtained from the survey. Working hours were summed from the respondent's reports of their (i) official working hours per day and (ii) mean hours of paid or unpaid overtime and their mean hours in another job per day. The daily working hours were multiplied by 5 to obtain the weekly hours in paid work and then dichotomised as less than 40 vs 40 h or more per week. Standard questionnaires were used to assess heavy drinking (>210 g of absolute alcohol per week vs less) and smoking status (current smoker vs non-smoker). The respondents self-reported their weight and height. Body mass index (BMI) was calculated as self-reported weight (kilograms) divided by self-reported height (metres) squared.

    The presence of chronic physical illness was derived from the Drug Reimbursement Register which contains information on persons entitled to special reimbursement for the treatment of chronic conditions and diseases, and the date when the special reimbursement is granted. Patients who apply for special reimbursement must submit a detailed medical statement prepared by the treating physician confirming the diagnoses. All participants with hypertension, cardiac failure, ischaemic heart disease, diabetes, asthma or other chronic obstructive lung disease, and rheumatoid arthritis at the end of the baseline survey year were identified.26 Data on cancer diagnosed during the baseline survey year or four preceding years were obtained from the Finnish Cancer Registry.27 The presence of any of these illnesses was coded (yes/no). Suboptimal self-rated health was assessed with the question ‘In general, would you say your health is very good, good, fair, poor, or very poor?’ (fair to very poor indicated suboptimal health).28 The presence of common mental disorders was assessed with the 12-item version of the General Health Questionnaire (scores ≥ 4 indicated common mental disorder).29 The selected covariates have been associated with physical activity in earlier studies.24 ,30

    Statistical analysis

    To analyse within-subject changes, the fixed-effects method using conditional logistic regression with time-discrete variables was applied to model the effect of change in workplace stress on change in physical activity among those with a change in both the exposure and outcome across the three survey phases (N=6665). In relation to the exposure, ‘change’ refers to moving from the low-stress group to the high-stress group during the follow-up, or from the high-stress group to the low-stress group. In a similar way, in relation to the outcome, ‘change’ refers to moving from the insufficiently active group to the sufficiently active group, or vice versa. In within-subjects analysis, the aim is to examine whether in repeated measurements the changes in the exposure and outcome variables of interest are in the same direction. Fixed-effects methods can be applied in cohort studies using a case-control design, in which the individual is at the same time his/her own case and control. This is possible with repeated measurements when the same individual is, for example, insufficiently active (case) at one study phase and sufficiently active (control) at another study phase. The research question is whether the individual reports high workplace stress when he/she is a case compared with when he/she is a control. More specifically, this analysis enabled us to examine whether physical activity decreases when workplace stress increases. In the analysis of longitudinal data, the fixed-effects method offers the advantage of controlling for stable characteristics of individuals, whether measured or not, by using within-subject variation only to estimate the regression coefficients.31 Because the case and the control share all stable (eg, sex, genes) and non-measured (eg, personality) characteristics, all examined exposures and covariates need to be time variant.

    In addition, logistic regression analysis was applied to examine the associations between repeated exposure to workplace stressors at Time 1 and Time 2 and insufficient physical activity at Time 3. The results are presented as adjusted ORs and their 95% CIs. The contribution of the covariates to the associations between workplace stressors and insufficient physical activity was examined by including each of the following sets of factors in turn: baseline insufficient physical activity, socio-demographics (sex, age, SES, marital status, working hours and employer type), and health status and health behaviours (chronic illness, sub-optimal self-rated health, common mental disorders, BMI, smoking and heavy drinking). Finally, the analysis was conducted with a simultaneous adjustment for all these factors. To test the robustness of our findings and to further examine the temporality between the variables, a sensitivity analysis excluding those participants who were insufficiently active at baseline was run.

    The analyses were conducted in the combined sample of men and women, and in all SES groups together, since the sex and SES interactions were not significant (all p>0.05 in the final models).

    SAS V.9.2 programme package was used for all analyses (SAS Institute Inc, Cary, North Carolina, USA).

    Results

    The characteristics of the study cohort and the associations between the study variables at baseline (Time 1) and insufficient physical activity at baseline (Time 1) and at follow-up (Time 3) are displayed in table 1. The majority (81%) of the participants were women and 53% represented intermediate SES group. Overall, 82% were contracted to municipalities. The proportion of participants with insufficient physical activity was 24% at baseline and 26% at follow-up. In all, 19% of the participants who were sufficiently active at baseline became insufficiently active at follow-up. Both at Time 1 and Time 3, insufficient physical activity was more prevalent in men, increased with age and was related to low SES. Participants with chronic illness, common mental disorder, suboptimal self-rated health and current smokers reported significantly more often insufficient physical activity both at Time 1 and Time 3 than their counterparts. Moreover, participants living without a partner, municipal employees and heavy drinkers reported more insufficient physical activity at Time 3. Mean BMI at baseline was higher in insufficiently active employees both at Time 1 and Time 3.

    View this table:
    Table 1

    Baseline (Time 1) characteristics of the participants and the prevalence of insufficient physical activity (<14 MET h/week) at baseline (Time 1) and follow-up (Time 3), the Finnish Public Sector Study, 2000–2008 (N=13 976)

    Table 2 presents the results from the within-individual analyses among the 6665 participants who had a change in physical activity between the study phases. As the table shows, slightly higher ORs of insufficient physical activity at Time 3 were observed among those who had experienced an increase in workplace stress while these odds were lowered among those who had experienced a decrease in workplace stress, measured as low job control, high strain and low effort. These associations changed little after adjustment for self-rated health.

    View this table:
    Table 2

    Within-individual analyses of the effect of workplace stressors on insufficient physical activity

    When workplace stressors were assessed by coworker reports, higher ORs of insufficient physical activity at Time 3 were observed among those whose coworkers reported decrease in job control.

    Online supplementary table 3(web only) summarises the results from logistic regression analyses on the associations between repeated exposure to job strain stressors at Time 1 and Time 2, assessed by individuals and coworkers, and insufficient physical activity at Time 3. In unadjusted model, the OR for insufficient physical activity was 1.5-fold higher in employees with repeated reports of low individual job control compared with their counterparts with no reports of low job control. Adjustments led to attenuation in the OR but the relationship remained statistically significant (OR=1.18, 95% CI 1.06 to 1.31; Model 5). The result was replicated when repeated exposure to low job control was assessed using coworker reports (OR=1.11, 95%: 1.00 to 1.24; Model 5). Job strain and high job demands were not associated with insufficient physical activity.

    When those who were insufficiently active at baseline (N=3296) were excluded from the analysis, the weak dose–response association between repeated exposure to low job control at Time 1 and Time 2 and insufficient physical activity at Time 3 remained statistically significant (OR=1.15; 95% CI 1.01 to 1.31 in the final model, at the individual level; data not shown).

    The associations between repeated exposure to ERI stressors at Time 1 and Time 2 and insufficient physical activity at Time 3 are presented in online supplementary table 4 (web only). After adjustment for all covariates, those participants who reported repeated exposure to low rewards had a slightly higher likelihood of insufficient physical activity compared with those participants who did not report any low rewards at Time 1 and Time 2 (OR=1.14, 95% CI 1.02 to 1.28; Model 5), and the association was to the same direction when coworker assessed scores were used. Experiencing low effort at one time point was associated with an increased likelihood of insufficient physical activity (OR=1.19, 95% CI 1.06 to 1.34; Model 5) but this result was not replicated when coworker assessments were used. ERI was not associated with insufficient physical activity.

    The weak dose–response relationship between repeated exposure to low rewards at Time 1 and Time 2 and subsequent insufficient physical activity at Time 3 remained after exclusion of participants who were physically inactive at baseline (OR=1.21; 95% CI 1.05 to 1.39 in the final model, at the individual level; data not shown).

    Discussion

    The present study investigated the relationship of repeated exposure to and a change in workplace stressors, and insufficient physical activity in a large sample of Finnish public sector employees. The results from the fixed-effect analyses suggest that an increase in workplace stress is related only to a very slight increase in insufficient physical activity within an individual. Moreover, the between-individual comparisons showed that repeated exposure to low job control and low rewards was weakly associated with an elevated likelihood of insufficient physical activity in a dose–response manner. Consistent effectswere obtained using both individual and coworker assessed scores, which supports the assumption that the health behavioural consequences of workplace stress may not depend only on the perceptions of an individual but also on external working conditions. However, again the effect sizes were small.

    Previous research on workplace stress and leisure-time insufficient physical activity predominantly relates to the job strain model. Some earlier cross-sectional6 ,8 and prospective9 studies have demonstrated an association between low job control and low physical activity.

    We found evidence for a weak association between chronic exposure to low job control and low rewards and the risk of insufficient physical activity. Lack of control at work may spill over to leisure time and be connected to feelings of helplessness, which may make participation in physical activities more challenging.9 Furthermore, it has been suggested that employees with low job control may have less time to plan opportunities or adjust their leisure time for participating in physical activities.9 Repeated exposure to low rewards may be associated with insufficient physical activity potentially through its association with fatigue. Low rewards have predicted fatigue in previous studies.32

    Study strengths and weaknesses

    To the best of our knowledge, this is the first large-scale study which examined the relationship between repeated exposure to both job strain and ERI stressors in relation to insufficient physical activity. A particular strength of this study is its longitudinal design where we can employ analysis of change. Other merits of this study include simultaneous inclusion of a number of covariates, non-response patterns that are unlikely sources for major selection bias and the operationalisation of insufficient physical activity corresponding to the contemporary recommended guidelines of minimum level of physical activity for adults.2

    Moreover, coworker assessment was used to measure workplace stress. The advantage of using coworker assessment is that common method bias, which is of particular concern when both the independent and dependent variables are perceptual measures derived from the same respondent, can largely be avoided. A further strength was the use of fixed-effects regression, which examines the effect of change in workplace stressors to a change in physical activity within an individual, a robust method to take into account all the observed and unobserved variables that are constant over time, thereby controlling for potential omitted time-invariant variables that could confound the associations.31

    Nevertheless, some limitations need to be taken into account. First, physical activity was measured by self-reports. This method is common practice in large-scale epidemiological studies,24 but is affected by reporting bias. Second, the use of coworker assessed scores in measuring workplace stress may reduce self-report bias, but at the same time it is insensitive to true differences in workplace stressors between the employees within a work unit. Third, even if prospective data were used it is not possible to fully exclude the possibility of reverse causation, that is, if employees experience more workplace stress because of lack of physical activity. Last, although the large size and diversity of the sample guarantee a certain generalisation of the results, the present data were female-dominated and from the Finnish public sector and cannot be assumed to represent the general population.

    Conclusions

    This large-scale prospective study shows that an increase in workplace stressors, such as low control, high job strain and low effort, was weakly associated with an increase in insufficient physical activity within an individual. In addition, we found a weak dose–response association between repeated exposure to workplace stressors and the likelihood of insufficient physical activity. Our findings provide one plausible mechanism mediating the previously observed effects of workplace stressors on morbidity, such as depression33 and heart disease.34 This study suggests that interventions to support physical activity among stressed employees could prevent some of the adverse health effects of chronic workplace stress, but job stress intervention studies are needed to confirm this. The fact that in the present study the effect sizes were small indicates that other factors such as physical inactivity in childhood35 may be more important predictors of insufficient physical activity in working populations.

    References

    1. ↵
      World Health Organization. Physical inactivity. A global public health problem. http://www.who.int/dietphysicalactivity/factsheet_inactivity/en/index.html (accessed 24 Oct 2011).
    2. ↵
      1. Haskell WL,
      2. Lee IM,
      3. Pate RR,
      4. et al
      . Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Med Sci Sports Exerc 2007;39:1423–34.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Noblet A,
      2. Lamontagne AD
      . The role of workplace health promotion in addressing job stress. Health Promot Int 2006;21:346–53.
      OpenUrlAbstract/FREE Full Text
    4. ↵
      1. Payne N,
      2. Jones F,
      3. Harris P
      . The impact of working life on health behaviour: the effect of job strain on the cognitive predictors of exercise. J Occup Health Psychol 2002;7:342–53.
      OpenUrlCrossRefPubMed
    5. ↵
      1. Bennett GG,
      2. Wolin KY,
      3. Avrunin JS,
      4. et al
      . Does race/ethnicity moderate the association between job strain and leisure time physical activity? Ann Behav Med 2006;32:60–7.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Kouvonen A,
      2. Kivimäki M,
      3. Elovainio M,
      4. et al
      . Job strain and leisure-time physical activity in female and male public sector employees. Prev Med 2005;41:532–9.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Johansson G,
      2. Johnson JV,
      3. Hall EM
      . Smoking and sedentary behaviour as related to work organization. Soc Sci Med 1991;32:837–46.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Choi B,
      2. Schnall PL,
      3. Yang H,
      4. et al
      . Psychosocial working conditions and active leisure-time physical activity in middle-aged US workers. Int J Occup Med Environ Health 2010;23:239–53.
      OpenUrlCrossRefPubMed
    9. ↵
      1. Smith P,
      2. Frank J,
      3. Bondy S,
      4. et al
      . Do changes in job control predict differences in health status? Results from a longitudinal national survey of Canadians. Psychosom Med 2008;70:85–91.
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Yang X,
      2. Telama R,
      3. Hirvensalo M,
      4. et al
      . The benefits of sustained leisure-time physical activity on job strain. Occup Med (Lond) 2010;60:369–75.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Kouvonen A,
      2. Kivimäki M,
      3. Elovainio M,
      4. et al
      . Effort/reward imbalance and sedentary lifestyle: an observational study in a large occupational cohort. Occup Environ Med 2006;63:422–7.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Lallukka T,
      2. Sarlio-Lähteenkorva S,
      3. Roos E,
      4. et al
      . Working conditions and health behaviours among employed women and men: the Helsinki Health Study. Prev Med 2004;38:48–56.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. van Loon AJM,
      2. Tijhuis M,
      3. Surtees PG,
      4. et al
      . Lifestyle risk factors for cancer: the relationship with psychosocial work environment. Int J Epidemiol 2000;29:785–92.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      1. Gimeno D,
      2. Elovainio M,
      3. Jokela M,
      4. et al
      . Association between passive jobs and low levels of leisure-time physical activity: the Whitehall II cohort study. Occup Environ Med 2009;66:772–6.
      OpenUrlAbstract/FREE Full Text
    15. ↵
      1. Kivimäki M,
      2. Hamer M,
      3. Batty GD,
      4. et al
      . Antidepressant medication use, weight gain, and risk of type 2 diabetes: a population-based study. Diabetes Care 2010;33:2611–16.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Karasek RA,
      2. Brisson C,
      3. Kawakami N,
      4. et al
      . The job content questionnaire (JCQ): an instrument for internationally comparative assessments of psychological job characteristics. J Occup Health Psychol 1998;3:322–55.
      OpenUrlCrossRefPubMed
    17. ↵
      1. Siegrist J
      . Adverse health effects of high effort/low reward conditions. J Occup Health Psychol 1996;1:27–41.
      OpenUrlCrossRefPubMed
    18. ↵
      1. Virtanen M,
      2. Vahtera J,
      3. Pentti J,
      4. et al
      . Job strain and psychologic distress influence on sickness absence among Finnish employees. Am J Prev Med 2007;33:182–7.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Landsbergis PA,
      2. Schnall PL,
      3. Warren K,
      4. et al
      . Association between ambulatory blood pressure and alternative formulations of job strain. Scand J Work Environ Health 1994;20:349–63.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Kobayashi Y,
      2. Hirose T,
      3. Tada Y,
      4. et al
      . Relationship between two job stress models and coronary risk factors among Japanese part-time female employees of a retail company. J Occup Health 2005;47:201–10.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Kivimäki M,
      2. Vahtera J,
      3. Elovainio M,
      4. et al
      . Effort-reward imbalance, procedural injustice and relational injustice as psychosocial predictors of health: complementary or redundant models? Occup Environ Med 2007;64:659–65.
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Kouvonen A,
      2. Kivimäki M,
      3. Virtanen M,
      4. et al
      . Effort-reward imbalance at work and the co-occurrence of lifestyle risk factors: cross-sectional survey in a sample of 36,127 public sector employees. BMC Public Health 2006;6:24.
      OpenUrlCrossRefPubMed
    23. ↵
      1. Kujala UM,
      2. Kaprio J,
      3. Sarna S,
      4. et al
      . Relationship of leisure-time physical activity and mortality: the Finnish twin cohort. JAMA 1998;279:440–4.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Seiluri T,
      2. Lahti J,
      3. Rahkonen O,
      4. et al
      . Changes in occupational class differences in leisure-time physical activity: a follow-up study. Int J Behav Nutr Phys Act 2011;8:14.
      OpenUrlCrossRefPubMed
    25. ↵
      Statistics Finland. Classification of occupations. Helsinki, Finland: Statistics Finland, 1987.
    26. ↵
      1. Sjösten N,
      2. Vahtera J,
      3. Salo P,
      4. et al
      . Increased risk of lost workdays prior to the diagnosis of sleep apnea. Chest 2009;136:130–6.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. Teppo L,
      2. Pukkala E,
      3. Lehtonen M
      . Data quality and quality control of a population-based cancer registry. Experience in Finland. Acta Oncol 1994;33:365–9.
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      1. Idler EL,
      2. Benyamini Y
      . Self-rated health and mortality: a review of twenty-seven community studies. J Health Soc Behav 1997;38:21–37.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Goldberg D,
      2. Williams P
      . A user's guide to the general health questionnaire. Windsor, UK: NFER-Nelson, 1988.
    30. ↵
      1. Trost SG,
      2. Owen N,
      3. Bauman AE,
      4. et al
      . Correlates of adults’ participation in physical activity: review and update. Med Sci Sports Exerc 2002;34:1996–2001.
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Allison PD
      . Fixed Effects Regression Methods in SAS. SUGI 31 Proceedings San Francisco, California March 26–29, 2006. http://www2.sas.com/proceedings/sugi31/184–31.pdf (accessed 14 Dec 2011).
    32. ↵
      1. Sembajwe G,
      2. Wahrendorf M,
      3. Siegrist J,
      4. et al
      . Effects of job strain on fatigue: cross-sectional and prospective views of the job content questionnaire and effort-reward imbalance in the GAZEL cohort. Occup Environ Med 2012;69:377–84.
      OpenUrlAbstract/FREE Full Text
    33. ↵
      1. Netterstrøm B,
      2. Conrad N,
      3. Bech P,
      4. et al
      . The relation between work-related psychosocial factors and the development of depression. Epidemiol Rev 2008;30:118–32.
      OpenUrlAbstract/FREE Full Text
    34. ↵
      1. Kivimäki M,
      2. Virtanen M,
      3. Elovainio M,
      4. et al
      . Work stress in the etiology of coronary heart disease–a meta-analysis. Scand J Work Environ Health 2006;32:431–42.
      OpenUrlCrossRefPubMedWeb of Science
    35. ↵
      1. Telama R,
      2. Yang X,
      3. Viikari J,
      4. et al
      . Physical activity from childhood to adulthood: a 21-year tracking study. Am J Prev Med 2005;28:267–73.
      OpenUrlCrossRefPubMedWeb of Science
    View Abstract

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

      Files in this Data Supplement:

    Footnotes

    • Contributors AK: originated the idea for the study. AK: performed the analyses and led all aspects of the work including data analysis and writing. JV, MK: supervised all aspects of the study implementation. JV, TO, JP, AV, TH, PS, MV, MK: helped to conceptualise the ideas, interpret findings, and write and critically review drafts of the article. JP: additionally provided advice in the choice of statistical methods used and helped in the data analysis.

    • Funding The Finnish Public Sector Study is supported by the Academy of Finland (projects 124271, 124322, 129262 and 132944) and the participating organisations.

    • Competing interests None.

    • Ethics approval Ethics approval provided by the Ethics Committee of the Finnish Institute of Occupational Health.

    • Provenance and peer review Not commissioned; externally peer reviewed.