You are here

Article Text

Article menu

Download PDFPDF

Workplace
Original article
Psychosocial factors and well-being among Finnish GPs and specialists: a 10-year follow-up
  1. Marko Elovainio1,2,
  2. Paula Salo3,4,
  3. Markus Jokela1,2,
  4. Tarja Heponiemi1,
  5. Anne Linna3,
  6. Marianna Virtanen3,
  7. Tuula Oksanen3,
  8. Mika Kivimäki4,5,
  9. Jussi Vahtera3,6
  1. 1Department of Health Services Research, The Institute for Health and Welfare, Helsinki, Finland
  2. 2University of Helsinki, Helsinki, Finland
  3. 3Department of Psychology, Finnish Institute of Occupational Health, Helsinki, Finland
  4. 4Department of Psychology, University of Turku, Turku, Finland
  5. 5Department of Epidemiology and Public Health, University College London, London, United Kingdom
  6. 6Department of Public Health, University of Turku and Turku University Hospital, Finland
  1. Correspondence to Professor Marko Elovainio, National Institute for Health and Welfare, P O Box 30, Helsinki 271, Finland; marko.elovainio{at}thl.fi

Abstract

Background Identifying factors that determine well-being among physicians may help to improve the functioning of hospitals and healthcare centres. We examined associations of psychosocial factors with psychological distress and sleep problems in Finnish general practitioners (GPs) and specialists.

Methods In this prospective cohort study, data from repeated measures over 10 years, related to 886 physicians followed-up from 2000 to 2010 (the Finnish Public Sector Cohort Study). Psychological distress was assessed repeatedly using the 12-item General Health Questionnaire, and sleeping problems using the Jenkins scale in three or in four surveys. Psychosocial factors and potential confounders were measured in four surveys over the same period.

Results High job demands were associated with psychological distress in GPs but not in specialists (p for interaction 0.005). This association was slightly stronger in the within-individual analysis than in the ordinary (total effects) regression, suggesting that the association was not confounded by stable differences between individuals. There was suggestive evidence for a stronger association between effort/reward imbalance and psychological distress in GPs compared with specialists (p for interaction 0.06). High demands and effort-reward-imbalance were associated with elevated sleeping problems in both groups, whereas high job control was associated with lower psychological distress but not sleeping problems.

Conclusions These findings suggest that work-related psychosocial factors are partly responsible for the rise of health problems in physicians, such as psychological distress and sleeping problems. Increasing job demands may be a health risk, especially in GPs.

https://doi.org/10.1136/oemed-2012-100996

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.

    Introduction

    Mental disorders, especially depression, distress and even suicidal tendencies, are more common among physicians than in teachers and several other occupational groups.1 ,2 High levels of depression and psychological distress have been reported in general practitioners (GPs) in Britain,3 North America4 and New Zeeland.5 ,6 Also, a combination of poor mental health and poor subjective physical health, seems to be a common experience among many GPs.5 These health problems have great importance for the doctors concerned, but they may also affect the care of patients because of the associations between distress, depression and poor job performance.7 There is increasing awareness of the potential adverse effects of sleep problems among physicians.8 Long-term sleep deprivation has been shown to be associated with various health outcomes, including premature mortality, cardiovascular disease, hypertension, inflammation, obesity, diabetes and impaired glucose tolerance, and with psychiatric disorders, such as anxiety and depression.9

    Psychosocial work environment is associated with mental ill health in working populations.10–14 The most widely tested models of psychosocial work environment is the Karasek-Theorell job-strain model10 which includes two central components: high job demands (the need to work quickly and hard), and lack of control over skill use, time allocation and organisational decisions. The theory suggests that workers who have concurrent high demands and low job control (experience job strain) cannot moderate the stress caused by the high demands through time management or learning new skills. Thus, they become subject to high stress at work, and if on-going, are at increased risk of psychological distress and disease.15 ,16

    The more recently developed Effort Reward Imbalance (ERI) model focuses on the social reciprocity related to work life.17 ,18 The ERI model proposes that the effort spent in the job is rewarded not only by monetary comparable rewards but also by job security, promotion prospects and prestige. In an optimal situation, this exchange is balanced, but when excessive efforts are combined with low rewards at work (effort-reward-imbalance), the person will be at risk of stress.17 It has been shown that failed reciprocity in terms of high efforts and low rewards elicits strong negative emotions in combination with sustained autonomic activation and may thus have adverse long-term consequences for health.19 In addition, recurrent reward frustration reduces commitment and motivation of employees and increases withdrawal behaviour.17

    Compared with other groups of physicians, GPs seem to have a worse work environment and decreased well-being. The results of a previous cross-sectional study of a representative sample of 2500 Finnish physicians found differences in well-being among physicians working in the different healthcare sectors. GPs and medical specialists experienced more distress than private physicians, while GPs’ self-rated health and work ability were lower than those of other physicians. GPs also reported higher levels of work-related psychosocial stressors, such as high job demands and low job control, and lower commitment to their organisations than other physicians, and work-related psychosocial factors seemed to explain a significant part of these differences.20 A previous cross-sectional study using the same dataset as the current study found that periods of short sick leave were more common among GPs compared with specialists, but this difference disappeared after controlling for work characteristics.21

    The present paper extends these earlier analyses by investigating (A) the role of both the job-strain model and ERI model in predicting changes in psychological distress and sleeping problems among GPs and specialists; (B) testing the associations in a 10-year follow-up using up to four repeated measurements and (C) using data from participants working in comparable work units in the public sector.

    Methods

    Participants

    The data were drawn from the ongoing Finnish Public Sector Study which examines psychosocial factors and health of employees in 10 towns in Finland (for more information, see22). The ethics committees of the Finnish Institute of Occupational Health and the Hospital District of Helsinki and Uusimaa have approved the study. The target population of this study was all GPs and specialists working in 10 municipalities. Here, we used data from study phases 2000–2001, 2002, 2006 and 2010, and they are referred to here as study phases 1–4. At each phase, we included all participants who had data on all measurements of symptoms questionnaire up to that phase. The analytical sample included 886 physicians who contributed a total of 1418 person-observations over the follow-up (1121 person-observations from GPs, 297 person-observations from specialists). Nine percent of the respondents provided full data at each study phase. Of the GPs, 59% responded only once, 19% twice, 12% three times and only 10% four times. Among specialists, the corresponding figures were 71%, 16%, 7% and 6%, respectively. The questionnaire on symptoms of distress was not administered at phase 4, so analyses with psychological distress were carried out with a smaller sample size.

    Measures

    Psychological distress

    We assessed symptoms of distress at study phases 1, 2 and 4 using the self-administered 12-item General Health Questionnaire (GHQ)23 which is designed as a screening instrument for use in community settings. In each GHQ-12 item, an enquiry is made about a specific symptom; there are four response categories which are scored from 1 to 4 to indicate the severity/frequency, and for multitime measures, mean scores were used in the analysis as a continuous variable. The GHQ has been validated against standardised clinical interviews and has shown high reliability and predictive validity.24

    Sleeping problems

    Sleeping problems were measured at each of the study phases using the 4-item Jenkins Scale: ‘How often in the past month did you: (A) have trouble falling asleep?; (B) wake up several times per night?; (C) have trouble staying asleep (including waking far too early)? and (D) wake up after your usual amount of sleep feeling tired and worn out?’ These items measured three components of sleeping problems: item 1 sleep onset problems, items 2 and 3 sleep maintenance problems and item 4 unrefreshing sleep. Responses were given along a 6-point scale from 1=never to 6=every night; for multitime measures, mean scores were used in the analysis. This scale has been shown to offer good internal consistency and predictive validity.25 Responses to these four items were averaged to create an overall sleep problem score at each phase.

    Job control and job demands

    Job control and job demands were measured by using items derived from the Job Content Questionnaire.11 Job control was measured by a 9-item indicator (Cronbach α  range over study phases  0.82–0.88, items 4 and 9 reversed when computed): ‘My job allows me to make a lot of decisions on my own’. ‘My job requires me to be creative’. ‘My job requires that I learn new things’. ‘My job involves a lot of repetitive work’. ‘I have a lot of say what happens on my job’. ‘My job requires a high level of skill’. ‘I get to do a variety of different things on my job’. ‘I have an opportunity to develop my own special abilities’. ‘On my job, I have very little freedom to decide how I do my work’. The job-demands scale consists of three items enquiring about workload and pace of work (Cronbach α :s range over study phases  0.77–0.81): ‘I have to work very hard’. ‘My job involves an excessive amount of work’. ‘ I don't have enough time to get my work done’. All items used a 5-point Likert-type response format ranging from 1 (strongly disagree) to 5 (strongly agree).

    Effort-reward-imbalance (ERI)

    As described in detail elsewhere,26 effort was measured with the question: ‘How much do you feel you invest in your job in terms of skill and energy?’; and rewards were assessed with a scale containing three questions about perception of getting in return from work in terms of (A) income and job benefits, (B) recognition and prestige and (C) personal satisfaction (Cronbach α :s range over study phases  0.64–0.66). Response format for all the questions was a 5-point Likert scale ranging from 1=‘very little’ to 5=‘very much’. Rewards were assessed as a mean score of the three rewards questions. If half or more of the component items were missing, a value of ‘missing’ was recorded in the total reward score. To measure ERI, a ratio of effort (numerator) and the mean of rewards (denominator) was computed in accordance with the procedure used in recent publications.27 ,28

    Potential confounders

    In addition to sex and age, the potential confounders included information on the type of work contract (permanent or fixed-term).

    Statistical analysis

    The data were analysed using random-intercept multilevel regression. The data were structured so that each participant could contribute up to four person-observations to the dataset, depending on the number of study phases for which data on psychosocial factors and health outcomes for the participant were available (table 1). The analyses method allowed us to use all the available data of pooled person-observations without having to exclude individuals with missing data on some of the measurement times. Random-intercept multilevel regression takes into account the non-independence of repeated measurements (person-observations) collected from the same individuals by decomposing the residual term into error variance between different individuals and error variance of repeated measurements within the same individuals. In the present study, we used random-intercept regression to pool the data for longitudinal analysis, but we were not concerned with random effects beyond this, so the details of the random effects are not reported here.

    View this table:
    Table 1

    Descriptive statistics of the sample, means (SD) or %

    First, we examined how the psychosocial risk factors (job control, job demands and ERI) and health outcomes (psychological distress and sleeping problems) changed over time, and whether there were significant differences between GPs and specialists in these time trends. We then examined the associations between psychosocial risk factors and health outcomes over time, and whether these associations were different in magnitude in GPs compared with specialists.

    In a longitudinal setting, the regression coefficients of random-intercept regression are determined as weighted averages of between-individual associations (ie, comparing different individuals with different covariate values) and within-individual associations (ie, comparing covariate values from different measurement times within the same individual).

    If the average between-individual differences are removed from the model by using within-individual regression analysis (also known as fixed-effect regression), it is possible to estimate associations based only on the longitudinal variation in the covariates within individuals. Thus, to minimise confounding arising from differences in time-independent characteristics between individuals, we repeated the main analyses applying within-participant regression models of variation in psychosocial factors and health outcomes using the fixed-effect estimator. This analysis examines whether, across the repeated measurement times, an individual reports worse health when reporting higher work stress, compared with measurement times in which the individual reports lower work stress.

    Data analysis was carried out with STATA 12 software using the ‘xtreg’ multilevel regression procedure, and the associations were expressed as unstandardised regression coefficients (B).

    Results

    GPs in our data were slightly younger, more often women, and had more often permanent contracts compared with specialists (table 1). As the first step in our analyses, we examined whether being a GP or specialist predicted trends in psychological distress or sleeping problems. Figure 1 shows the mean trajectories of psychological distress and sleeping problems in GPs and specialists over study phases. The results suggest that there were no statistically significant differences in the initial levels or trends in symptoms of distress or sleeping problems between GPs and specialists.

    Figure 1
    Figure 1

    Sleeping problems and psychological distress General Health Questionnaire over the study period in general practitioners and specialists (others). The 95% CIs (error bars) are shown only for the smaller group of specialists.

    The trajectories of work-related psychosocial risk factors are presented in figure 2. There were no differences between GPs and specialists in the initial level, or time trends in job control or ERI. However, GPs reported significantly higher job demands compared with specialists, and this difference continued over time despite the slightly declining trend in both groups.

    Figure 2
    Figure 2

    Work-related psychosocial factors (job demands, job control and effort-reward-imbalance) over the study period in general practitioners and specialists (others). The 95% CIs (error bars) are shown only for the smaller group of specialists.

    We then examined whether the psychosocial factors were differently associated with health outcomes in GPs compared with specialists. A significant interaction effect (p=0.005) indicated that job demands predicted psychological distress only in GPs (B=0.15, 95% CI 0.11 to 0.18) but not in specialists (B=0.04, 95% CI −0.03 to 0.10) (table 2). The association between job demands and psychological distress among GPs was slightly stronger in the within-individual analysis than in the ordinary regression (p=0.06), suggesting that this association was not confounded by stable differences between individuals. There was also some evidence for a stronger association between ERI and psychological distress in GPs (B=0.30, CI 0.22 to 0.37) compared with the association in specialists (B=0.14, CI −0.02 to 0.29), but this interaction effect was not quite significant (p=0.06).

    View this table:
    Table 2

    Associations between job characteristics and well-being in GPs and specialists

    There were fewer person-observations from specialists compared with GPs, and the dropout rate was slightly higher in specialists compared with GPs (discrete survival analyses OR 1.55, p=0.006). However, adjusting the final model for differences in the maximum follow-up time (the pattern mixture modelling) did not change the results.

    Discussion

    The present study uncovered two main findings. First, Finnish GPs and specialist doctors did not experience increasing or decreasing levels of psychological distress23 or sleeping problems25 in the first decade of the 21st century, implying that there was essentially no period effect. However, the sample was ageing during that period, and despite adjustment for age, there might be some residual confounding as levels of psychological distress and sleeping problems tend to decline with age. Little change was also observed for job demands and job control, but there was a slightly increasing trend in ERI imbalance in specialists but not in GPs. The only significant difference between GPs and specialists was in job demands, which were higher in GPs than in specialists, and the relative difference in job demands even slightly increased between the groups over time.

    Second, our results supported both the demand-control29 and ERI models17 in predicting psychological distress and sleep problems. There was one significant difference in this pattern between GPs and specialists. Job demands were associated with psychological distress in GPs but not in specialists. This association was replicated in within-individual analysis, implying that the association was not confounded by stable differences between individuals, for instance, dispositional or personality differences. Similar difference was observed in ERI between the groups, but this difference was not statistically significant. Thus, GPs seem to perceive their job more demanding than specialists, and those demands are also more strongly associated with the GPs distress compared with specialists. Job demands and ERI were both associated with elevated sleeping problems in both groups, whereas high job control decreased distress but not sleeping problems.

    The associations between psychosocial factors measured in this study and health outcomes have been demonstrated in many previous studies,10–14 but evidence on longitudinal associations is scarce. According to our results, changes in psychosocial factors were associated with subsequent changes in psychological distress and sleeping problems. Job demands and job control are more or less expected and previously found stress factors in physicians in clinical work.20 ,30

    Primary care is the first contact part of the healthcare system for patients with any kind of health problems. Taking care of patients can be demanding in primary care, and our results suggest that working in primary care sets higher job demands for the physician than working in secondary care does. These job demands also appear to be more potent correlates of psychological distress among physicians working in primary care compared with those working in secondary care. One contributing factor might be the fact that GPs are often working alone without social support from the colleagues. Theoretically, lack of social support, and especially isolation, is suggested to potentiate the effects of high demands and low job control.15 Furthermore, it has been shown that lack of, or insufficient colleague consultation, may cause distress among physicians.30

    The explanation behind the association between ERI and sleeping problems in both groups of physicians is less self-evident. ERI at work is frequent in employees with limited alternative choice in the labour market, in those exposed to heavy competition, and in overly committed people. Only the last one of these seems to characterise Finnish physicians. There is a serious shortage of physicians in Finland and, thus, the position and status of physicians in the labour market is exceptionally good.

    The development of sleeping problems may involve direct neuroendocrine effects of chronic stress and indirect behavioural effects mediated by physical inactivity and excess alcohol consumption, or both direct and indirect pathways. One of the limitations in this study was that it was not possible to control for health behaviours using repeated measurements over the course of follow-up. It is possible that changes in health behaviours may have acted as mediators, or even confounders, affecting the associations found in this study. However, the associations between psychosocial factors and distress were largely similar to those between psychosocial factors and sleeping problems. The main potential confounder of the association observed here is age, which may be linked with job experience, expertise and experienced job demands and sleeping problems. In the analyses, age did not behave as a major confounder. Adjustment for age, sex and work contract, reduced the size of psychosocial factors effects only slightly.

    We are not able to rule out residual confounding due to factors such as marital status and ethnicity. However, Finnish physicians, ethnically, are a very homogeneous group with less than 6% from abroad, mostly Estonia and other European Union countries. Our results are based on self-reported data that may be biased, and thus, these findings need to be confirmed using more objective measures. The specialists were from a wide variety of specialities, and it would have been preferable to adjust the models for this information because psychosocial work characteristics may vary between specialties. Unfortunately, we did not have that information.

    We used a 12-item version of GHQ to measure distress instead of lengthier and potentially more reliable GHQ scales. Shorter versions are often preferred in epidemiological surveys, especially with groups such as physicians, that normally show low response rates. In addition, the 12-item GHQ has shown good structural, concurrent and predictive validity in population-based samples.31 GHQ has been developed as a screening instrument to be used in primary care, and with the cut-off point of 3/4, only about 18 person-observations of the 1053 could be considered as cases. Importantly, psychological distress, measured by GHQ-12, is associated with increased risk of mortality from several major causes in a dose-response pattern.32

    Our outcomes are timely because it has been estimated that sleep problems affect nearly 20% of the adult population in Western countries.33 It has also been shown that sleep problems lead to adverse physiologic changes. In experimental and epidemiologic studies, both short and long sleep hours have been related to many cardiovascular disease risks including hypertension,34 type-2 diabetes35 and increased body mass index.36 Furthermore, it has been shown that both mental health problems, in general, but especially sleeping problems and sleep deprivation, may cause cognitive problems,37 and even affect clinical performance of physicians.38

    In conclusion, our study linking work-related psychosocial factors to subjective health outcomes in GPs and specialists has public health implications, since it adds to the evidence that there may be a psychosocial dimension to the rapid rise of physicians’ health problems, especially psychological distress and sleeping problems. Thus, interventions should focus target-excessive job demands among GPs (eg, by reducing the list sizes or reorganising the work in a way that some of the clinical procedures are undertaken by other health professionals). This may reduce both high job demands and extra effort needed at GP work.

    What this paper adds

    • No evidence of steep negative trends in psychosocial profiles of physicians was found.

    • Changes in psychosocial factors are, however, associated with wellbeing outcomes in physicians.

    • There are significant differences in these associations between GPs and specialists.

    Acknowledgments

    ME is supported by the Academy of Finland (project #128002) and Finnish Work Environment Fund. MK is supported by the Academy of Finland, the EU New OSH ERA Research Programme and the Finnish Work Environment Fund. MV is the Academy of Finland research fellow (project #258598). The Finnish Public Sector study was supported by the Academy of Finland (projects #129262, and #132944) and the participating organisations.

    References

    1. ↵
      1. Lindfors PM,
      2. Heponiemi T,
      3. Meretoja OA,
      4. et al
      . Mitigating on-call symptoms through organizational justice and job control: a cross-sectional study among Finnish anesthesiologists. Acta Anaesthesiol Scand 2009;53:1138–44.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Lindfors PM,
      2. Meretoja OA,
      3. Luukkonen RA,
      4. et al
      . Suicidality among Finnish anaesthesiologists. Acta Anaesthesiol Scand 2009;53:1027–35.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Firth-Cozens J
      . Individula nd organizational predictors of depression in general practitioners. Br J Gen Pract 1998;48:1647–51.
      OpenUrlAbstract/FREE Full Text
    4. ↵
      1. Hsu K,
      2. Marshall V
      . Prevalence of depression and distress in a large sample of Canadian and residents, interns and fellows. Am J Psychiatry 1987;144:1561–66.
      OpenUrlPubMedWeb of Science
    5. ↵
      1. Appleton K,
      2. House A,
      3. Dowell A
      . A survey of job satisfaction, sources of stress and psychological symptoms among general practitioners in Leeds. Br J Gen Pract 1998;48:1059–63.
      OpenUrlAbstract/FREE Full Text
    6. ↵
      1. Dowell A,
      2. Hamilton S
      . Job satisfaction, psychological morbidity, and job stress among New Zealand general practitioners. N Z Med J 2000;113:269–72.
      OpenUrlPubMed
    7. ↵
      1. Firth-Cozens J
      . Stress, psychological problems and clinical performance. In: Vincent C, Ennis M, Audley RJ. eds. Medical accidents. Oxford: Oxford University Press, 1993:131–49.
    8. ↵
      1. Rose M,
      2. Manser T,
      3. Ware JC
      . Effects of call on sleep and mood in internal medicine residents. Behav Sleep Med 2008(6):75–88.
    9. ↵
      1. Ferrie J,
      2. Kumari M,
      3. Salo P,
      4. et al
      . Sleep epidemiology—a rapidly growing field. Int J Epidemiol 2011;40:1431–37.
      OpenUrlFREE Full Text
    10. ↵
      1. Karasek RA
      . Job demands, job decision latitude and meantal strain: implications for job redesign. Adm Sci Q 1979;24:285–95.
      OpenUrlCrossRefWeb of Science
    11. ↵
      1. Karasek RA,
      2. Theorell T
      . Healthy work: stress, productivity and the reconstruction of working life. New York: Basic Books, 1990.
    12. ↵
      1. Stansfeld S,
      2. Candy B
      . Psychosocial work environment and mental health–a meta-analytic review. Scand J Work Environ Health 2006;32:443–62.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Stansfeld SA,
      2. North FM,
      3. White I,
      4. et al
      . Work characteristics and psychiatric disorder in civil servants in London. J Epidemiol Commun Health 1995;49:48–53.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      1. Virtanen M,
      2. Honkonen T,
      3. Kivimaki M,
      4. et al
      . Work stress, mental health and antidepressant medication findings from the Health 2000 Study. J Affect Disord 2007;98:189–97.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Theorell T,
      2. Karasek RA
      . Current issues relating to psychosocial job strain and cardiovascular disease research. J Occup Health Psychol 1996;1:9–26.
      OpenUrlCrossRefPubMed
    16. ↵
      1. Theorell T,
      2. Tsutsumi A,
      3. Hallquist J,
      4. et al
      . Decision latitude, job strain, and myocardial infarction: a study of working men in Stockholm. The SHEEP Study Group. Stockholm Heart epidemiology Program. Am J Public Health 1998;88:382–8.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Siegrist J
      . Adverse health effects of high-effort/low-reward conditions. J Occup Health Psychol 1996;1:27–41.
      OpenUrlCrossRefPubMed
    18. ↵
      1. Siegrist J
      . Reciprocity in basic social exchange and health: can we reconcile person-based with population-based psychosomatic research? J Psychosom Res 1998;45:99–105.
      OpenUrlCrossRefPubMed
    19. ↵
      1. Chandola T,
      2. Britton A,
      3. Brunner E,
      4. et al
      . Work stress and coronary heart disease: what are the mechanisms? Eur Heart J 2008;29:640–8.
      OpenUrlAbstract/FREE Full Text
    20. ↵
      1. Kuusio H,
      2. Heponiemi T,
      3. Aalto AM,
      4. et al
      . Differences in well-being between GPs, medical specialists, and private physicians: the role of psychosocial factors. Health Serv Res 2012;47(1 Pt 1):68–85.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Virtanen P,
      2. Oksanen T,
      3. Kivimäki M,
      4. et al
      . Work stress and health in primary health care physicians and hospital physicians. Occup Environ Med 2008;65:364–6.
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Vahtera J,
      2. Kivimäki M,
      3. Pentti J,
      4. et al
      . Organizational downsizing, sickness absence, and mortality: 10-town prospective cohort study. Br Med J 2004;328:555–9.
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Goldberg DP
      . Detecting pychiatric illness by questionnaire. London: Oxford University Press, 1972.
    24. ↵
      1. Pevalin DJ
      . Multiple applications of the GHQ-12 in a general population sample: an investigation of long-term retest effects. Soc Psychiatry Psychiatr Epidemiol 2000;35:508–12.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Jenkins CD,
      2. Stanton BA,
      3. Niemcryk SJ,
      4. et al
      . A scale for the estimation of sleep problems in clinical research. J Clin Epidemiol 1988;41:313–21.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Kivimaki 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
    27. ↵
      1. Kouvonen A,
      2. Kivimaki 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
    28. ↵
      1. Kouvonen A,
      2. Kivimaki 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
    29. ↵
      1. Theorell T,
      2. de Faire U,
      3. Johnson J,
      4. et al
      . Job strain and ambulatory blood pressure profiles. Scand J Work Environ Health 1991;17:380–5.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Kuusio H,
      2. Heponiemi T,
      3. Sinervo T,
      4. et al
      . Organizational commitment among general practitioners: a cross-sectional study of the role of psychosocial factors. Scand J Prim Health Care 2010;28:108–14.
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Aalto AM,
      2. Elovainio MMK,
      3. et al
      . The Beck Depression Inventory and General Health Questionnaire as measures of depression in the general population: a validation study using the Composite International Diagnostic Interview as the gold standard. Psychiatry Res 2012;15:163–71.
      OpenUrl
    32. ↵
      1. Russ TC,
      2. Stamatakis E,
      3. Hamer M,
      4. et al
      . Association between psychological distress and mortality: individual participant pooled analysis of 10 prospective cohort studies. BMJ 2012;345.
    33. ↵
      1. Ohayon MM,
      2. Carskadon M,
      3. Guilleminault C,
      4. et al
      . Meta-analysis of quantitative sleep parameters from childhood to old age in healthy individuals: developing normative sleep values across the human life-span. Sleep 2004;27:1255–73.
      OpenUrlPubMedWeb of Science
    34. ↵
      1. Gangwisch J,
      2. Heymsfield S,
      3. Boden-Albala B
      . Sorth sleep duration as a risk factor for hypertension: analyses of the ifrst National Health and Nutrition Examination Survey. Hypertension 2006;47:833–9.
      OpenUrlCrossRef
    35. ↵
      1. Yaggi H,
      2. Araujo A,
      3. McKinlay JB
      . Sleep duration as a risk factor the development of type 2 diabetes. Diabetes Care 2006;29:657–61.
      OpenUrlAbstract/FREE Full Text
    36. ↵
      1. Stranges S,
      2. Cappuccio FP,
      3. Kandala N,
      4. et al
      . Cross-sectional versus prospective associations of sleep duration with changes in in relative weight and body distribution: the Whitehall ll study. Am J Epidemiol 2008;167:321–9.
      OpenUrlAbstract/FREE Full Text
    37. ↵
      1. Virtanen M,
      2. Singh-Manoux A,
      3. Ferrie JE,
      4. et al
      . Long working hours and cognitive function: the Whitehall II Study. Am J Epidemiol 2009;169:596–605.
      OpenUrlAbstract/FREE Full Text
    38. ↵
      1. Grantcharov TP,
      2. Bardram L,
      3. Funch-Jensen PG,
      4. et al
      . Laparoscopic performance after one night on call in a surgical department: prospective study. BMJ 2001;323:1222–3.
      OpenUrlFREE Full Text

    Footnotes

    • Contributors All authors listed have contributed significantly to the manuscript.

    • Competing interests None.

    • Ethics approval The Ethics Committees of the Finnish Institute of Occupational Health and the Hospital District of Helsinki and Uusimaa.

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