- Review previous evidence on associations between psychosocial job demands/control and future labor market situation.
- Summarize the new findings on associations between job demands/control and occupational outcomes 11 years later, based on Swedish population data.
- Discuss the findings for specific labor market outcomes including differences by gender.
Some reviews have shown that being in jobs with high demands and low control is associated with a higher risk of sickness absence (SA)1–3 and disability pension (DP).4 Associations have also been found between psychosocial work characteristics and retirement, with some studies showing that women's retirement decisions are more associated with psychosocial job demands than men,5 although this is not unequivocal.6 Job control has also been shown to be associated with future retirement, both before and after standard retirement age.6–9 A systematic review concluded that while there was strong evidence to suggest that high job control was associated with longer time to retirement, there was only very limited evidence for any association between job demands and retirement.10
Review previous evidence on associations between psychosocial job demands/control and future labor market situation. Summarize the new findings on associations between job demands/control and occupational outcomes 11 years later, based on Swedish population data. Discuss the findings for specific labor market outcomes including differences by gender.
However, less is known about the association of job demands and control with future long-term SA/DP (>6 months), long-term unemployment, and old-age pension, especially with a long follow-up time. Most longitudinal studies in this field have had follow-up times of six months to five years,11,12 although a few that study associations between job demands/control and physical health outcomes have longer follow-up times (one to six years).13 There is also a need for studies with a longer follow-up time, to learn whether job demands/control are associated with labor market situation in a long-term perspective.
Also, outcomes that are related, for example, long-term SA/DP, long-term unemployment, and old-age pension, have not included in the same study before. Since demands and control have been shown to be associated with each of these labor market situations, and some of them are mutually exclusive, there is a risk that results can be biased when only using one of the outcomes. Therefore, merely studying these on their own may lead to mistaken conclusions. One study of people aged ≥50 years, based on survey data from smaller samples and a 4-year follow-up9 reported that low job control was associated with early retirement, unemployment, and disability pension, while another study on nursing aides in Denmark found that high demands were associated with early voluntary retirement, but not disability pension,14 and a study of a random sample from Denmark found that high job demands were associated with a lower risk of unemployment, but not SA or DP.15 Thus, there is a need for more detailed knowledge based on larger cohort and with longer follow-up regarding associations between job demands/control and future different labor market situations, since the associations might differ between labor market situations.
Furthermore, many of the previous studies on job demands and job control were conducted with specific occupational groups, not on the entire working population.16–20 Since the cut-offs for high and low job demands and job control in such studies are set for each specific studied group, comparisons between studies are problematic, and also limits the generalizability of these studies. Thus, there is a need for studies that examine job demands and job control among all in paid work.
Most previous studies in this area used individual's self-reported data on how they experienced their job demands and job control. This means that there might be a bias regarding outcomes; there may be inverse causation that leads people with worse health to estimate the demands of their job as higher and the job control as lower than others with the same job situation.21 Job exposure matrices (JEM), using extensive survey data to estimate mean values of job demands and job control from many respondents in each occupation, are a possible remedy.22–24 In Sweden, there are two JEMs for psychosocial work environment conditions; a JEM constructed by Johnson et al23 and an updated version of this JEM by Fredlund et al.22 In a previous study, we used the latter to study cross-sectional associations between job demands/control and being in paid work, unemployment, and long-term SA/DP in 2001: here the same cohort will be followed prospectively.25
The aim of this study was to analyze the association between combinations of job demands and job control among women and men in different occupations and their future labor market situations in terms of long-term sickness absence/disability pension, long-term unemployment, old-age pension, emigration, and death.
We conducted a population-based prospective cohort study with an 11-year follow-up.
We used anonymized microdata from the Longitudinal Integration Database for Health Insurance and Labor Market Studies (LISA) held by Statistics Sweden to identify the cohort and for information on age, sex, occupation, income, country of birth, type of living area, family situation, and educational level in 2001 for number of net days with unemployment benefits, SA, and DP in 2001 and 2012, and status regarding old-age pension, emigration, and death in 2012.
We included women and men who according to LISA in December 2001 were living in Sweden, were aged 30 to 54 years, had an income from work or work-related benefits ≥8856 SEK (=1003 Euros by the 2001 conversion rate), and an occupation (registered according to the Swedish Standard for Occupational Classification [SSYK by Swedish acronym]). The income limit of 8856 SEK from work and/or benefits was used to include only those with an income high enough to qualify for SA benefits from the Swedish Social Insurance Agency. We excluded those who were on full-time DP in all of 2001, had full-time SA in both 2000 and 2001,26,27 or were on early old-age pension, yielding a study population of 2,194,694 individuals. Each individual was followed up in 2012.
Job Demands and Job Control
For information on job demands and job control, we used the psychosocial JEM from 1999.22 Occupations were classified according to the Nordic Job Classification (NYK), and the JEM was based on the Swedish Working Conditions Surveys from 1989, 1991, 1993, 1995, to 1997. The JEM includes mean scores for job demands and for job control for each NYK occupation by sex in three age groups: 16 to 29, 30 to 44, and 45 to 64 years. Occupations with too few respondents were assigned values of neighboring occupations.22 We translated SSYK categorizations of occupations into NYK, in order to assign each individual a JEM value (for more details on this process, see Norberg et al).25
In order to capture more detailed values on the combination of demands and control, we classified individuals in tertiles of control and demands in a modified “quadrant” model.25 This lead to the introduction of a model with nine groups (instead of the usual four groups) of individuals based on their values of demands and control: (1) high demands/high control, (2) high demands/medium control, (3) high demands/low control, (4) medium demands/high control, (5) medium demands/medium control, (6) medium demands/low control, (7) low demands/high control, (8) low demands/medium control, and (9) low demands/low control. We used group (5) medium demands/medium control as the reference category for the analyses.
Outcome Measures in 2012
We used the following labor market situations: long-term SA/DP (>183 net days with SA and/or DP benefits from the Social Insurance Agency); long-term unemployment (registered as unemployed or in active labor market programs for >183 days, or no income registered); old-age pension (more than half of income from old-age pension); emigration; death. Individuals who did not fulfil any of these conditions were coded as being in paid work.
Sickness Absence Insurance System and Unemployment Benefits in Sweden
All people living in Sweden with income from work or unemployment benefits are from age 16 years covered by the public SA insurance system and can be granted SA benefits if their work capacity is reduced due to disease or injury. Day 1 is a waiting day, with 100% loss of income. After 7 days, a medical certificate is required. The employer usually reimburses income loss during days 2 to 14, after which SA benefits are administered by the Swedish Social Insurance Agency, for unemployed individuals this happens from day 2. SA spells can go on for years.
Also, all people aged 19 to 64 years can be granted DP if their work capacity is long-term or permanently reduced due to disease or injury. Both SA and DP benefits can be granted for full-time (100%) or part-time (75%, 50%, or 25%) of ordinary work hours. This means that it is possible to have both partial SA and DP at the same time, and consequently we combined SA and DP, using net days as described above. SA benefits cover 80% and DP covers 64% of lost income, up to a certain level.
In Sweden, unemployment benefits consist of two parts: a basic insurance and an optional earnings-related insurance. The earnings-related part covers up to 80% of lost income up to a certain level. Individuals 20 years or older enrolled at the employment service and with a job-seeking plan can have a basic insurance, also without previous work experience. This is a subsistence-level fixed amount. In this study, we considered the days with either type of unemployment benefits. Ordinary age for old-age pension was 65 years in 2012, but could be taken earlier.
As covariates we included age (30 to 34, 35 to 44, or 45 to 54 years), sex (woman or man), country of birth (Sweden, other Nordic country, other EU25, or rest of world (including missing)); educational level (elementary (≤9 years or missing), high school (10–12 years), or university/college (>12 years)), type of living area (large city (Stockholm, Gothenburg, Malmö with surrounding municipalities), medium-sized town (municipalities with >90,000 inhabitants within 30 km of municipal center), or rural (municipalities with <90,000 inhabitants within 30 km of municipal center)), family situation (married/cohabiting with children living at home, married/cohabiting with no children living at home, single with children living at home, or single/with no children living at home), and partial unemployment or partial SA/DP for more than half the year in 2001.
We calculated descriptive statistics of the sociodemographic characteristics of the cohort.
We used multinomial logistic regression to calculate odds ratios (ORs) with 95% confidence intervals (CIs) of the associations between combinations of job demands and job control in 2001 and future long-term SA/DP, long-term unemployment, old-age pension, emigration, and death, respectively in 2012, unadjusted and adjusted for all included sociodemographic covariates. We also calculated corresponding ORs for each sociodemographic covariate, mutually adjusted for all others.
Sociodemographic descriptive statistics (n and %) for the cohort in 2001 are presented in Table 1. Nearly half of the cohorts were women (49.3%), the majority (88.9%) were born in Sweden, and about a third (35.6%) had college/university training.
Distribution of Labor Market Situations in 2012
In Table 2, numbers and percent of individuals with each of the studied labor market situations are shown at follow-up in 2012. Of the 2,194,694 participants, the majority (1,824,601; 83.1%) were still in paid work, whereas 370,084 (16.9%) were out of paid work. Women were slightly more likely than men to be out of paid work (17.4% and 16.5%, respectively). The main reason for being out of paid work was in both sexes unemployment (6.3% and 6.8%, respectively), followed by SA/DP (6.1% and 4.7%, respectively), and old-age pension (2.6% and 2.7%, respectively).
Figure 1 shows the proportions of individuals with each of these labor market situations in 2012 by the nine combinations of job demands and job control in 2001. In all the nine groups, most individuals were in paid work, followed by unemployment, SA/DP, and old-age pension. Very few individuals in either of the nine groups had emigrated or died.
The two groups with the highest proportion of individuals in paid work in 2012 were those with the combinations of high demands and high control and high demands and medium control. The group with the smallest proportion in paid work was those initially in jobs with the combination of low demands and low control. Among men, the two groups with low demands and high control, and medium demands and low control also had smaller proportions in paid work, whereas this was not the case among women.
Those in jobs in 2001 with the combination of low job demands and low control or low demands and medium job control had the highest proportion unemployed in 2012, while those with the combination of low job demands and low job control had the highest proportion with long-term SA/DP. Differences in proportions with each of the other outcomes (old-age pension, emigration, death) between the groups were small.
Figure 2 shows proportions with each of the previously mentioned labor market situation in 2012 based on labor market situation in 2001: those who had partial long-term unemployment, partial long-term SA/DP, and those who had neither of those benefits in 2001, respectively. Among those in paid work with no long-term SA/DP or unemployment benefits in 2001, 85% were still in paid work in 2012. Among those with partial long-term SA/DP or partial unemployment in 2001 about half (48%) were in paid work in 2012 in both groups. However, for those who had long-term SA/DP in 2001, 33% also had that in 2012, whereas of those who had long-term unemployment in 2001, 25% also had that in 2012. There was a slightly larger proportion of those who were on SA/DP in 2001 who died before the end of 2012 than in the other groups.
Associations Between Combinations of Job Demands/Job Control in 2001 and Labor Market Situation in 2012
Table 3 shows ORs for long-term SA/DP, long-term unemployment, old-age pension, emigration, and death, respectively, relative to being in paid work and not having any of those specified labor market situations in 2012, by the nine different combinations of job demands and job control in 2001.
In general, the ORs were rarely above 1.5 or below 0.65. There were, however, a few exceptions to this, and there were also some notable sex differences. In the whole population, those in jobs with the combination of low demands and low control were more likely to be on long-term SA/DP at follow-up (OR 1.53; 95% CI 1.49–1.56). Those with the combination of high demands and high control were more likely to have emigrated (OR 1.67; 95% CI 1.59–1.76). Among women, both those in jobs with high demands and low control, and those in jobs with the combination of low demands and medium control were more likely to be on old-age pension (OR 1.91; 95% CI 1.82–2.00 and OR 1.20; 95% CI 1.14–1.26, respectively), whereas among men, both these groups were less likely to be on old-age pension (OR 0.59; 95% CI 0.25–0.66 and OR 0.66; 95% CI 0.62–0.69, respectively).
Supplementary Table 1, http://links.lww.com/JOM/A725 shows that both partial long-term SA/DP and partial long-term unemployment in 2001 were strongly associated with a higher risk of long-term SA/DP, long-term unemployment, and death in 2012 (Suppl. Table 1, http://links.lww.com/JOM/A725). Those born outside Sweden were more likely to have emigrated. Those who were married/cohabitants without children at home in 2001 were more likely to be on old-age pension in 2012, even after adjustment for age group.
Summary of Results
In this population-based prospective cohort study, covering all 2.2 million people in Sweden aged 30 to 54 and in paid work in 2001, we found that the majority (85%) were still mainly in paid work eleven years later, in 2012.
Those in jobs with the combination of low demands and low control in 2001 were more likely to be on long-term SA/DP at follow-up, and those with the combination of high demands and high control were more likely to have emigrated. There were some sex differences in the association between job demands/control and early old-age pension.
We also found a few situations where combinations of job demands/job control were differently associated with labor market situations among women than men, for example, women with the combinations of high demands and low control and low demands and medium control were more likely to be on old-age pension eleven years later, whereas men with the combinations of high demands and low control and low demands and medium control were less likely to be on old-age pension.
Among the individuals on long-term SA/DP in 2001, one-third (33%) were also on long-term SA/DP 11 years later while only about half of them were in paid work. Having long-term SA/DP or unemployment in 2001 was the factor with the strongest association with labor market situation at follow-up in our analysis.
Discussion of Results
We found that those initially in jobs with low demands and low control were more likely to be on long-term SA/DP at follow-up. In our previous study of cross-sectional associations between job demands/job control in the same study population, we found that low demands were associated with a higher risk of long-term SA/DP for all levels of job control.25 Our current results support this also regarding future labor market situation; however, the association was weaker in this prospective cohort study.
In the previous cross-sectional study, we also found that high demands and high control were associated with lower risk of SA/DP and of unemployment among both women and men, although more so for men.25 These results are also in line with the findings from the current study. Those with high demands and high control in 2001 were, however, slightly more likely to be on old-age pension in 2012, but this was not statistically significant for men. This is contrary to findings in three previous studies; that people in jobs with high control often retire later.6–8 That our results differ from those studies might be due to our longer follow-up period, the division of demands and control into tertiles, rather than dividing at the median, or that our study considers the entire population in paid work in Sweden, rather than a sample24,32 or people from specific occupations.16–20,33
Contrary to the theory that high demands and low control is the combination of demands and control that is the most adverse for health, we found that the highest risk for SA/DP was among those with low demands and low control. This may be due to either a health selection into jobs with low demands by those with poorer work ability, or possibly that those who have low demands and low control experience under-stimulation. This latter hypothesis is supported by findings that individuals who report low demands and low control have a higher risk of cognitive decline and dementia,34,35 myocardial infarction,36 depression and anxiety,37 as well as general SA38 and SA due to depression.39 A review of studies regarding the association between job demands/control and DP found that both high demands/low control and low demands/low control were associated with a higher risk of DP in several studies,4 leading the authors to speculate that level of job control is more strongly associated with the risk of DP than level of job demand. Our findings that the risk of SA/DP was higher for all groups with low control support this hypothesis.
The gender differences we found show the need for not only adjusting analyses for sex, but also for stratifying by sex, as done here. The reasons for this gender difference also need to be investigated further, potentially in relation to marital status (individuals tend to retire when their partner does, and women tend to marry older men) or in relation to physical job demands or types of occupation held by women and men.
Strengths and Limitations
The main strengths of this study are that it covers the entire working population rather than a specific occupation or a sample, the use of high-quality register data,28,29 administrative data rather than self-reports, the prospective cohort design without loss to follow-up, the long follow-up period (11 years) compared to previous studies, and the large study population, allowing several sub-group analyses. That several different possible labor-market outcomes could be used, not only one, is also an important strength. Other strengths are that all in paid work for part or all of the inclusion year were included, also those who had some unemployment and/or SA/DP, which means that our results are less affected by selection bias. Another strength is that we used a JEM22 to measure job demands and job control, meaning that the analyses were less vulnerable to reporting bias and reverse causation from those with morbidity, SA/DP, or unemployment estimating their job demands and job control differently from those without. A further strength is that the employment rates in Sweden were and are high for both women (73.9%) and men (77.5%)30 as well as in higher ages, that is, up to 65 years31—this means that our findings are less affected by health selection effects into the labor force. It is also strength that we could use JEM divided in tertiles instead of by the median, as this allowed more nuanced combinations of demands and control.
There are also certain limitations to this study. One is that we only had information on SA reimbursed by the Social Insurance Agency, which means that most SA spells ≤14 days were not included—however, since we consider only SA/DP >183 net days in a year, that will likely not have impacted our results. Another limitation is that the information on occupation in LISA is not updated every year for all individuals; thus, some individuals might have had another occupation in 2001 than the one registered, or that we do not include information on job demands/control in the intervening years between exposure and outcome.
The use of JEM, while it is a strength of the study, also involves certain limitations, namely that the values of demands and control for each individual are based on the average value for an individual of the same gender, age group and occupation, not his or her actual values. This may lead to an under- or overestimation of associations between job demands/control and labor market situations.
Our study was focused on a specific conceptualization of the psychosocial work environment, the combination of job demands and job control. It is possible that other psychosocial work environment exposures or exposures to physical and chemical factors at work that we did not measure in our study are equally or more important for labor market participation than job demands and job control.
Combinations of job demands and job control in 2001 were associated with labor market situations in 2012. Low demands/low control was associated with risk of SA/DP at follow-up among both women and men. Further, high demands/low control was associated with a higher likelihood of old-age pension among women and with a lower likelihood among men.
1. de Vries H, Fishta A, Weikert B, Rodriguez Sanchez A, Wegewitz U. Determinants of sickness absence and return to work among employees with common mental disorders: a scoping review. J Occup Rehabil
2. White M, Wagner S, Schultz IZ, et al. Modifiable workplace risk factors contributing to workplace absence across health conditions: a stakeholder-centered best-evidence synthesis of systematic reviews. Work
3. Williams-Whitt K, White MI, Wagner SL, et al. Job demand and control interventions: a stakeholder-centered best-evidence synthesis of systematic reviews on workplace disability. Int J Occup Environ Med
4. Knardahl S, Johannessen HA, Sterud T, et al. The contribution from psychological, social, and organizational work factors to risk of disability retirement: a systematic review with meta-analyses. BMC Public Health
2017; 17:176doi: 10.1186/s12889-017-4059-4.
5. Fisher GG, Chaffee DS, Sonnega A. Retirement timing: a review and recommendations for future research. Work Aging Retirement
6. Carr E, Hagger-Johnson G, Head J, et al. Working conditions as predictors of retirement intentions and exit from paid employment: a 10-year follow-up of the English Longitudinal Study of Ageing. Eur J Ageing
7. Virtanen M, Oksanen T, Batty GD, et al. Extending employment beyond the pensionable age: a cohort study of the influence of chronic diseases, health risk factors, and working conditions. PLoS One
8. Wahrendorf M, Akinwale B, Landy R, Matthews K, Blane D. Who in Europe works beyond the state pension age and under which conditions? Results from SHARE. J Popul Ageing
9. Robroek SJ, Schuring M, Croezen S, Stattin M, Burdorf A. Poor health, unhealthy behaviors, and unfavorable work characteristics influence pathways of exit from paid employment among older workers in Europe: a four year follow-up study. Scand J Work Environ Health
10. Browne P, Carr E, Fleischmann M, Xue B, Stansfeld SA. The relationship between workplace psychosocial environment and retirement intentions and actual retirement: a systematic review. Eur J Ageing
11. de Lange AH, Taris TW, Kompier MA, Houtman IL, Bongers PM. The very best of the millennium: longitudinal research and the demand-control-(support) model. J Occup Health Psychol
12. Van der Doef M, Maes S. The Job Demand-Control (-Support) Model and psychological well-being: a review of 20 years of empirical research. Work Stress
13. Van der Doef M, Maes S. The job demand-control(-support) model and physical health outcomes: a review of the strain and buffer hypotheses. Psychol Health
14. Jensen LD, Ryom PK, Christensen MV, Andersen JH. Differences in risk factors for voluntary early retirement and disability pension
: a 15-year follow-up in a cohort of nurses’ aides. BMJ Open
15. Thielen K, Nygaard E, Andersen I, Diderichsen F. Employment consequences of depressive symptoms and work demands individually and combined. Eur J Public Health
16. Virtanen M, Vahtera J, Pentti J, Honkonen T, Elovainio, Kivimäki M. The influence of job strain on sickness absence among 7986 Finnish employees with psychological distress. Eur J Public Health
17. Bourbonnais R, Mondor M. Job strain and sickness absence among nurses in the Province of Quebec. Am J Ind Med
18. Melchior M, Niedhammer I, Berkman LF, Goldberg M. Do psychosocial work factors and social relations exert independent effects on sickness absence? A six year prospective study of the GAZEL cohort. J Epidemiol Community Health
19. North FM, Syme SL, Feeney A, Shipley M, Marmot M. Psychosocial work environment and sickness absence among British civil servants: The Whitehall II Study. Am J Public Health
20. Niedhammer I, Bugel I, Goldberg M, Leclerc A, Gueguen A. Psychosocial factors at work and sickness absence in the Gazel cohort: a prospective study. Occup Environ Med
21. Baumberg B. Fit-for-work—or work fit for disabled people? The role of changing job demands
and control in incapacity claims. J Soc Pol
22. Fredlund P, Hallqvist J, Diderichsen F. Psychosocial Job Exposure Matrix. An Updated Version of a Classification System for Work-related Psychosocial Exposure. Stockholm: Swedish National Institute for Working Life; 2000.
23. Johnson J, Stewart W, Fredlund P, Hall E, Theorell T. Psychosocial Job Exposure Matrix. Stress Research Report Nr 221. Stockholm: National Institute for Psychosocial Factors and Health and WHO Psychosocial Centre; 1990.
24. Madsen IEH, Gupta N, Budtz-Jørgensen E, et al. Physical work demands and psychosocial working conditions as predictors of musculoskeletal pain: a cohort study comparing self-reported and job exposure matrix measurements. Occup Env Med
25. Norberg J, Alexanderson K, Framke E, Rugulies R, Farrants K. Job demands
and control and sickness absence, disability pension
, and unemployment
among 2,194,692 individuals in Sweden. Scand J Public Health
26. Sickness, Disability and Work: Breaking the Barriers: OECD; 2010.
27. Sickness, Disability and Work: Breaking the Barriers: Sweden: OECD; 2010.
28. Ludvigsson JF, Almqvist C, Bonamy AK, et al. Registers of the Swedish total population and their use in medical research. Eur J Epidemiol
29. Ludvigsson JF, Andersson E, Ekbom A, et al. External review and validation of the Swedish national inpatient register. BMC Publ Health
30. Arbetskraftsundersökningarna 2001 [Labour Force Surveys 2001]. Stockholm: Statistics Sweden; 2002.
31. Trends in Retirement and in Working at Older Ages. Pensions at a Glance 2011. Paris: OECD; 2011.
32. Suominen S, Vahtera J, Korkeila K, Helenius H, Kivimaki M, Koskenvuo M. Job strain, life events, and sickness absence: a longitudinal cohort study in a random population sample. J Occup Environ Med
33. Laine S, Gimeno D, Virtanen M, et al. Job strain as a predictor of disability pension
: the Finnish Public Sector Study. J Epidemiol Community Health
34. Pan KY, Xu W, Mangialasche F, Dekhtyar S, Fratiglioni L, Wang HX. Working life psychosocial conditions in relation to late-life cognitive decline: a population-based cohort study. J Alzheimer's Dis
35. Wang H-X, Wahlberg M, Karp A, Winblad B, Fratiglioni L. Psychosocial stress at work is associated with increased dementia risk in late life. Alzheimers Dement
36. Malinauskiene V. Psychosocial factors at work and myocardial infarction among men in Kaunas, Lithuania. Scand J Public Health
37. D'Souza RM, Strazdins L, Clements MS, Broom DH, Parslow R, Rodgers B. The health effects of jobs: status, working conditions, or both? Aust N Z J Public Health
38. Lidwall U, Bergendorff S, Voss M, Marklund S. Long-term sickness absence: changes in risk factors and the population at risk. Int J Occup Med Environ Health
39. Clumeck N, Kempenaers C, Godin I, et al. Working conditions predict incidence of long-term spells of sick leave
due to depression: results from the Belstress I prospective study. J Epidemiol Community Health