4. Labor market participation

The effects of stress on disease may also result in reduced labor market participation through sickness absence and disability pension. Also sickness presenteeism (being at work while ill) may be one form of non-participation in terms of ability to perform. In this section, however, the focus is also on whether sickness absence (or presenteeism) predicts further health related outcomes, such as disability pension or mortality.


4.1 Stress related diagnoses and future sickness absence

In several prospective collaborative (I+S & I+E) population-based cohort studies we found that sickness absence due to stress related diagnoses was related to all cause and specific premature mortality and to suicidal behavior. We also found this for sickness absence due to mental diagnoses in general. A particular strength in these, as well as other studies, is that we could control also for previous morbidity. In a study of police officers, we found that sickness presenteeism was affected the most by the work environment in otherwise healthy subjects (E+I). In addition, we have shown that fatigue and sleep problems predict later long-term sickness absence. Moreover, insomnia, usually attributed to exposure to stress, is associated with all-cause and diagnosis-specific sickness absence and disability pension, as is self-reported sleep problems.

4.2 Effects of sickness presenteeism

Another effort of the center (I+W) has been to determine whether self-reported sickness presenteeism (SP) and self-reported sickness absence (SA), regardless of cause, are specific risk factors for future health. A cohort of 2181 employees aged 25 – 50 years in 2004 with data for 2005 and 2006 was studied. The results suggest that both SP and SA are strong predictors of future poor health, physical complaints, low mental well-being and low work ability. This prospective association was replicated (E+I+W) in a sample of 7445 workers from the SLOSH cohort. When comparing SP and SA regarding the strength of the associations with health/ill- health (five data sets, (n=425-3,622) certain groups in the labor market seem to have poorer health than the measure of SA would indicate. However, we (E+I) also found that SP, far from being an alternative to SA, was strongly correlated with the latter, also after adjustment for a large number of health and work environment factors. Fixed-term employees reported lower levels of working while ill than permanent staff, whereas subjective job insecurity was associated with higher levels, indicating that fear of job loss may motivate people to work despite illness.

4.3 Effects of sickness absence

In several prospective population-based cohort studies we found that sickness absence (disregarding the specific diagnosis) is a risk in- dicator for suicidal behaviour and all- and cause-specific premature death, also when controlling for morbidity and sociodemographics. We also found this for disability pension due to some diagnoses. In particular, in 26,000 young individuals granted disability pension due to either mental or somatic diagnoses in Sweden, we showed that the risk of suicide attempt increased continuously up to the year preceding the granting of disability pension, after which the risk declined. In a prospective study of 45,000 Swedish twins with 13 years of follow-up, sickness absence was a risk factor for disability pension and premature death, in both women and men. Those granted disability pension in 1995 had a higher risk for premature death, with a slight indication of familial influences in both sexes. A prospective study of 46,000 twins showed that genetic factors were of importance for the liability to disability pension due to different diagnoses. Based on stress theories, we asked 5,800 persons on long-term sickness absence about their encounters with health or insurance professionals, and studied the effects of these experiences on emotions and return to work.

4.4 Effects of disability pension

In a prospective cohort study of >500,000 young adults (I+E), disability pension due to either somatic or mental diagnoses (including stress-related mental disorders) was associated with a higher risk of suicide attempt and suicide, also when adjusting for the effects of so- cio-economic status, previous suicide attempts, and parental risk factors (SES, psychiatric health care and suicide attempt/suicide).

4.5 Retirement

One obvious cause of non-participation in working life is retirement. This may be one the potential stressors of working life. Both having to leave work earlier than desired (e.g. for health or labor market reasons) and having to work longer (for financial reasons possibly related to policies aiming to extend working life) could be stressful for the individual. In a series of groundbreaking collaborative (E+I) papers we have shown that retirement for many people, especially those with poor health and poor working conditions, is associated with substantial improvements in self-rated health, fatigue and mental health, headaches, health functioning as well as increased physical activity. However, there were no immediate changes in the risk of incident chronic disease at retirement and there was a worrying increase in non-adherence with preventive medication, as well as increases in heavy drinking, which was temporary in most, but not all, groups. These results are based on data from France, Finland, and the UK, with new results supporting the overall picture currently emerging from Swedish and partly US data. In a study based on the US Health and Retirement Survey, however, physical functioning was shown to decline faster in retirement than in full-time work in employees aged 65 years or older, a difference not explained by chronic diseases or lifestyle. To understand the changes in exposures and behaviors that could underlie these apparent health effects of retirement, we are currently starting a unique study with in-depth data from questionnaires, diaries, actigraphy recordings and other intense methods of data collection, where people planning to retire are followed during one week, six months prior to retirement, and for another week six months after, and finally a third week yet another year later. In other joint ongoing or planned studies we follow people before and after retirement with regard to healthcare and pharmacological consumption, sick leave, morbidity and mortality using our nationwide register of all people in Sweden. Note that the work on retirement has obvious links to issues of aging, covered elsewhere in this report.

4.6 Encounters

Based in stress theories we studied long-term 5,800 sickness absentees’ experiences of their encounters with health or insurance profes- sionals and effects on emotions and return to work.


Planned research

Sick leave is common in Sweden as in many other welfare states and there is a need for further knowledge on both risk factors for sick leave or disability pension as well as for on different types of consequences of being sickness absent for short or long periods, with specific conditions. The latter there is hardly any knowledge on. The different and large databases we have on this within SSC will be updated and managed and several studies are ongoing and planned. SSC funding will mainly be used for coordination and facilitating collaboration to make large data set available and to analyze major research questions.

4.1 Psychosocial factors and sickness absence

In a series of collaborative studies (I+E) we are going to continue and extend previous work and investigate sickness absence data for specific groups of individuals (related to type of occupation, type of condition, age and sex), using new different types of analytical methods to gain knowledge on this over longer times and in some studies even over life.

4.2 Effects of sickness absence and disability pension due to stress related disorders

Of special interest is our future studies on long-term outcomes of sickness absence and disability pension due to stress-related mental disorders. In those collaborative (I+S) studies, as well as in other studies we will be able to account for effects of patient characteristics (age, sex, education, birth country, (co)morbidity, of working environment, health care, pharmacological treatment, and parental factors) – and annual changes in this with long follow ups.

4.3 Burnout and mental disorder

In collaborative (I+S) twin studies of stress and burnout, focus will be on understanding the complex relationships between burnout and mental disorders as well as further investigation of environmental stressors, accounting for genetic aspects. Further, studies will focus on investigating the pathways to sick leave and disability pension.

4.4 Life events

Associations between stressful life events and future morbidity, life situation, sick leave, disability or old-age pension, and premature death will be studied in several collaborative (I+E) projects with 2-20 years of follow up, to understand different mechanisms, in general and from a gender perspective. Examples of such life events are: death of own child (all cause and due to suicide), immigrating, giving birth, diagnosed with a severe disorders, long-term unemployment. Both register (All-Sweden data base; Swedish twin register) and survey data will be used.

4.5 Encounters with the insurance system

Sickness absentees’ experiences of encounters with healthcare and social insurance professionals will be studied in several large projects, using different theories about mechanisms for effects, including stress-related theories and theories on social emotions. We will also study this for specific diagnoses (for example cancer) and for possible changes over time. Effects of interventions to promote positive encounters will be studied. Also, people who have had their claims for continuous sickness benefits denied will be studied in two large projects.

4.6 Presenteeism

We will also extend our collaborative (I+E and I+W) studies of sickness presenteeism, in different occupational groups and diagnoses, using several of the different datasets available in the center.

4.7 Aging

Rates of people working after 65 years of age are increasing and more knowledge is warranted about effects of this. We will investigate sickness absence and morbidity among those working after 65, using data from four different prospective cohorts (all people in Sweden above 65 years in 1995, 2000, 2005, and 2010, respectively), and compare to this to sick leave, disability pension and health among them in the five preceding years – in different socioeconomic groups and occupations. This project will be coordinated with the other Center projects on aging and retirement.

4.8 Retirement

A new nationally representative prospective cohort study of work, health and retirement among persons aged 50+ with frequent repeat measures is planned to be launched within the next few years by E in collaboration with the other SSC partners (all should enter their own questions subjective health, sleep, etc). The study will be based on the model provided by the US Health and Retirement Study (HRS) and the English Longitudinal Study of Ageing (ELSA), but include substantial improvements based on the experience in those and related studies, as well as specific adaptations to the Swedish context and a clear aim to facilitate the application of the latest ad- vances in longitudinal and life course modelling.

4.9 Retirement preferences and bridge employment

We will continue our research on retirement preferences, the timing of retirement and the engagement in bridge employment among senior individuals. Some of the planned research will also in the future involve collaboration (W+E+I), but also include new collab- orators from countries such as Belgium and Portugal. These planned research efforts aim at identifying individual characteristics and structural factors that could either promote successful mobility patterns or complicate and hinder mobility.

4.10 Health consequences of prolonged working life in physically demanding jobs

We have applied for funding from Forte (2014-2488) to investigate the relationship between physically demanding work, retirement and health. This will be done partially by secondary analyses in large epidemiological cohorts, partially by an in-depth study with nested case-control design. In the latter study. 160 older workers in physically demanding jobs (both male and female dominated) will be fol- lowed with intense measurements of physical activity, postures and other exposures under three 7-day periods, 6 months before retirement (of the retiring cases), 6 months after, and finally after another 12-month period. The participants will be selected in such a way that 80 persons (40 men and 40 women) plan to retire, and 80 plan to work at least 2 more years. The study aims to investigate, among other things, how physical demands at work influences timing of retirement and how people’s lives are affected when they retire, how these changes differ from the corresponding changes in people who retire from less physically demanding jobs, and how retirement and (changes in) health behaviors influence health and quality of life in people retiring from physically demanding jobs. This research will be carried out by E and I in collaboration with the Danish National Research Centre for the Working Environment (A Holtermann) and Jönköping University (E Fransson).