LAUSR

A large number of studies have found job control to be consistently associated with an increased risk of disability pension. However, most previous studies have measured job control by self-report, introducing possible reporting bias inflating the risk estimates. Furthermore, previous studies have not accounted for the potential selection of individuals with pre-existing risk factors for disability pensioning into low control jobs. We analyzed data from the nationwide register-based Danish Work Life Course Cohort (DAWCO; n = 960,562 with approx. 6 million person-years). We measured job control annually by a job exposure matrix, based on a scale of five self-reported items from The Danish Work Environment Cohort Study, and disability pension using registers on public transfer payments. To account for potential selection into occupations with lower levels of job control, we included numerous life course confounders, including parental socioeconomic position and psychiatric and somatic diagnoses. Employees in jobs with lower levels of job control had increased risk of disability pensioning. The association attenuated after adjustment for confounders but was not explained by selection into job groups with lower levels of job control (hazard ratio: 1.16 (95% CI: 1.03-1.31). Our findings suggest that lower levels of job control are associated with an increased risk of disability pension, and that this association is explained by neither reporting bias nor a selection of individuals with an increased risk of disability pensioning into job groups with lower levels of job control. Lower levels of job control appear associated with an increased risk of disability pension independent of life course confounders. Further research is needed on preventive measures in occupations with low levels of job control.