Prospective cohort studies of disability pension and mortality in a Swedish county

Abstract

The overall aim of this thesis was to study risk factors for disability pension, focusing on sociodemographic risk factors and sick-leave diagnoses, and also to investigate the risk of mortality among disability pension recipients. The research comprised four register studies of all people in the Östergötland County who were 16 64 years of age in December 1984, representing a total cohort of 245,704 persons. The first study was a 12-year follow-up based on the 229,864 individuals in the cohort who at baseline were not receiving disability pension or old-age pension. The second investigation constituted a 10-year follow-up of the 19,379 people aged 16-60 years who had a new period of long-term (≥ 56 days) sick-leave spell in 1985 1987. The third and fourth studies were 12-year follow-ups of, respectively, the entire cohort, and all individuals in the cohort except those receiving a medically-based disability pension and missing disability pension diagnosis.

The likelihood of being granted a disability pension within 12 years was 14 % for women and 11 % for men, and this increased with age. Women younger than 54 years were at higher risk than men (P < 0.001). The disability pensions were granted for full-time in 69% of the cases and more women received part-time pensions.

Five years after inclusion, 28% of the cohort with long-term sick leave had been granted disability pension. The risk of such pension was greater for those who were older, had a low income, previous sick leave, no employment, no young children, or were born in a country other than Sweden. There was a gender-related difference in sick-leave diagnoses (P < 0.001): among the men, those with psychiatric diagnoses were at highest risk; among women, the risk was greatest for those with musculoskeletal diagnoses. This overall pattern of disability pension predictors was still apparent, albeit attenuated 5 10 years after inclusion. However, this was not true for income, the effect of which was reversed over time.

The relative risk (RR) of all-cause mortality was higher for disability pension recipients than for individuals not collecting these benefits, and this applied to both women (RR 2.8, 95% confidence interval [CI]: 2.6 3.0) and men (RR 3.0, 95% CI: 2.8 3.1), as well as all age groups. The relative risk of mortality was highest for the youngest disability pensioners, and it was lower for those with part-time benefits than for those collecting full pension.

All groups receiving medically-based disability pension showed increased mortality due to all main causes of death both directly and not directly associated with the disability pension diagnoses. The magnitude of this increase varied depending on the diagnosis legitimating disability pension. There was a strong relationship between disability pension and future suicide among those granted disability pension due to psychiatric disorders (hazard ratio [HR] =12.0, 95% CI: 8.4 17.2 for women; and corresponding values 5.5, 95% CI: 4.0 7.6 for men), and the same was seen for men with cardiovascular diagnoses (HR = 2.4, 95% CI: 1.3 4.4). No increase in mortality due to any cause was found in the subjects over 55 years of age who were granted disability pension due to labour-market reasons.