Several studies have found insomnia and long sleep duration to be independently associated with subsequent disability pension. However, the issue of a possible gender-based pattern in this context has received little attention. The aim of the present study was to assess the impact of different sleeping problems and sleep duration on the rate of disability pension during a 12-year follow-up period.

The cohort was recruited by random invitations from the general population of Malmö, Sweden, aged 45–65 years in 1992; the participation rate was 41% (n=14,555). The participants in this study were the 2,254 men and 2,065 women who were healthy and working at least 30 hours per week at baseline. Baseline data included socio-demographic factors, lifestyle, body mass index, the Job Content Questionnaire, social support and participation, ‘stress outside work’, self-rated health and an instrument assessing sleep quality and duration. Information on disability pension was obtained through record linkage from the National Health Insurance Register.

Disability pensions were granted to 9% of the men and 15% of the women during the follow-up period. Affirming moderate or very large problems with any of ‘initiating sleep’, ‘waking up during the night’, ‘waking up too early’, and ‘not feeling rested by sleep’ was defined as ‘problems with sleep’, and this was present in 33% of the women and 41% of the men. The age-adjusted hazard ratios for problems with sleep and subsequent disability pension were in men 1.8 (95% CI 1.4 to 2.4) and in women 1.6 (95% CI 1.3 to 2.0). In the full model, after adjustment for sleep duration and for potential confounders and/or mediators, these HRs decreased to 1.4 (95% CI 1.1 to 2.0) in men and 1.4 (95% CI 1.1 to 1.7) in women.

Short sleep duration, i.e. ≤6 hours/night on weekdays, was associated with having problems with sleep, long sleep duration was not. Only two percent of the population slept ≥9 hours/night. In women, but not in men, long sleep duration was strongly associated with the subsequent granting of a disability pension; the HR was 2.8 (1.7 to 4.6). In conclusion, sleeping problems seem to be a significant risk-factor for disability pension in the middle-aged working population. There were also clear gender differences in the pattern regarding which type of sleeping problems that were most linked to disability pension.