Single and additive effects of health behaviours on the risk for disability pensions among Swedish twins.

BACKGROUND This prospective cohort study aimed to investigate single and additive effects of health behaviours for risk of future disability pension (DP) due to musculoskeletal diagnoses, DP due to mental diagnoses, DP due to circulatory system diagnoses or DP due to other diagnoses. A secondary aim was to evaluate the role of familial confounding in these associations. METHODS A sample of 31,206 Swedish twin individuals was followed from 1998-2003 to the end of 2008 via national registries for DP supplemented with additional baseline questionnaire data. Cox proportional hazard ratios (HRs) were estimated. RESULTS The single effect of moderate to heavy alcohol consumption was associated with decreased risk of DP irrespective of diagnosis group in analyses accounting for background and familial factors (HR 0.42-0.88). Low frequency of physical activity (HR 1.17-1.63) or being past or regular user of tobacco products (HR 1.38-3.49) predicted increased risk for DP in all diagnosis groups. Interactions between physical activity and alcohol or tobacco consumption were significant for DP due to musculoskeletal diagnoses and mental diagnoses. CONCLUSION Health behaviours are important risk factors for DP irrespective of diagnosis groups. The effect of health behaviours is independent from familial effects (genetics and shared environment) pointing to a direct effect of health behaviours on DP, but education, body mass index, number and severity of diseases and marital status may confound the associations. The complex interrelationships between the health behaviours revealed by interactions suggest that prevention of future DP should attempt to influence these co-existent behaviours.