Unfavourable sedentary and physical activity behaviour before and after retirement: a population-based cohort study

BACKGROUND: During transition to retirement there is often a rearrangement of daily life which might provide a key opportunity for interventions to promote a non-sedentary and active lifestyle. To be able to design effective interventions, it is essential to know which sedentary and physical behaviour domains (eg, at home or during leisure time) have potential to facilitate healthy ageing during the retirement transition. OBJECTIVE: To determine whether unfavourable sedentary and physical activity behaviour before retirement predict unfavourable sedentary and physical activity behaviour after retirement. DESIGN: Population-based cohort. SETTING AND PARTICIPANTS: Adults (n=3272) employed in 2010 but retired in 2014. METHODS: Self-reported preretirement job activity, sedentary leisure time, physical activity at home, and walking-cycling and exercise were assessed as predictors for unfavourable sedentary and physical activity behaviours after retirement using logistic regression. Unfavourable behaviours were defined based on the respective median of the cohort distribution. Furthermore, the OR for having multiple unfavourable behaviours after retirement was determined, based on the amount of unfavourable behaviours before retirement. All models were adjusted for gender and education. RESULTS: Unfavourable preretirement physical activity and sedentary behaviour at home or during leisure time were the strongest predictors of the same behaviours after retirement. Unfavourable job activity did not predict physical activity but did predict unfavourable sedentary behaviour after retirement (OR=1.66, 95% CI 1.41 to 1.96). Unfavourable exercise behaviour before retirement predicted unfavourable sedentary and physical activity after retirement in all domains. With all behaviours being unfavourable before retirement, the OR of having at least three unfavourable behaviours after retirement was 36.7 (95% CI 16.8 to 80.5). CONCLUSIONS: Adults with a higher number of unfavourable preretirement physical activity and sedentary behaviours are likely to carry these unfavourable behaviours into retirement age. Interventions should target those with more unfavourable preretirement physical activity and sedentary behaviours before retirement, and those interventions focusing on exercise might have greatest potential.