This thesis aimed to examine changes in physical activity and body mass index (BMI) during working years and changes in physical activity concurrently with changes in BMI, physical functioning, and self-rated health during retirement transition. It examined two cohorts of public sector workers, from the Finnish Public Sector study (n=66,852) and the Finnish Retirement and Aging study (n=3,550). The data were collected using a repeated survey. Group-based multi-trajectory analysis was the main statistical method used.

During a 16-year follow-up in this working-age population, physical activity decreased and BMI increased, regardless of age or sex. The differences in the four identified groups were mainly related to their initial level of physical activity and BMI.

During the retirement transition, physical activity temporarily increased, independently of the pre-retirement activity level. BMI remained stable. BMI and physical activity were inversely connected: low activity was associated with a high BMI. During the retirement transition, lower physical activity was associated with a lower level of physical functioning and poorer self-rated health. Physical functioning and self-rated health improved along with physical activity during the transition, but the change was only temporary. Female sex, manual occupation and living alone were associated with lower physical activity, higher BMI, poorer physical functioning, and suboptimal self-rated health during the transition to retirement.

In conclusion, physical activity, physical functioning, and self-rated health temporarily increased during the retirement transition, but BMI remained stable. The changes in physical activity, physical functioning and self-rated health were interconnected. The results suggest that retirement is timepoint-sensitive to changes in lifestyle habits and that physical activity should be promoted among those about to retire, especially among those at risk of low physical activity after retirement.