Abstract
This study investigated longitudinal physical activity (PA) profiles over 7 years in the Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability (FINGER). Cognition, depression, pain, and PA motives were included as determinants of the PA profiles. The 1259 participants, aged 60-77 years at baseline, were randomized into either a control group receiving general health advice, or an intervention group offered a comprehensive 2-year multidomain intervention including physical exercise, diet advice, cognitive training, and vascular risk factor management. The participants reported weekly moderate-intensity PA at baseline and 1, 2, 5, and 7 years after the baseline. Those providing PA data at two or more time points were included (n = 1188). Longitudinal PA profiles were determined using latent class growth analysis, and their associations with baseline determinants via multinomial logistic regression analysis. Interaction terms were added to investigate whether the intervention modified these associations. Six PA profiles were identified: Very high-stable (6%), High-stable (22%), Moderate-declining (47%), Moderate-steeply declining (5%), Low-increasing (9%), and Constantly low (12%). Participants in the intervention group and those motivated by distal and proximal benefits of exercise were likelier to maintain high PA level. Conversely, depressive symptoms and pain were predictors of Constantly low profile. Results show that high baseline PA was generally maintained, while greater variability in PA changes was observed among initially less active participants.