Using cohort data to emulate lifestyle interventions: Long-term beneficial effects of initiating physical activity on cognitive decline and dementia

peer reviewedBackground: Intervention studies have shown beneficial short-term effects of physical activity on cognitive decline and reduced risk of dementia. However, randomized controlled trial data of lifestyle interventions over long time spans are not available due to lack of resources, feasibility or ethical reasons. Drawing from the principles of emulating a ‘target trial’, which apply design principles of randomized trials to the analysis of observational data, cohort data of a large European survey were analyzed to understand the long-term effects of physical activity changes. Method: Biennial assessments of the economic, social, and health situation of respondents aged 50 and older came from the Survey of Health, Ageing and Retirement in Europe (2004-2017). Cognitive functioning (immediate recall, delayed recall, and verbal fluency) and self-reported diagnosis of dementia were assessed at each follow-up. The target trial included sedentary respondents at t1 who, at follow-up (t2), stayed sedentary (“control group”) or newly reported vigorous physical activity more often than once a week (initiators, “treatment group”). Inclusion and exclusion criteria were implemented as close as possible to those of the FINGER trial. Inverse-probability weighting accounted for the probability of initiating physical activity with a large set of predictor variables. Selecting respondents aged 50-85 years old who met the target trial inclusion and exclusion criteria, assessments of cognitive functioning and self-reported diagnosis of dementia were available for 8,781 respondents at t3 (on average 3.02 years later), 3,858 respondents at t4 (5.84 years), and 2,304 respondents at t5 (7.72 years). A total of 304 respondents reported a diagnosis of dementia. Result: Initiators of vigorous physical activity had higher cognitive functioning at two follow-ups compared to non-initiators (t3: “average treatment effect on the treated”, ATET=0.059, CI: 0.028, 0.090), which remained significant after implementing inclusion and exclusion criteria. Initiators had lower risk of dementia compared to non-initiators at all three follow-ups (t3: ATET=-0.009, CI: -0.015, -0.005, relative risk decrease -46.7%), remaining significant after implementing inclusion and exclusion criteria. Conclusion: Emulating a target trial showed long-term benefits of initiating physical activity for cognitive functioning and dementia risk. Multidomain interventions related to nutrition, social, cognitive activities etc. can be similarly emulated