Objective: To assess leisure-time physical activity (LTPA) as a modifier of the diabetes/cognitive decline association in middle-aged and older participants in the Estudo Longitudinal de Saude do Adulto (ELSA-Brasil) study.
Research design and methods: ELSA-Brasil is a cohort of 15,105 participants (age 35-74 years) enrolled between 2008 and 2010. We evaluated global cognitive function, summing the scores of six standardized tests evaluating memory and verbal fluency, including the Trail-Making Test, at baseline and follow-up. Incident cognitive impairment was defined as a global cognitive function score at follow-up lower than -1 SD from baseline mean. Participants reporting ≥150 min/week of moderate to vigorous LTPA at baseline were classified as physically active. We assessed the association of LTPA with global cognition change in those with diabetes in the context of our overall sample through multivariable regression models.
Results: Participants' (N = 12,214) mean age at baseline was 51.4 (SD 8.8) years, and 55.5% were women. During a mean follow-up of 8.1 (SD 0.6) years, 9,345 (76.5%) inactive participants and 1,731 (14.1%) participants with diabetes at baseline experienced faster declines in global cognition than those who were active (β = -0.003, -0.004, and -0.002) and those without diabetes (β = -0.004, -0.005, and -0.003), respectively. Diabetes increased the risk of cognitive impairment (hazard ratio [HR] 1.71; 95% Cl 1.22, 2.39) in inactive but not in active adults (HR 1.18; 95% CI 0.73, 1.90). Among participants with diabetes, those who were active showed a delay of 2.73 (95% CI 0.94, 4.51) years in the onset of cognitive impairment.
Conclusions: In adults living with diabetes, LTPA attenuated the deleterious association between diabetes and cognitive function.
© 2024 by the American Diabetes Association.