Patterns Of Physical Activity And Cognitive Performance Over Time In Midlife Women In SWAN

Physical Activity and Cognitive Performance over Time in Midlife Women in SWANObjective: To determine how increase in physical activity in midlife affects cognitive performance with aging in the Study of Women’s Health Across the Nation (SWAN) diverse cohort. METHODS: We converted Kaiser Permanente Physical Activity Scale scores to MET hours per week to provide a meaningful interpretation for clinicians. Since physical activity (PA) changed minimally in the first 6 years of observation, we used the average MET hours from visits 0 through 6 as baseline PA. At visits 9, 12, 13, and 15 (visits with both cognitive and PA assessments), we calculated the difference between the MET hours at the time of visit and baseline average. We eliminated the first two cognitive assessments for each participant to minimize practice effects. Longitudinal associations were analyzed with mixed linear regression models using the z-scores of processing speed (Symbol Digit Modalities Test, SDMT) and working memory (Digit Span Backwards, DSB); episodic memory (East Boston Memory Test- delayed recall, EBMT-D) was reverse coded and analyzed using Poisson regression. RESULTS: Our sample consisted of racially diverse midlife women (N = 2020) with 25.2% Black, 9.9% Chinese, 10.8% Japanese, 4.3% Hispanic, and 49.9% non-Hispanic White. At visit 9, mean (SD) age 59.4 (3.2) and mean (SD) MET hours/week was 6.8 (2.4). For every 1 MET hour/week higher at baseline, processing speed (SDMT) was 0.010 units higher, even when adjusted for sociodemographic, medical, and emotional-behavioral factors. An increase of 1 MET hour per week from baseline was significantly associated with 0.003 units higher SDMT z-score in fully adjusted models. There were no significant associations of PA or change in PA with DSB or EBMT-D. CONCLUSION: Our study showed a positive impact on cognitive processing speed in women who had higher PA and increase PA during their midlife. Processing speed is a more sensitive predictor of early cognitive decline than memory tests, which may explain the significance of the association in this relatively young study population. Increasing physical activity in midlife may delay decline in processing speed as women age.SWAN has multiple grant support from NIA, NINR and ORWH. Abstract does not necessarily reflect the official views of the NIH. Supported by: The Study of Women's Health Across the Nation has multiple NIA, NINR, and ORWH grants from the NIH