Physical activity trajectories and cognitive function: A national cohort study

To examine the prospective associations between physical activity trajectories, measured from repeated assessments over time, and cognitive function. A total of 2972 participants aged 45 years old and over (median age: 56.0 [interquartile range - IQR 50.0–62.0], 50.8% males]) from the Chinese Health and Retirement Longitudinal Study (CHARLS) study were included. First, our study used the growth mixture modeling to identify physical activity trajectories from the first three surveys of the CHARLS. Second, we performed regression analysis to explore the associations of the trajectories with 3-year cognitive function. We identified four physical activity trajectories, characterized by persistently low (N = 1880), initially low then increasing (154), initially moderate then decreasing (584), and initially high then decreasing (354). After 3-year follow-up, compared to individuals with persistently low trajectory, those with initially moderate then decreasing (β = −0.74, 95% CI = (−1.38, −1.10), p = 0.024) and initially high then decreasing (β = −1.12, 95% CI = (−1.91, −0.33), p = 0.005) were significantly associated with cognitive decline. The effects of the decline in physical activity on cognitive function varied by sex. Females’ cognitive function (β = −1.79, 95% CI = (−2.82, −0.77) was more likely to be influenced by decreasing physical activity, but this association was not significant in males. We did not find the significant effect of initially low then increasing trajectory on cognitive function. Fast decline in physical activity is related to a higher risk of cognitive decline, especially in females.