Background: Physical activity (PA) is believed to improve mental health, including depression. However, whether recommended PA levels have a similar impact in individuals with poststroke depression is unclear.
Objective: The aim of this study was to apply a quasi-experimental propensity score (PS) matching control for covariate differences and compare the effects of PA on reducing depression risk among people with stroke.
Design: A cross-sectional design was used for this study.
Methods: Health-related information for community-dwelling adults (N=4,555) who reported having had a stroke was extracted from the 2013 Korean Community Health Survey data set. The survey participants were asked a series of questions about depression, recommended PA levels (moderate and vigorous intensity), and chronic conditions. A multivariable regression model, inverse probability weighting adjustment, and Greedy algorithms with 1:1 matching and covariate adjustment were used to estimate the effects of PA on risk of depression. The dependent variable was diagnosis of depression, and the primary independent variable was PA. Baseline covariates were 10 demographic and 9 chronic condition variables.
Results: Without PS methods, there were significant differences in baseline covariates (16 out of 19) between people who performed PA and those who did not perform PA. After applying 1:1 matching, the number of patients in each group comparing the effect of PA numbered 1,970, and 13 covariates did not differ significantly between the 2 groups. Physical activity reduced the risk of poststroke depression by 36.1% to 42.4% (odds ratio=0.639-0.376) across the 3 methods.
Limitations: Unaccounted-for covariates, including stroke severity, predepression status, and history of depression treatments, may have biased the results.
Conclusions: The findings suggest that recommended PA levels have protective relationships with the risk of poststroke depression.
© 2017 American Physical Therapy Association