Depressive symptoms and physical activity among community-dwelling perimenopausal women: a prospective longitudinal study

BackgroundWomen in perimenopause are vulnerable to depressive symptoms, and physical activity was reported to be a potential protective factor. The trajectories of physical activity and depressive symptoms over time and their longitudinal relationships in Chinese perimenopausal women have not been explored yet, leaving a research gap hindering us from better understanding and managing perimenopause depressive symptoms.MethodsA multi-center prospective longitudinal study was conducted in four cities in Sichuan Province, China. Depressive symptoms and physical activity in perimenopausal women were collected in March 2019, June 2019, September 2019, and December 2019, respectively. Multivariable linear regression by generalized estimation equation was used to identify the relevant factors associated with depressive symptoms and physical activity. A four-wave autoregressive and cross-lagged panel model was performed to explore their longitudinal relationships.ResultsA total of 1875 women who completed the four-wave data collection were included in the data analysis. Depressive symptoms exacerbated over time and were associated with women's age, monthly income, marital status, chronic disease, and negative life events. Physical activity decreased over time and was associated with educational background and monthly income. According to the cross-lagged panel model, perimenopausal women with more severe depressive symptoms tended to be less physically active, and similarly, perimenopausal women with less physical activity were more prone to report more severe depressive symptoms.ConclusionThe cross-lagged panel model disclosed longitudinal bidirectional relationships between depressive symptoms and physical activity in perimenopausal women. Appropriate physical activity should be recommended for perimenopausal women to improve their mental well-being. Tailored physical activity duration and maintenance measures should be proposed based on different sociodemographic statuses.