Bidirectional association between cognitive function and depressive symptoms in UK community elderly based on physical activity as mediator

Abstract

Introduction Previous studies have shown that physical activity (PA) can improve cognitive function (CF) and depressive symptoms in middle-aged and older adults, and that there is a bidirectional relationship between the two. However, there is a lack of research on the temporal sequence and potential mechanisms of these three factors. This study conducted a large-scale longitudinal study to explore the mediating role of PA in the bidirectional association between CF and depressive symptoms in middle-aged and older adults. Methods A longitudinal study was conducted using data from the English Longitudinal Study of Aging (ELSA) collected in 2014−2015 (T1), 2016−2017 (T2), and 2018−2019 (T3). PA, CF, and depressive symptoms were assessed via questionnaires, with Spearman’s correlation analysis used to examine variable correlations. A Cross-Lagged Panel Model (CLPM) was constructed to test the longitudinal causal relationships among the three variables, and a longitudinal mediation model was developed to analyze the mediating role of PA. Results A total of 6,787 participants aged 50 years or older were included, comprising 3,777 women (55.7%) and 3,010 men (44.3%), with a median baseline age of 66 years. Spearman correlation analysis indicated a significant correlation among the three variables (r s =[−0.225, 0.447], P < 0.001). CLPM results showed that poorer CF predicted more severe depressive symptoms in the future (β T1-T2  = −0.038, P < 0.001; β T2-T3  = −0.040, P < 0.001), and more severe depressive symptoms predicted poorer future CF (β T1-T2  = −0.022, P = 0.005; β T2-T3  = −0.021, P = 0.005). The Wald test indicated that CF was the dominant factor in the bidirectional negative relationship between CF and depressive symptoms (Wald χ² = 7.546, P = 0.006). Longitudinal mediation model results showed that poorer CF at T1 predicted lower PA at T2 (β T1-T2  = 0.043, P < 0.001), and lower PA at T2 predicted more severe depressive symptoms at T3 (β T2-T3  = −0.053, P < 0.001). Bootstrap results further confirmed the significant indirect effect of T1-CF on T3-depressive symptoms through T2-PA (β = −0.002, 95% CI=[−0.004, −0.001]), with the mediation effect accounting for 5.71%. However, the indirect effect of T1-CF on T3 depressive symptoms through T2-PA was not significant (β = −0.001, 95% CI=[−0.002, 0]). Conclusion CF was negatively correlated with depressive symptoms. PA mediated the positive effect of CF on depressive symptoms in middle-aged and older adults, but failed to alleviate the negative effect of depression on CF. This study indicates that active participation in PA contributes to maintaining cognitive and mental health in middle-aged and older adults.