Effect of physical activity intervention on late-life depressive symptoms: a target trial emulation using two cohort studies

Abstract

Aim: To estimate the effect of a sustained physical activity intervention on 12-year risk of late-life depressive symptoms among older adults.

Methods: We emulated a target trial of a sustained intervention of moderate to vigorous physical activity at least twice a week using data from the Health and Retirement Survey (HRS) and English Longitudinal Study of Ageing (ELSA). The HRS included 10,735 eligible participants (57.5 % women; 65.8y ± 9.4) and the ELSA included 4898 participants (54.0 % women, 66.3y ± 9.2). We adjusted for pre-baseline, baseline and time-varying risk factors using parametric g-formula to estimate the 12-year risk of depressive symptoms under the physical activity intervention vs under natural course.

Results: The estimated 12-year risk of developing depressive symptoms was 44.3 % (95%CI: 36.2; 51.1) under natural course and 39.1 % (32.5; 45.9) under the physical activity intervention, resulting in a risk difference of -5.20 percentage points (-7.40; -2.37) in the HRS. In the ELSA, the risk was 64.8 % (58.9; 73.4) under natural course vs 56.1 % (49.3; 64.5) under the physical activity intervention, resulting in a risk difference of -8.70 percentage points (-11.78; -6.14).

Conclusion: We estimated that older adults have experienced a lower incidence of depressive symptoms under a long-term sustained intervention of physical activity.

Keywords: Depression; Exercise; G-formula.