The longitudinal association between physical health and depressive symptoms over eight years: Evidence from the health and retirement study

Background: The bidirectional relationship between physical health (PH) and depressive symptoms (DS) remains unclear.

Methods: Data were extracted from the Health and Retirement Study in the United States. PH was measured with a composite of chronic diseases, functional limitations and difficulties in basic and instrumental activities of daily living, and DS with a modified Center for Epidemiological Studies of Depression. Latent growth curve models (LGCM) were employed to examine how the change in PH or DS affected their mutual trajectories in later life. In addition, multilevel models were utilized.

Results: There were 6144 participants included, with an average age of 69.82 ± 6.85 years at baseline, of whom 3686 (59.99 %) were women. PH scores increased from 5.65 in 2010 to 7.72 in 2018, while depression scores increased from 1.14 to 1.31. LGCM results showed that the initial levels of PH and DS were associated (β = 0.558, P < .001), and the initial level of PH could predict the trajectory of DS (β = 0.089, P < .001). Likewise, the initial level of DS was also related to initial PH (β = -0.563, P < .001) but couldn't predict the trajectory of PH. Furthermore, the slopes of PH and DS were predicted bidirectionally by each other. Two-level logistic models further demonstrated the bidirectional association between PH and DS.

Conclusion: There was a bidirectional association between physical health and depressive symptoms, which highlights the necessity of comprehensive health management for older adults with poor physical health or depression symptoms.