Abstract

Objectives: To disentangle the direction between health-related behaviours (HRBs) and depression in East Asia, we examined the bidirectional longitudinal relationship of HRBs clustering and depression in Chinese and Japanese ageing populations gender-specifically.

Methods: Data were drawn from the China Health and Retirement Longitudinal Study (CHARLS) (N = 18,739) (2011-2015) and the Japan Study of Ageing and Retirement (JSTAR) (N = 7116) (2007-2011). Cross-lagged panel models (CLPM) were applied to assess bidirectional longitudinal relationships between HRB clustering and depression in adjacent waves, stratified by cohort and gender. Sensitivity analyses were conducted to explore key risk factors in risky clusters.

Results: The prevalence of baseline depression was approximately 20 % in China and 4 % in Japan, with over 70 % of participants belonging to the risky cluster. A unidirectional negative association between the risky cluster and depression was observed only in Chinese females (βstandardized = -0.039, 95%CI: -0.054 to -0.024, P < 0.001). However, no association was found in Japanese females or males in either country. Further sensitivity analyses revealed that this association was primarily driven by overweight/obesity cluster. Sensitivity analyses also identified a positive effect of depression on the smoking and drinking cluster and a negative effect on the overweight/obesity cluster in Chinese females.

Limitation: This study was limited by reliance on self-report data, the CES-D cannot be used to diagnose depression and some other confounding factors were not considered (e.g., use of antidepressants).

Conclusions: Our findings highlight a unidirectional negative association of the risky cluster on depression in Chinese females, with notable variations observed across different countries and genders.