Joint trajectories of body mass index and depressive symptoms and incident liver disease: A multicohort study with frailty mediation analysis.

BACKGROUND: Obesity and depression are independently associated with liver disease, yet to our knowledge, no prior study has specifically examined their joint longitudinal trajectories or the mediating role of frailty in this association. OBJECTIVES: To identify joint trajectories of body mass index and depressive symptoms, examine their associations with incident liver disease, and evaluate frailty as a potential mediator. METHODS: This multicohort prospective study included 17,582 participants aged 45 years or more from the China Health and Retirement Longitudinal Study (2011-2015; n = 10,487) and the Korean Longitudinal Study of Aging (2014-2018; n = 7095). Joint body mass index-depressive symptom trajectories were identified using group-based multi-trajectory modeling in the Chinese cohort and replicated in the Korean cohort via trajectory mean squared error matching. Associations with incident liver disease were estimated using Cox proportional hazards models with progressive covariate adjustment. Mediation analysis assessed the mediating role of mid-wave frailty index with 1000 bootstrap iterations. RESULTS: Four joint trajectory groups were identified: Normal Weight-Low Depression (25.3%), Overweight-Low Depression (31.8%), Normal Weight-High Depression (22.3%), and Obese-Moderate Depression (20.6%). During follow-up, 925 incident liver disease events occurred (5.3%). In fully adjusted models, compared with Normal Weight-Low Depression, hazard ratios for Obese-Moderate Depression were 3.53 (95% confidence interval 2.59-4.79; p < 0.001) in the Chinese cohort and 3.14 (95% confidence interval 2.21-4.47; p < 0.001) in the Korean cohort. Normal Weight-High Depression was associated with hazard ratios of 2.45 (95% confidence interval 1.80-3.35; p < 0.001) and 2.33 (95% confidence interval 1.62-3.33; p < 0.001), respectively. Frailty index mediated 8.5%-15.3% of these associations in the Chinese cohort and 6.8%-12.7% in the Korean cohort. Results remained robust across subgroup and sensitivity analyses, although marginal attenuation was observed for the Overweight-Low Depression group in the Korean cohort. CONCLUSIONS: Joint body mass index-depressive symptom trajectories are differentially associated with incident self-reported physician-diagnosed liver disease, partially mediated through frailty accumulation. Integrated metabolic, psychological, and frailty assessment may inform liver disease risk stratification, although further research with objective outcome ascertainment is needed.