Four-year changes in metabolic burden and transitions to cardiovascular disease and death: a longitudinal CHARLS study

Background Metabolic abnormalities evolve over time, yet their associations may differ between cardiovascular disease (CVD) onset and death before or after CVD. We examined whether four-year change in metabolic burden was associated with transitions among CVD-free, CVD, and death states in Chinese middle-aged and older adults. Methods This longitudinal cohort study used the China Health and Retirement Longitudinal Study. Five metabolic components were assessed in 2011 and 2015, and subsequent transitions were observed during 2015–2018 and 2018–2020. Change in the component count was modelled continuously and as improvement, stability, or deterioration. Transition-specific discrete-time logistic models estimated odds ratios (ORs) for CVD-free to CVD, CVD-free to all-cause death, and CVD to all-cause death, with participant-clustered robust standard errors. Results Among 4,716 participants, 1,855 (39.3%) improved, 1,730 (36.7%) remained stable, and 1,131 (24.0%) deteriorated. The 8,732 interval records included 939 CVD-free-to-CVD, 220 CVD-free-to-death, and 103 CVD-to-death transitions. Each one-component increase in metabolic burden was associated with higher odds of CVD onset (OR 1.14, 95% confidence interval [CI] 1.06–1.21). Improvement versus stability was associated with lower odds of CVD onset (OR 0.79, 95% CI 0.66–0.94). Continuous metabolic change was not associated with direct death from the CVD-free state (OR 1.08, 95% CI 0.93–1.25) or death after CVD (OR 0.91, 95% CI 0.74–1.12). The CVD-onset estimate was similar after sampling and attrition weighting, exclusion of baseline diabetes, and restriction to the later follow-up interval. Conclusions Four-year worsening of metabolic burden was associated with subsequent CVD onset, while metabolic improvement was associated with lower odds of moving from a CVD-free state to CVD. Associations with mortality transitions were imprecise. Repeated metabolic assessment may provide information about CVD development that is not captured by a single baseline measurement.