Obesity-metabolic heterogeneity and its transitions in relation to incident cardiovascular disease and all-cause mortality among adults with CKM stages 0–3: a nationwide prospective cohort study

Abstract Background Cardiovascular disease (CVD) remains a major cause of mortality, and obesity, diabetes, and chronic kidney disease frequently coexist and increase cardiovascular risk. The American Heart Association cardiovascular-kidney-metabolic (CKM) framework shifts attention toward prevention before overt CVD. Whether metabolic obesity phenotypes and their changes further stratify CVD risk in CKM stages 0–3, or relate similarly to all-cause mortality, remains unclear. Methods We studied adults aged ≥ 45 years in the China Health and Retirement Longitudinal Study (2011–2020). Baseline phenotypes combined body mass index and metabolic health: MHNW, MHOO, MUNW, and MUOO. Cox models evaluated incident CVD and all-cause mortality. A 2015 landmark analysis related 2011–2015 metabolic-health transitions to outcomes. Death before CVD was treated as a competing event; proportional hazards, restricted mean survival time (RMST), subgroup, sensitivity, and bootstrap C-index analyses were performed. Results The baseline CVD analysis included 8852 participants and 2221 events. Compared with MHNW, fully adjusted hazard ratios (HRs) were 1.54 (95% confidence interval [CI] 1.35–1.75) for MHOO, 1.33 (95% CI 1.18–1.50) for MUNW, and 1.73 (95% CI 1.55–1.94) for MUOO. The landmark analysis included 5738 participants and 1109 CVD events. Relative to stable metabolic health, HRs were 1.24 (95% CI 1.01–1.52) for improvement, 1.69 (95% CI 1.40–2.03) for deterioration, and 2.01 (95% CI 1.72–2.35) for stable metabolic unhealthiness. Fine–Gray and sensitivity results were concordant. Phenotype information increased the baseline and landmark C-indices by 0.0170 (95% CI 0.0110–0.0257) and 0.0362 (95% CI 0.0233–0.0506), respectively. All-cause mortality associations were not robust: all adjusted 95% CIs crossed 1.0, RMST differences were small, and C-index increments were near zero. Conclusions In middle-aged and older Chinese adults across CKM stages 0–3, incident CVD risk differed meaningfully across metabolic obesity phenotypes and their longitudinal changes. Persistent metabolic health marked the lowest-risk course, whereas metabolic deterioration and sustained unhealthiness were associated with higher subsequent risk. Serial phenotyping may add useful information for early CKM risk assessment and prevention, although the modest gain in discrimination does not support its use as a stand-alone prediction tool.