Multi-trajectory modeling of metabolic syndrome indicators and cardiovascular disease risk: a study based on health examination big data.

BACKGROUND: The clinical utility of metabolic syndrome (MetS) in predicting cardiovascular disease (CVD) and diabetes has been questioned due to its binary definition, which results in substantial loss of information from metabolic indicators. This study aims to longitudinally evaluate MetS measurement indicators, identify their multi-trajectory patterns over time, and assess the associated CVD risk. METHODS: Group-based multi-trajectory modeling (GBMTM) was applied to metabolic syndrome indicators to identify multi-trajectory patterns and describe baseline characteristics of subgroups. Subsequently, with CVD as the outcome, trajectory groups were incorporated into interval-censored Cox proportional hazards models to estimate the associated CVD risks across different trajectory patterns. RESULTS: This study identified five distinct metabolic patterns through trajectory typing. The “progressive hyperglycemia trajectory” (characterized by high and continuously increasing fasting plasma glucose [FPG] levels) and the “persistent dyslipidemia trajectory” (characterized by persistently high triglycerides [TG], low high-density lipoprotein cholesterol [HDL-C] with continuous decrease) demonstrated higher CVD incidence density. After adjusting for age and sex, their hazard ratios (HR) were 1.469 (95% CI: 1.319–1.635) and 1.355 (95% CI: 1.189–1.544), respectively. CONCLUSIONS: This study identified five distinct longitudinal trajectories of MetS indicators and demonstrated their significant associations with CVD risk. The progressive hyperglycemia trajectory and the persistent dyslipidemia trajectory were associated with higher CVD risk, underscoring the importance of longitudinal monitoring of FPG, TG, and HDL-C for CVD prevention in elderly populations. CLINICAL TRIAL NUMBER: Not applicable.