The association between the visceral obesity indices and the future diabetes mellitus risk: A prospective cohort study

Abstract Aim As abnormal visceral fat accumulation is the core pathophysiology of diabetes mellitus (DM), this study evaluated five novel visceral obesity indices to provide optimized clinical metabolic risk assessment tools. Materials and Methods This study included 6575 participants aged ≥40 years without baseline diabetes. Waist‐to‐height ratio (WHtR), body roundness index (BRI), visceral adipose index (VAI), lipid accumulation product (LAP) and abdominal body shape index (ABSI) were exposure variables. Multivariable‐adjusted Cox proportional hazards models analysed associations with diabetes onset; restricted cubic splines (RCS) explored dose–response relationships, stratified analyses and receiver operating characteristic (ROC) curve assessed model stability and predictive efficacy. Results During an average follow‐up period of 3.19 years, 752 (11.4%) of participants developed diabetes. Multivariable Cox regression showed that each visceral obesity index independently predicted the risk of diabetes (all P for trend <0.05), with LAP showing the strongest association (HR = 2.93, 95% CI = 2.27–4.01). The RCS model revealed the characteristics of a nonlinear dose–response relationship. Stratified analysis and sensitivity analysis confirmed the high stability of the association between LAP and the risk of diabetes. ROC curve analysis indicated that LAP had the optimal predictive efficacy (AUC = 0.752). Conclusions Visceral obesity indices are closely linked to the risk of diabetes onset, highlighting the potential benefits of reducing visceral fat accumulation. Among these indices, LAP emerges as the most robust clinical indicator for predicting DM risk.