Linear associations of triglyceride-glucose body mass index and atherogenic index of plasma with the risk of diabetes: a retrospective cohort study

Background This research intends to examine the relationship between triglyceride glucose-body mass index (TyG-BMI) and the Atherogenic Index of Plasma (AIP) in relation to the risk of diabetes, as well as evaluate their predictive capabilities. Methods This retrospective cohort study included data from 116,662 Chinese adults with normal baseline glucose levels. Participants were categorized according to tertiles of triglyceride-glucose body mass index (TyG-BMI; TT1-TT3) and atherogenic index of plasma (AIP; TA1-TA3), as well as nine combined categories (TC1-TC9). Incident diabetes mellitus (DM) was assessed after a mean follow-up of 37.4 months, and the primary endpoint was new-onset DM. Results The overall incidence of DM was 2.3% (n = 2,681). After adjustment for covariates, both TyG-BMI and AIP were independent risk factors for DM (TyG-BMI: OR = 1.0100, p < 0.001; AIP: OR = 2.7800, p < 0.001). In combined analyses, participants in the highest combined category (TC9) had a 15.02-fold higher DM risk than those in the lowest combined category (TC1) in Model 1 and a 3.06-fold higher risk after multivariable adjustment in Model 2. Restricted cubic spline analyses demonstrated dose-response associations. The observed inflection points were 296.292 for TyG-BMI and 0.677 for AIP within this cohort, and addition of TyG-BMI, AIP, or their combination significantly improved DM risk prediction (all p < 0.001). Conclusions In this Chinese adult cohort, TyG-BMI and AIP were significantly associated with incident DM and showed added predictive value. Incorporating TyG-BMI, AIP, or their combination into the baseline risk model significantly improved diabetes risk prediction.