Comparison of the Cholesterol, High-Density Lipoprotein, and Glucose (CHG) Index, Atherogenic Index of Plasma (AIP), and Triglyceride-Glucose (TyG) Index in Predicting the Risk of New-Onset Hypertension Among Prehypertensive Individuals: A Cohort Study

BackgroundEarly identification of individuals at high risk for hypertension development is crucial for implementing timely preventive strategies. Metabolic indices such as the cholesterol-glucose (CHG) index, atherogenic index of plasma (AIP), and triglyceride-glucose (TyG) index have emerged as potential biomarkers for metabolic and cardiovascular disorders. However, their comparative predictive value for new-onset hypertension in prehypertensive individuals remains unclear. MethodsThis prospective cohort study utilized data from the China Health and Retirement Longitudinal Study (CHARLS), including 2,859 adults with prehypertension followed from 2011 to 2015. Participants were stratified based on progression to incident hypertension. Baseline characteristics and metabolic indices were evaluated. Multivariable logistic regression models, restricted cubic spline (RCS) analyses, receiver operating characteristic (ROC) curves, and subgroup analyses were conducted to assess the associations between CHG, AIP, and TyG indices and the risk of developing hypertension. ResultsDuring the 4-year follow-up, 31.34% (896/2,859) of participants developed new-onset hypertension. All three metabolic indices were independently associated with an increased risk of hypertension after multivariable adjustment. The CHG index demonstrated the strongest association (odds ratio [OR]: 1.96, 95% confidence interval [CI]: 1.45-2.66, P < 0.001), followed by the TyG index (OR: 1.31, 95% CI: 1.07-1.60, P = 0.010). RCS analysis revealed a significant nonlinear relationship between the CHG index and hypertension risk (P for nonlinear = 0.042), whereas AIP and TyG showed linear trends. ROC analysis indicated that the CHG index had the highest discriminatory ability for predicting hypertension (fully adjusted area under the curve [AUC] = 0.7010), outperforming both AIP (AUC = 0.6997) and TyG (AUC = 0.6980). Subgroup analyses showed that the association between the CHG index and incident hypertension was significantly stronger among individuals with lower educational attainment (illiterate), those aged 60-70 or [&ge;]70 years, and widowed individuals (P for interaction < 0.05). ConclusionAmong prehypertensive individuals, higher baseline levels of CHG, AIP, and TyG indices are significantly associated with an increased risk of developing hypertension. The CHG index demonstrates superior predictive performance and may serve as a valuable tool for early risk stratification and targeted intervention in clinical practice.