Nonlinear association of a composite metabolic index (ZJU index) with hypertension: a cross-sectional study of NHANES 2003–2018

Objective To explore the association between the composite metabolic index (ZJU index) and hypertension using data from the National Health and Nutrition Examination Survey (NHANES). Methods NHANES data from 2003 to 2018 were analyzed. Participants were categorized into hypertension and non-hypertension groups. Logistic regression models evaluated the relationship between ZJU index and hypertension. Restricted cubic spline (RCS) and threshold effect analyses assessed nonlinear associations. Subgroup and interaction analyses tested robustness and heterogeneity. The predictive ability of the ZJU index across age groups was evaluated using receiver operating characteristic (ROC) curves. Results After adjusting for covariates, each unit increase in ZJU index was associated with a 7% higher odds of hypertension (OR = 1.07; 95% CI: 1.06–1.07). Participants with higher ZJU index values had significantly increased risk compared to the reference group (OR = 3.73; 95% CI: 3.25–4.29). RCS analyses indicated a nonlinear positive association, with a threshold inflection point at 53.22. Subgroup analyses confirmed consistent associations across all subgroups, while significant interactions were observed for age, education, diabetes history, and smoking status ( P < 0.05). The ZJU index showed moderate predictive ability in individuals under 60 years (AUC = 0.691) and low predictive value in those aged 60 and above (AUC = 0.604). Conclusions An elevated ZJU index is significantly associated with increased hypertension risk among U.S. adults, with a nonlinear dose-response relationship observed.