Multimetallic exposure and risk of hypertension: A multi-model analysis based on the China Southwest prospective cohort

OBJECTIVE: Real-world metal exposure is usually mixed, and single-metal analyses cannot reflect actual health risks. This study used multiple statistical approaches to investigate the effects of mixed-metal exposure on incident hypertension and identify key metals and exposure patterns. METHODS: A total of 3749 participants were enrolled from the Guizhou Population Health Cohort Study. Cox proportional hazards model, factor analysis, weighted quantile sum regression (WQS), quantile g-computation (qgcomp), and Bayesian kernel machine regression (BKMR) were employed. RESULTS: During a median follow-up of 6.63 years, 806 incident hypertension cases were identified. After multivariable adjustment, Fe, Zn, Mg, and K were associated with decreased hypertension risk, while As, Mn, and Na were associated with increased risk (P < 0.050). Factor analysis identified three factors: Factor 1 (eight essential metals), Factor 2 (Mn alone), and Factor 3 (Sr‑As). Factor 1 was inversely associated with hypertension (HR=0.920, 95%CI:0.855‑0.990), whereas Factor 2 was positively associated (HR=1.205, 95%CI:1.120‑1.297). In WQS, each 1‑percentile increase in Factor 1 index corresponded to an 11.8% risk reduction (RR=0.882, 95%CI:0.788‑0.988), with K as the main contributor. In qgcomp, each 1‑quantile increase in Factor 1 mixture reduced risk by 10.9% (RR=0.891, 95%CI:0.795‑0.999), with Na and Se driving positive effects, and K, Fe, Mg driving negative effects. BKMR revealed an overall low‑dose beneficial and high‑dose harmful pattern, with Mn as the dominant contributor (PIP=0.807). CONCLUSION: Metal mixtures showed correlated exposure patterns and non‑linear relationships with hypertension. Hypertension prevention should ensure sufficient essential metals and control exposure to hazardous metals including Mn and Na.