Association between a composite serum oxidative stress biomarker score and systemic inflammation: A multi-provincial...

Oxidative stress and chronic inflammation share common upstream triggers and can mutually reinforce each other; however, few tools based on routine blood biomarkers are available to capture this interplay in Chinese adults. Using data from 6,044 participants in the 2009 China Health and Nutrition Survey (CHNS), a large multi-provincial population-based study, we developed a composite serum oxidative stress-related biomarker score (COSBS), hereafter referred to as a serum redox-metabolic biomarker score to reflect its overlap with metabolic phenotypes.The score was constructed by summing quintile-based points for six fasting biomarkers: uric acid, triglycerides, fasting glucose, and ferritin as pro-oxidant components, and HDL cholesterol and magnesium as antioxidant components, yielding a total score ranging from 6 to 30. Equal weighting was applied because no validated differential weights exist for this population and to avoid circularity with the outcome. Reliability and structural properties of COSBS were preliminarily evaluated using Cronbach's α (0.615) and principal component analysis (PC1 = 35.5% of variance), reflecting acceptable but moderate internal consistency for a heterogeneous biomarker panel. COSBS effectively discriminated established high-risk groups (all P < 0.001). Analyses did not apply post-stratification weights; a cluster-robust sensitivity analysis confirmed that community-level clustering did not alter conclusions. Each one-unit increase in COSBS was independently associated with higher log(hs-CRP) (β = 0.0255, 95% CI: 0.0213-0.0297, P = 1.73 × 10⁻³²), with a strictly linear dose-response relationship (P for non-linearity = 0.372). Results were consistent across all pre-specified subgroups and sensitivity analyses, including a clinical threshold count score (β = 0.081, P = 5.51 × 10⁻²⁸). These findings suggest that COSBS, derived entirely from routine blood tests, is independently and linearly associated with systemic inflammation as reflected by hs-CRP, providing a reproducible exposure index for population-based research on oxidative-inflammatory burden. Independent external validation in NHANES 2015-2018 (n = 2,752 U.S. adults) replicated the association directionally and quantitatively (β = 0.048, P = 2.33 × 10⁻¹³; 25.2% higher geometric mean hs-CRP per SD of COSBS-5), supporting cross-population generalizability. Prospective validation against hard endpoints is required before any clinical application can be considered.