Association between systemic redox balance and osteoporosis: prospective evidence, polygenic modification, and proteomic and inflammatory mediation in the UK Biobank

Purpose
Redox signaling governs bone remodeling, but whether systemic redox balance is associated with incident osteoporosis, genetic susceptibility, and proteomic/inflammatory pathways at population scale remains unclear.
Methods
We analyzed UK Biobank participants free of osteoporosis at baseline. Serum redox balance score (SRBS) combined albumin, total bilirubin, and γ-glutamyl transferase. Cox proportional hazards models adjusted for prespecified covariates. Effect modification by an osteoporosis polygenic risk score (PRS) was tested on multiplicative and additive scales. Mediation was evaluated in a proteomics subset and an inflammatory-panel subset using counterfactual mediation with multiple-testing control.
Results
Over a median follow-up of 12.8 years (IQR, 11.7–13.7), 12,893 incident osteoporosis cases were observed. SRBS demonstrated a nonlinear inverse association with osteoporosis, displaying a J-shaped pattern: relative to Q1, multivariable hazard ratios (95% CIs) were 0.82 (0.78–0.86) for Q2, 0.75 (0.71–0.78) for Q3, and 0.72 (0.68–0.75) for Q4; the per–standard-deviation increase corresponded to an HR of 0.85 (0.83–0.87). Cumulative-incidence curves diverged early and showed a stepwise gradient across quartiles. Associations were stronger among men and physically inactive participants. SRBS interacted with the osteoporosis PRS on the multiplicative scale (interaction HR, 1.03; 95% CI, 1.02–1.04), whereas evidence for additive interaction was limited. Proteomic mediation implicated EGFR, TNFRSF10A, CBLN4, CD27, and IGDCC4 (≈8–11% each); inflammatory mediation implicated C-reactive protein (≈8%), platelets (≈4%), and neutrophils (≈4%).
Conclusion
Systemic redox balance values were linked to osteoporosis risk, with partial mediation through plasma-protein and inflammatory pathways and only modest modification by polygenic risk.