Frontiers in Endocrinology · Oct 30, 2026
Association of complete blood cell count-derived inflammatory biomarkers with the prevalence and mortality risk of osteoporosis: evidence from NHANES and an independent Chinese hospital-based case-control cohort
By Di Gao, Xiaotong Zhang, Qiang Huang et al.
Purpose: This study aimed to examine the associations between complete blood count (CBC)-derived inflammatory biomarkers and the prevalence and mortality of osteoporosis, and to further examine these associations in an independent hospital-based case-control cohort. Methods: We evaluated the association of six CBC-derived inflammatory biomarkers with osteoporosis prevalence and all-cause/cardiovascular disease (CVD) mortality from NHANES 2005-2018. Additionally, an independent Chinese hospital-based cohort comprising 63 osteoporosis patients and 63 controls frequency-matched by age and sex was analyzed using unpaired and (Mann-Whitney U) and matched-pair (Wilcoxon signed-rank, 53 pairs) analyses, together with Spearman rank correlations and rank-biserial effect sizes with 95% confidence intervals (CIs). Results: Among 16,539 included participants, 1,223 had osteoporosis. In crude analyses, NLR, MLR, NMLR, and SIRI were associated with osteoporosis prevalence, but none remained statistically significant after multivariable adjustment and FDR correction. NLR, MLR, NMLR, and SIRI were positively associated with all-cause mortality (all FDR P ≤0.006), whereas no biomarker was significantly associated with CVD mortality. In the hospital-based case-control cohort, NLR, MLR, NMLR, and SIRI were significantly elevated in osteoporosis patients in both unpaired and matched-pair analyses (all P < 0.05), with SIRI showing the largest effect size (rank-biserial r = 0.30, P = 0.004) and remaining significant after age and sex adjustment (P = 0.006). Sex-stratified analyses suggested that these differences were confined to females. In exploratory correlation analyses, five of six biomarkers correlated with age in controls (consistent with inflammaging); in osteoporosis patients the point estimates were lower, but the between-group differences were not statistically significant (Fisher z tests, all P > 0.05), and this observation should be considered hypothesis-generating. Conclusions: Elevated NLR, MLR, NMLR, and SIRI are associated with increased all-cause mortality in osteoporosis, whereas their associations with osteoporosis prevalence and CVD mortality were not significant after multivariable adjustment and FDR correction. NLR, MLR, NMLR, and SIRI were elevated in an independent Chinese hospital-based case-control cohort, particularly in female patients. A possible attenuation of the age-inflammation correlation in osteoporosis was observed but is exploratory and requires confirmation. These findings suggest that CBC-derived inflammatory biomarkers may serve as accessible markers of inflammatory burden in osteoporosis.