C-reactive protein and fracture risk: an updated systematic review and meta-analysis of cohort studies through the use of both frequentist and Bayesian approaches.

Abstract

The association between the inflammatory marker CRP and osteoporotic fracture has remained uncertain. In this study, we conducted a systematic review and meta-analysis to examine the association of serum hs-CRP and fracture risk. We performed a systematic literature search of relevant databases, including PubMed, Embase, and MEDLINE publications from January 1950 through April 2020. Three reviewers independently performed the study selection, quality assessment, and data abstraction. Frequentist and Bayesian hierarchical random-effects models were used separately for the analysis. Statistical heterogeneity was assessed using Higgin's I2 and Cochran's Q statistic, and publication bias was examined using funnel plots and rank correlation tests. Fourteen cohort studies that reported t fracture outcomes were eligible for the systematic review. Only ten studies (n = 29,741) qualified for meta-analysis. In the frequentist approach, the RR for fracture in a comparison of the top tertile group to the bottom tertile group of hs-CRP was 1.54 (1.18, 2.01). The estimated risk of fracture remained significant in all sensitivity and subgroup analyses. Higgin's I2 (30.52%) and Cochran's Q statistic (p < 0.01) suggested there was moderate heterogeneity for the meta-analysis. In the Bayesian approach, the pooled RR was 1.60 (95% CI (1.07-2.49)), and the probabilities that the high level of hs-CRP increased fracture risk by more than 0%, 10%, and 20% were 99%, 98%, and 93%, respectively. A high level of hs-CRP is associated with a significantly increased risk of osteoporotic fracture.