C-reactive protein and risk of fracture: a systematic review and dose-response meta-analysis of prospective cohort studies

This study systematically reviews prospective cohort studies evaluating the relationship between C-reactive protein (CRP) concentrations and subsequent fracture risk. The positive association cannot completely explain the existing evidence, and further studies are needed to demonstrate the shape of the association. We aimed to perform a systematic review and dose-response meta-analysis of published prospective studies evaluating associations of high-sensitivity CRP (hs-CRP) levels with fracture risk in general populations. We identified relevant studies by searching MEDLINE and EMBASE databases from their inception to May 20, 2014. We included published prospective studies evaluating the associations of hs-CRP levels with risk of fracture in general populations. Two reviewers working independently abstracted the data. Eight prospective cohort studies involving 34,840 participants and 3,407 incident fracture events were eligible for the present analyses. A meta-analysis of six prospective studies showed that the overall risk for incident fracture in a comparison of individuals in the top tertile with those in the bottom tertile of baseline hs-CRP levels was 2.14 [95 % confidence interval (CI) 1.51-3.05, I (2) = 62.3 %]. The moderate heterogeneous disappeared when one study was excluded. However, the remaining two studies reported inconsistent results. One study with the biggest sample size showed a U-shaped association for CRP and fracture risk (the association was positive when CRP > 1 mg/L). Similarly, another study reported that per doubling of CRP was positive only when CRP > 3 mg/L. In summary, the present analysis showed that the relationship between CRP concentrations and subsequent fracture risk is still inconsistent. The positive association cannot completely explain the existing evidence, and further larger prospective cohorts with more power are needed to demonstrate the shape of the association, especially for the relatively low CRP concentrations, such as less than 3 mg/L.