Association between serum level of C-reactive protein and risk of cardiovascular events based on cohort studies

Abstract

Although the association between serum level of C-reactive protein (CRP) and risk of cardiovascular events (CVEs) has been reported, the comprehensive assessment of the quantitative association of CRP level with risk of CVEs has not been reported. Our meta-analysis aims to quantitatively evaluate the association of CRP level and risk of CVEs. We searched PubMed and Embase databases for articles published up to December 6, 2019. Studies with data on men and women, different types of CVEs and multiple cohorts within a study were treated as independent studies. Generalized least-squares regression models were used to assess the quantitative association between CRP level and risk of CVEs. Restricted cubic splines were used to model the possible linear association between CRP and CVEs. We included 36 articles (60 studies; 227,715 participants) in the analysis. The pooled relative risks (RRs) of high versus low CRP level for cardiovascular disease (CVD), stroke and coronary heart disease (CHD) were 1.64 (95% confidence interval [CI], 1.49-1.82), 1.46 (95% CI, 1.35-1.58), and 1.55 (95% CI, 1.47-1.63), respectively. A linear association was found between CRP level and CVD (P = 0.429), stroke (P = 0.940), and CHD (P = 0.931); with each 1-mg/L increase in CRP level, the pooled RRs for CVD, stroke, and CHD were 1.18 (95% CI, 1.12-1.24), 1.07 (95% CI, 1.04-1.09), and 1.12 (95% CI, 1.08-1.16), respectively. This meta-analysis suggests that risk of CVEs increases with increasing serum CRP level.