C-reactive protein and risk of heart failure. The Rotterdam Study

Background Experimental studies have shown that the known biologic effects of proinflammatory cytokines could explain many aspects of the syndrome of heart failure. The inflammatory marker that presently seems most suitable to assess inflammation is C-reactive protein (CRP). This study was designed to investigate the association between serum CRIP levels, as determined by high-sensitivity assay, and the occurrence of heart failure. Methods Serum CRP levels were available from 6437 men and women without heart failure, aged >= 55 years, from the prospective population-based Rotterdam Study. Cox proportional hazards analysis was used to determine risk of heart failure for sex-specific quartiles of CRP. Results C-reactive protein levels in the highest versus the lowest quartile showed increased hazard ratios of incident heart failure. The age- and sex-adjusted hazard ratio was 2.64 (95% CI 2.04-3.43) for all participants. For men, the age adjusted hazard ratio was 4.37 (2.87-6.66), and for women, 1.86 (1.32-2.62). The interaction term of CRP with sex was highly significant. After additional adjustment for established cardiovascular risk factors, the association attenuated slightly in men and substantially in women, becoming 3.73 (2.40-5.78) and 1.42 (0.99-2.03), respectively. Excluding participants with prevalent coronary heart disease and accounting for incident coronary heart disease resulted in a further attenuation of the hazard ratios, which was proportionately larger in men than in women. Conclusions C-reactive protein is strongly and independently associated with occurrence of heart failure in men. In women, the association is weaker and does not persist after accounting for established cardiovascular risk factors.