C-reactive protein haplotype predicts serum C-reactive protein levels but not cardiovascular disease risk in a dialysis cohort

Background: C-Reactive protein (CRP) gene variation has been associated with serum CRP levels in the general population. We examined the associations of CRP gene variation with longitudinal CRP measurements and incident cardiovascular disease (CVD) risk in a cohort of 504 white and 244 African-American incident dialysis patients. Methods: Seven tagging single-nucleotide polymorphisms in the CRP gene were selected by using the Carlson method (r(2) > 0.65). High-sensitivity CRP was measured every 6 months (mean, 4.6 months). Haplo.glm was used to determine the association of haplotypes with serum CRP levels and CVD risk. Global tests from Haplo.score were conducted to determine statistical significance. Results: Compared with the most common haplotype, 1 haplotype was associated with a 52% lower CRP level at baseline among African Americans (ratio, 0.48; 95% confidence interval [Cl], 0.28 to 0.82; global P-value = 0.0005). Furthermore, this haplotype was associated significantly with lower serum CRP levels during 36 months of follow-up. Among whites, this haplotype was associated with an 18% (ratio, 0.82; 95% Cl, 0.56 to 1.22; n = 6 carriers) lower CRP level compared with the most common haplotype with a frequency of 1% (global P-value = 0.048). No association was detected between CRP gene variation and CVD risk in either whites or African Americans. Conclusion: Compared with the most common haplotype of the CRP gene, 1 haplotype predicts a lower serum CRP level over time, but no association exists between haplotype of CRP gene and incident CVD in this incident dialysis population. Serum CRP level might be a biomarker, rather than a causal factor, in CVD development. CRP variation may lead to susceptibility to inflammation, but not risk for CVD; however, replication in multiple settings is necessary.