Baseline characteristics of the overall ARIC cohort are presented in by race. African Americans had significantly higher mean high-density lipoprotein cholesterol and body mass index measures, as well as a significantly higher percentage of diabetics, hypertensives, and smokers. Whites had significantly higher mean white blood cell counts and had a significantly higher frequency of the CFH 402HH homozygous genotype. CFH genotype distributions were in accordance with Hardy–Weinberg equilibrium expectations for each race (data not shown). Because of the allele frequency differences between whites and African Americans, analyses were performed separately according to racial group. For all analyses, 402YH heterozygotes and 402HH homozygotes were individually compared to the referent genotype (402YY homozygotes) within the same model.
Baseline characteristics of the overall Atherosclerosis Risk in communities (ARIC) cohort, by race
Analyses evaluating whether the characteristics presented in differed according to CFH genotype (while adjusting for age and gender) showed no significant mean or frequency differences among CFH genotypes for any of the variables considered (data not shown). Results from the Cox proportional hazards models used to estimate the HRRs of both incident CHD and ischemic stroke, for the CFH Y402H polymorphism, are presented in by racial group. The 402HH homozygous genotype was a significant predictor of incident ischemic stroke in whites (HRR 1.47, 95% confidence interval (CI) 1.05–2.05). No significant findings were observed for incident ischemic stroke in African Americans, or for incident CHD in whites or African Americans. We observed that the HRR trends were heterogeneous between whites and African Americans for both incident CHD and ischemic stroke.
Hazard rate ratios (HRRs) relating CFH variant genotypes to incident CHD and ischemic stroke (ISC), in the overall ARIC cohort and by hypertension statue
Due to the involvement of CFH in the inflammatory process, and the association of inflammation with hypertension, we evaluated the interaction effects between genotype and hypertension in CHD, stroke and carotid IMT. Significant interaction effects between genotype and hypertension were observed in the case of CHD in whites with the 402YH genotype (interaction P = 0.05), for carotid IMT in whites with the 402YH and 402HH genotypes (interaction P = 0.009 and 0.04, respectively), and for carotid IMT in African Americans with the 402HH genotype (interaction P = 0.04). Therefore, we performed additional analyses investigating those participants who were normotensive at baseline and those participants who were classified as being hypertensive at any of the four clinic examinations (or period prevalent hypertensives) (). Although no significant interaction effects between genotype and hypertension were observed in the case of ischemic stroke, we performed stratified analyses in this subgroup for consistency. In the baseline normotensives, the 402HH homozygous genotype remained a significant predictor of incident ischemic stroke in whites (HRR 1.62, 95% CI 1.03–2.56). Again, no significant findings were observed for African Americans, but the HRR trend was similar to that observed in whites. Although not significant, a similar trend for an increased risk for ischemic stroke associated with the 402H allele was also observed in whites for the period prevalent hypertensives. As shown in , the frequency of white incident ischemic stroke cases with the 402HH genotype is higher than the frequency of white noncases with the 402HH genotype, and this is observed in the overall cohort as well as in each subgroup evaluated.
CFH genotype frequencies for incident CHD cases, incident ischemic stroke (ISC) cases, and noncases, in the overall ARIC cohort and by hypertension status
Further analyses of incident CHD in baseline normotensives showed a similar nonsignificant trend for an increased risk for incident CHD associated with the 402HH genotype in both whites and African Americans (). However, genotypes carrying the 402H allele were a significant predictor of incident CHD only in whites who were period prevalent hypertensives (402YH: HRR 1.19, 95% CI 1.01–1.40; 402HH: HRR 1.28, 95% CI 1.04–1.57). provides Y402H genotype frequencies in white and African-American incident CHD cases and noncases, both for the overall ARIC cohort and according to hypertension status. The frequency of white incident CHD cases with the 402HH genotype is higher than the frequency of white noncases with the 402HH genotype, and this is observed in the overall cohort as well as in each subgroup evaluated.
Carotid IMT measures are presented in according to race and Y402H genotype for the overall cohort and each subgroup (baseline normotensives and period prevalent hypertensives). The 402H allele was associated with higher IMT measures for whites in the overall cohort and each subgroup. Similar to our analyses of incident disease, the trend for higher IMT measures observed in whites with the 402H allele was only observed in baseline normotensive African Americans.
Carotid IMT measures by CFH genotype, in the overall ARIC cohort and by hypertension status
For all analyses presented, further adjustment for additional covariates did not alter the results (data not shown). Additional covariates included in the analyses are provided in the Methods