Context: Associations between chromosome 9p21 single-nucleotide polymorphisms (SNPs) and heart disease have been reported and replicated. If testing improves risk assessments using traditional factors, it may provide opportunities to improve public health.
Objectives: To perform a targeted systematic review of published literature for effect size, heterogeneity, publication bias, and strength of evidence and to consider whether testing might provide clinical utility.
Data sources: Electronic search via HuGE Navigator through January 2009 and review of reference lists from included articles.
Study selection: English-language articles that tested for 9p21 SNPs with coronary heart/artery disease or myocardial infarction as primary outcomes. Included articles also provided race, numbers of participants, and data to compute an odds ratio (OR). Articles were excluded if reporting only intermediate outcomes (eg, atherosclerosis) or if all participants had existing disease. Twenty-five articles were initially identified and 16 were included. A follow-up search identified 6 additional articles.
Data extraction: Independent extraction was performed by 2 reviewers and consensus was reached. Credibility of evidence was assessed using published Venice criteria.
Data synthesis: Forty-seven distinct data sets from the 22 articles were analyzed, including 35 872 cases and 95 837 controls. The summary OR for heart disease among individuals with 2 vs 1 at-risk alleles was 1.25 (95% confidence interval [CI], 1.21-1.29), with low to moderate heterogeneity. Age at disease diagnosis was a significant covariate, with ORs of 1.35 (95% CI, 1.30-1.40) for age 55 years or younger and 1.21 (95% CI, 1.16-1.25) for age 75 years or younger. For a 65-year-old man, the 10-year heart disease risk for 2 vs 1 at-risk alleles would be 13.2% vs 11%. For a 40-year-old woman, the 10-year heart disease risk for 2 vs 1 at-risk alleles would be 2.4% vs 2.0%. Nearly identical but inverse results were found when comparing 1 vs 0 at-risk alleles. Three studies showed net reclassification indexes ranging from -0.1% to 4.8%.
Conclusion: We found a statistically significant association between 9p21 SNPs and heart disease that varied by age at disease onset, but the magnitude of the association was small.