An evaluation of inflammatory gene polymorphisms in sibships discordant for premature coronary artery disease: the GRACE-IMMUNE study
1 Leeds Institute of Genetics, Health and Therapeutics (LIGHT), University of Leeds, UK
2 Leeds Institute of Molecular Medicine (LIMM), University of Leeds, UK
3 Roche Molecular Systems, Pleasanton, California, USA
4 Department of Cardiovascular Sciences, University of Leicester, UK
BMC Medicine 2010, 8:5 doi:10.1186/1741-7015-8-5Published: 13 January 2010
Inflammatory cytokines play a crucial role in coronary artery disease (CAD). We investigated the association between 48 coding and three non-coding single nucleotide polymorphisms (SNPs) from 35 inflammatory genes and the development of CAD, using a large discordant sibship collection (2699 individuals in 891 families).
Family-based association tests (FBAT) and conditional logistic regression (CLR) were applied to single SNPs and haplotypes and, in CLR, traditional risk factors of CAD were adjusted for.
An association was observed between CAD and a common three-locus haplotype in the interleukin one (IL-1) cluster with P = 0.006 in all CAD cases, P = 0.01 in myocardial infarction (MI) cases and P = 0.0002 in young onset CAD cases (<50 years). The estimated odds ratio (OR) per copy of this haplotype is 1.21 (95% confidence interval [95CI] = 1.04 - 1.40) for CAD; 1.30 (95CI = 1.09 - 1.56) for MI and 1.50 (95CI = 1.22 - 1.86) for young onset CAD. When sex, smoking, hypertension and hypercholesterolaemia were adjusted for, the haplotype effect remained nominally significant (P = 0.05) in young onset CAD cases, more so (P = 0.002) when hypercholesterolaemia was excluded. As many as 82% of individuals affected by CAD had hypercholesterolaemia compared to only 29% of those unaffected, making the two phenotypes difficult to separate.
Despite the multiple hypotheses tested, the robustness of family design to population confoundings and the consistency with previous findings increase the likelihood of true association. Further investigation using larger data sets is needed in order for this to be confirmed.