Author + information
- Sonny Dandona, MD,
- Robert Roberts, MD and
- Alexandre F.R. Stewart, PhD⁎ ()
- ↵⁎University of Ottawa Heart Institute, 40 Ruskin Street, H3100, Ottawa, Ontario, K1Y 4W7, Canada
We thank Mr. Chan and colleagues for the interest in our paper (1). They are able to replicate the association of gene dosage of the risk variant located at 9p21 and severity of coronary artery disease (CAD) in the SAS (Southampton Atherosclerosis Study) when diabetic patients are excluded from their analysis. However, when diabetic patients are included, this association no longer achieves statistical significance. Given the findings of Doria et al. (2) that poor glycemic control tended to amplify the risk association of CAD with 9p21, we had postulated in our original paper that the inability to demonstrate such an association in other cohorts might potentially have been secondary to a confounding effect of diabetes (1). The finding of Mr. Chan and colleagues would lend credence to this hypothesis. We do recognize that further similar confirmatory studies are required. However, these findings do underscore the need for meticulous phenotyping in genetic association studies. Furthermore, it demonstrates the potential complex interplay between genotype, the phenotype of interest, and other associated disease processes that may obscure the finding of such associations.
- American College of Cardiology Foundation