|| Checking for direct PDF access through Ovid
The absence of a clear relation between genotype and phenotype has complicated molecular analyses of the susceptibility to common diseases such as cardiovascular disease. Gene-gene and gene-environment interactions may have biased genetic association studies on cardiovascular disease. In the general population, susceptibility to cardiovascular disease is probably the result of many loci with frequently occurring alleles that have relatively small effects. We hypothesize that addition of mortality analyses to genetic association studies may provide important information on the clinical relevance of molecular findings. We propose to use a parent-offspring model for this purpose: the parents contribute large follow-up and the index cases (offspring) are eliminated from the analyses to remove selection on cardiovascular disease. In particular, indirect estimation of mortality risk has high statistical power and may establish the role of common genetic variation with small effects on cardiovascular disease in the general population. Moreover, the effect of a candidate gene on excess mortality illustrates the quality of such a gene or protein as a novel target for intervention. The results of genetic association studies may have been of little clinical relevance for cardiovascular disease, but we conclude that the methodology still has possibilities for improvements and we propose to use analyses of mortality in a parent-offspring model.