Large-scale test of hypothesised associations between the angiotensin-converting-enzyme insertion/deletion polymorphism and myocardial infarction in about 5000 cases and 6000 controls

The original report of a possible association between myocardial infarction and the insertion/deletion (I/D) polymorphism of the gene for the angiotensin-1-converting enzyme (ACE) indicated a risk ratio for myocardial infarction with the DD genotype of 1·34 (95% Cl 1·05–1·70), and the association was claimed to be particularly strong in a retrospectively defined low-risk subgroup (3·2 [95% Cl 1·7–5·9). Subsequent investigations reached varying conclusions, but all were small, and much larger studies were needed. 4629 myocardial infarction cases and 5934 controls were compared. Cases were UK men aged 30–54 years and women aged 30–64 years recruited on presentation to hospital with confirmed myocardial infarction. Controls were aged 30–64 years with no history of cardiovascular disease, but were siblings or children of myocardial infarction survivors, or spouses of such relatives. All risk-ratio calculations allow for this relatedness of some of the controls. An updated meta-analysis of previous studies was also conducted. The ACE DD genotype was found in 1359 (29·4%) of the myocardial infarction cases and in 1637 (27·6%) of the controls (risk ratio 1·10 [95% Cl 1·00–1·21]). The association between myocardial infarction and the DD genotype did not seem to be stronger in the subgroup defined as low risk by previously used criteria (234 [28%] of 836 cases and 911 [28%] of 3253 controls: risk ratio 1·04 [95% Cl 0·87–1·24]), or in any other subgroup. Nor was the ACE I/D genotype predictive of subsequent survival. This study involved many more cases than any previously reported study of this question, but did not confirm the existence of any substantial association. In an updated meta-analysis of these results with those of previously published studies, the risk ratio for myocardial infarction with the DD genotype seems to lie in the range 1·0 to about 1·1. Although an increase in risk of up to about 10–15% cannot be ruled out, substantially more extreme risks can be. Moreover, there are not especially strong associations in the subgroups previously selected for emphasis. These findings illustrate the need for some studies of candidate genes to involve much larger populations than is customary, without undue emphasis on retrospectively defined subgroups.