To the Editor.
—In the article on sexual activity and triggering of MI by Dr Muller and colleagues,1 the data were gathered with an intruiging study design—the case-crossover method.2 The quintessence of the technique is that risk esTime Between Sexual Activity and Onset of Myocardial Infarction, hRelationship between induction time of onset of myocardial infarction and sexual activity. Data are from Muller et al1 and are presented as point estimates with 95% confidence intervals using a log scale. timates are derived from self-reported sexual activities preceding MI compared with the "usual" activity during the previous year. In contrast to an ordinary case-control study, there is no doubt on the choice of the control group because it consists of the patients themselves.After having read the article, one might question one's own sexual activities and conclude that coitus directly preceding a life-threatening illness would certainly be remembered. On the