The effect of diltiazem on long-term outcome after acute myocardial infarction (AMI) was assessed in 2,377 patients enrolled in the Multicenter Diltiazem Post-Infarction Trial and subsequently followed for 25 +/- 8 months. The study population included 855 patients (36%) with at least 1 prior AMI before the index infarction and 1,522 patients (64%) with a first AMI, of whom 409 (27%) had a first non-Q-wave AMI, 664 (44%) a first inferior Q-wave AMI, and 449 (30%) a first anterior Q-wave AMI. This post hoc analysis revealed that, among patients with first non-Q-wave and first inferior Q-wave AMI, there were fewer cardiac events during follow-up in the diltiazem than in the placebo group, and that the reverse was true for patients with first anterior Q-wave AMI or prior infarction. The diltiazem:placebo Cox hazard ratio (95% confidence limits) for the trial primary end point (cardiac death or nonfatal reinfarction, whichever occurred first) was: first non-Q-wave AMI-0.48 (0.26, 0.89); first inferior Q-wave AMI-0.66 (0.40, 1.09); first anterior Q-wave AMI-0.82 (0.51, 1.31); and prior AMI-1.11 (0.85, 1.44). Use of cardiac death alone as an end point gave an even more sharply focused treatment difference: first non-Q-wave AMI-0.46 (0.18, 1.21); first inferior Q-wave AMI-0.53 (0.27, 1.06); first anterior Q-wave AMI-1.28 (0.68, 2.40); prior infarction-1.26 (0.90, 1.77). Further analysis revealed that these differences in the effect of diltiazem in large part reflected the different status of the 4 electrocardiographically defined subsets in terms of left ventricular function.(ABSTRACT TRUNCATED AT 250 WORDS)