To the Editor.
—Lowenstein et al1 have contributed an interesting study of the role of alcohol in new-onset atrial fibrillation (AF). Although they appropriately pointed out its limitations, it is, altogether, a valuable report. Since the Archives has a teaching role, however, I should like to clarify a few points.First, the authors state that "AF has a well-recognized association with pericarditis...."1(p1882) This may be said in one or more of the references given, but the only formal studies of pericarditis have shown that, in fact, pericarditis per se has no association (and, perhaps, a negative association) with arrhythmias.2,3 (Recent, as yet unpublished, findings from Holter ECG monitor studies by us are in accord.) Indeed, notable arrhythmias with pericarditis do not occur unless there is accompanying heart disease or cardiac tamponade, which is tantamount to the same thing.2,3 This is not to say that there may