Cigarettes, Ulcerative Colitis, and Inferences from Uncontrolled Data

Abstract

The myriad components of cigarette smoke include some potent toxic and pharmacologic agents. Of those whose effects are known, most are unequivocally bad.¹ That is no reason, however, to assume a priori that all the effects must be bad, and indeed an apparent beneficial effect on Parkinsonism was reported many years ago.² In this issue of the Journal, Jick and Walker³ report a statistically significant negative association between cigarette smoking and ulcerative colitis, and they cite several earlier and more limited reports of the same relation. What should we make of this? Four possibilities come to mind: the observed . . .