
When Rhodes and his colleagues (1) described the tendency for patients with ulcerative colitis to be non-smokers it would have been natural to assume that medical advice as a consequence of developing chronic illness had led those who had colitis and who were smokers to stop smoking. The other possibility was that the fracture clinic controls chosen were unrepresentative of the ordinary population and contained an unduly large proportion of smokers. This latter explanation was unlikely, however, because only 8