THE AUTHORS REPLY

Extract

We thank Dr. Phillips for his views (1). Phillips’ original and novel observation in a small case-control study of increased risk of myocardial infarction among men who shaved infrequently (2) was the starting point for our investigation. Our finding of very similar strengths of association between less frequent shaving and lung cancer, all-cause mortality, and cardiovascular diseases demonstrated a lack of specificity in the relation (3). Appropriate adjustment for potentially confounding factors attenuated the hazard ratios towards the null value for all causes of death, although hazard ratios for stroke events remained substantially raised among infrequent shavers.

We interpreted these findings as probably being due to confounding by smoking and social factors, whereas Phillips suggests that smoking lies on a causal pathway between frequency of shaving and coronary heart disease, mediated by hyperestrogenemia, which may be caused by smoking or may be a cause of smoking. Which is the more plausible explanation?