If we consider the significance of one additional self-reported drink / day in the context of lifetime exposure, the dose-response relationship looks quite different. Assuming for the purpose of illustration that each additional 10 grams/day reported by a woman upon entering a study corresponds to an additional 10 g/day from the age of beginning drinking (18 years old) to age 55, then this corresponds to ≈ 135,000 g, or 135 kg of ethanol over that time period. Even smaller differences in self-reported amounts at study entry would correspond to large additional amounts consumed over a lifetime. Of course, in reality the relationship between self-reported drinking by postmenopausal women and lifetime consumption is not so simple or linear. However, if alcohol consumption in postmenopausal women is proportional to lifetime consumption, the basic point still holds; small differences in self-reported drinks/day at study entry correspond to large differences in the amount of alcohol consumed over a lifetime, and it is these large differences in cumulative carcinogen exposure over the lifetime that are responsible for the small differences in breast cancer risk.
