The curve that relates mortality (vertical axis) to alcoholic drink consumption (horizontal axis) is J-shaped; i.e. as consumption rises above zero the all-cause mortality declines, then levels off, and finally rises.
The benefit is largely a reduction of deaths due to cardio- and cerebrovascular disease for regular drinkers of 1-2 units/d for men over 40 years and postmenopausal women. Consumption over 2 units/d does not provide any major additional health benefit. The mechanism may be an improvement in lipoprotein (HDL/LDL) profiles and perhaps a reduction in platelet aggregation.
The effect appears to be due mainly to ethanol itself but nonethanol ingredients (antioxidants, phenols, flavinoids) may contribute (see below).
The rising (adverse) arm of the curve is associated with known harmful effects of alcohol (already described), but also, for example, with pneumonia (which may be secondary to direct alcohol effects, or with the increased smoking of alcohol users).
Whether the cardioprotective effect differs between classes of alcoholic drink remains an open issue. Suggestion that wine confers greater advan-
35 For pictures see Streissguth A P et al 1985 Lancet 2: 85-91.
tage than spirits was not supported by a review of 12 ecological, 3 case-control and 10 prospective cohort studies.36 The social importance of alcohol combined with the very substantial scientific problems posed by these studies (including the problem of unreliably reported intakes) render the whole matter controversial.
Was this article helpful?