| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
Pantox Laboratories, 4622 Santa Fe Street, San Diego, CA 92109
Dear Sir:
Jéquier's (1) recent editorial regarding the alcohol paradox notes that "epidemiologic evidence does not show a clear relation between daily alcohol energy intake and body weight." This is true with respect to men. However, large cross-sectional epidemiologic surveys show that women who are moderate drinkers tend to be much lighter than women who do not drink (2, 3). In a cohort of nearly 90000 women participating in the extensive Nurses' Health Study, the average body mass index (BMI) of women who were moderate drinkers (1–2 drinks daily) was 
15% lower than that of nondrinkers—even though self-reported daily energy consumption was higher for the drinkers (2). In the British Health Survey, women who were moderate drinkers were about half as likely to be obese as nondrinkers (3). Group differences in age, smoking habits, or physical activity do not appear to account for these remarkable (and remarkably overlooked) findings.
An explanation for this paradox may be readily at hand. Cross-sectional studies also show that moderate drinkers tend to be much more insulin sensitive than abstainers; this sensitivity is associated with lower fasting and postprandial insulin concentrations (4–6). Controlled studies documenting induction of insulin sensitivity by chronic ethanol consumption are lacking, but the magnitude of the sensitization associated with alcohol use is not likely to be explained by confounding factors. If we assume that the insulin sensitization produced by chronic ethanol ingestion is specific to skeletal muscle (perhaps induced by muscle metabolism of acetate?), the accompanying down-regulation of diurnal insulin secretion will diminish net insulin activity on adipocytes and hepatocytes, discouraging fatty acid storage while disinhibiting hepatic ketogenesis and gluconeogenesis, which are thermogenic.
Indeed, results of both clinical and animal research suggest that diminution of diurnal insulin secretion tends to promote leanness, independent of any effect on energy consumption (7). A decrease in insulin secretion may mediate, at least in part, reductions in body fat associated with exercise training, very-low-fat whole-food vegan diets, or administration of the drugs diazoxide, acarbose, and metformin. Arguably, the most effective strategy for achieving and maintaining leanness is to minimize daily insulin secretion within the context of a low fat intake.
Even though most of the metabolic energy derived from ethanol is presumably available to participate in feedback control of appetite, ethanol itself may have an appetite-stimulating effect, as recently shown in this Journal (8). Nevertheless, epidemiologic analysis suggests that chronic alcohol consumption tends to quell carbohydrate cravings, most notably in women (2). Thus, although alcohol use tends to increase daily energy consumption in both sexes, it does so more markedly in men, which perhaps explains why alcohol promotes leanness in women but not in men. Note, however, that men drinkers are no heavier than nondrinkers despite decidedly higher energy intakes (from food plus ethanol).
Although down-regulation of insulin secretion should promote fat oxidation, the immediate effect of ethanol ingestion is a selective inhibition of fat oxidation (9). This may explain why the BMI of women in the highest category of alcohol consumption tended to be higher than that of women who drank more moderately (2). Theoretically, the minimal daily dose of ethanol that produces substantial insulin sensitization should have the most favorable effect on body composition—and should be reasonably safe from the standpoint of risks of breast cancer, hypertension, and liver damage while promoting vascular health.
The effect of alcohol on body composition in women may be of more than just cosmetic significance. In the Nurses' Health Study, women who had 
2 drinks daily were 70% less likely to develop diabetes than nondrinkers during 4 y of follow-up (10). The authors of this study used statistical corrections for BMI to conclude that the true reduction in risk associated with alcohol was 40%, but this correction may not have been appropriate if alcohol use was primarily responsible for the lower BMI of the drinkers. Thus, the direct insulin-sensitizing effect of ethanol and its longer-term favorable effect on body weight may collaborate to substantially reduce diabetes risk in women.
Jéquier is absolutely right to call for longer-term studies of the metabolic effects of ethanol ingestion; the short-term studies completed to date only deepen the sense of paradox. In all probability, a better understanding of the long-term adaptive response to regular alcohol consumption will enable a definitive resolution of this issue.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
