Correspondence

Alcohol Consumption and Mortality among Women

N Engl J Med 1995; 333:1081-1082October 19, 1995

Article

To the Editor:

The article by Fuchs et al. (May 11 issue)1 on the relation between alcohol consumption and mortality among women should set women's minds at ease, especially women at risk for coronary heart disease who drink. It appears that such women profit most from a beneficial effect of alcohol, which is already known for men. Consumption of alcohol by men is cardioprotective in large part because of an increase in serum high-density lipoprotein (HDL) levels,2 whereas cigarette smoking has a deleterious effect on serum HDL levels.3

The standard dietary interview used in the lipid research clinic at our hospital has enabled us to extend these observations in men to women at risk for coronary heart disease. Experienced dietitians estimated daily alcohol consumption among a total of 631 patients with hypercholesterolemia (272 women and 359 men), 44 percent of whom had coronary heart disease. Serum lipoprotein levels were reported as the mean of two serum samples, measured under standardized conditions.3 Thirty-one percent of the women were nondrinkers, 64 percent were light-to-moderate drinkers (up to 30 g daily), and 5 percent were heavier drinkers (more than 30 g daily). Fifteen, 64, and 21 percent of the men were nondrinkers, light-to-moderate drinkers, and heavier drinkers, respectively. The highest HDL levels were observed in the heavier drinkers, the women as well as the men. Using a regression model that simultaneously estimated the effects of alcohol consumption and cigarette smoking, we estimated an increase of 3.5 mg of HDL per deciliter (95 percent confidence interval, 1.3 to 5.6) for every 10 g of alcohol consumption per day in the women. The corresponding estimate in men was 1.1 mg per deciliter (95 percent confidence interval, 0.5 to 1.6), which is similar to earlier observations.2 In contrast, smoking was associated with lower HDL levels in both women (7.8 mg per deciliter lower in smokers; 95 percent confidence interval, 4.2 to 11.3) and men (3.8 mg per deciliter lower in smokers; 95 percent confidence interval, 2.7 to 4.8). An additional adjustment for serum triglyceride levels did not alter the effects of alcohol consumption and smoking on HDL levels. Apparently, urging women at risk for coronary heart disease to refrain from smoking is more advantageous than dissuading them from consuming alcohol.

Eric J.G. Sijbrands, M.D.
Augustinus H.M. Smelt, M.D., Ph.D.
Rudi G.J. Westendorp, M.D., Ph.D.
University Hospital, 2300 RC Leiden, the Netherlands

3 References

1. 1

   Fuchs CS, Stampfer MJ, Colditz GA, et al. Alcohol consumption and mortality among women. N Engl J Med 1995;332:1245-1250
   Full Text | Web of Science | Medline

2. 2

   Gaziano JM, Buring JE, Breslow JL, et al. Moderate alcohol intake, increased levels of high-density lipoprotein and its subfractions, and decreased risk of myocardial infarction. N Engl J Med 1993;329:1829-1834
   Full Text | Web of Science | Medline

3. 3

   Sijbrands EJ, Westendorp RG, Hoffer MJ, et al. Effect of insulin resistance, apoE2 allele, and smoking on combined hyperlipidemia. Arterioscler Thromb 1994;14:1576-1580
   CrossRef | Medline

To the Editor:

Fuchs and coworkers report that light-to-moderate drinkers have some survival advantage over nondrinkers. Their data suggest that the consumption of what might be considered homeopathic amounts of alcohol (i.e., a few grams per day) is beneficial. Are these results real, or do they reflect the arbitrary stratification of a group of “light” drinkers into several subgroups? Would it be wiser and safer to combine the lowest three categories of alcohol consumption, creating a more realistic base-line group of women who consume less than 5 g of alcohol per day?

In Table 2 of the article, in which nondrinkers are the reference group, J-shaped curves are noted for several diseases, including two known to be associated with increased alcohol consumption: death from external injuries and death from hepatic cirrhosis. It would be worthwhile to reanalyze the data after combining the lowest three consumption categories into a single base-line group of women consuming less than 5 g of alcohol per day. The J-shaped curves might disappear, along with the temptation for the large group of nonconsumers to become drinkers.

Albert B. Lowenfels, M.D.
New York Medical College, Valhalla, NY 10595

Author/Editor Response

The authors reply:

To the Editor: Sijbrands and colleagues provide further evidence of the protective effect of alcohol consumption and the deleterious effect of smoking on serum lipoprotein levels. Their findings are consistent with our results, which demonstrate that, among women, the mortality benefit associated with light-to-moderate drinking is largely confined to women with risk factors for coronary heart disease (defined as a history of hypertension or diabetes, smoking, a high cholesterol level, a myocardial infarction in a parent <60 years of age, and a body-mass index [the weight in kilograms divided by the square of the height in meters] >29.0). For women without cardiac risk factors, we observed no significant mortality benefit associated with light-to-moderate drinking.

Dr. Lowenfels questions the association of relatively light drinking (1.5 to 4.9 g of alcohol per day) with a significant reduction in mortality among women. As compared with men, women have higher blood alcohol concentrations for a given amount of alcohol consumed,1 are more susceptible to alcoholic liver disease,2,3 and as Sijbrands and colleagues note, have similar increases in HDL levels with considerably lower levels of alcohol consumption.

Dr. Lowenfels also expresses concern about the categorization of alcohol intake. As we state in our article, all categories of alcohol consumption were established before any data were analyzed. Furthermore, to address the issue of a bias resulting from such categorization, we performed analyses in which alcohol consumption was measured continuously, without any potentially “arbitrary” categorization. This continuous variable was entered into the multivariate model as both linear and quadratic terms (alcohol and alcohol²). As previously demonstrated,3 there was a negative regression coefficient for the linear term (β = -1.3×10-², P = 0.02) and a positive coefficient for the quadratic term (β = 4.9×10-⁴, P = 0.001) — findings consistent with a U- or J-shaped relation. Independently of any categories one might choose, we observed a U-shaped relation between the level of alcohol consumed and overall mortality among women.

Finally, we do not agree with Dr. Lowenfels that our results will encourage a “large group of nonconsumers to become drinkers.” We observed that light-to-moderate drinking was associated with a mortality benefit only among women with risk factors for coronary heart disease. Many of these risk factors are modifiable (e.g., hypertension, hypercholesterolemia, obesity, and smoking),4 and our study demonstrates the deleterious as well as the protective effects of drinking.

Charles S. Fuchs, M.D., M.P.H.
Brigham and Women's Hospital

Meir J. Stampfer, M.D., Dr.P.H.
Walter C. Willett, M.D., Dr.P.H.
Harvard School of Public Health, Boston, MA 02115

4 References

1. 1

   Frezza M, di Padova C, Pozzato G, Terpin M, Baraona E, Lieber CS. High blood alcohol levels in women: the role of decreased gastric alcohol dehydrogenase activity and first-pass metabolism. N Engl J Med 1990;322:95-99
   Full Text | Web of Science | Medline

2. 2

   Norton R, Batey R, Dwyer T, MacMahon S. Alcohol consumption and the risk of alcohol related cirrhosis in women. BMJ 1987;295:80-82
   CrossRef | Web of Science | Medline

3. 3

   Saunders JB, Davis M, Williams R. Do women develop alcoholic liver disease more readily than men? BMJ 1981;282:1140-1143
   CrossRef | Web of Science | Medline

4. 4

   Rich-Edwards JW, Manson JE, Hennekens CH, Buring JE. The primary prevention of coronary heart disease in women. N Engl J Med 1995;332:1758-1766
   Full Text | Web of Science | Medline

Citing Articles (1)

Citing Articles

1. 1

   R. M. Pinkhasov, J. Wong, J. Kashanian, M. Lee, D. B. Samadi, M. M. Pinkhasov, R. Shabsigh. (2010) Are men shortchanged on health? Perspective on health care utilization and health risk behavior in men and women in the United States. International Journal of Clinical Practice 64:4, 475-487
   CrossRef