Correspondence

Primary Prevention of Coronary Heart Disease in Women through Diet and Lifestyle

N Engl J Med 2000; 343:1814-1815December 14, 2000

Article

To the Editor:

Stampfer et al. (July 6 issue)1 report impressively low levels of coronary heart disease among women with healthy lifestyles. Unfortunately, their definition of lifestyle includes at least one variable, the body-mass index, that is itself a measure of disease status, rather than a matter of behavior. Although poor diet and lack of exercise presumably contribute to many cases of overweight and obesity, one cannot simply assume that they are the only factors leading to a high body-mass index. In fact, the study by Stampfer et al. provides evidence that factors other than lifestyle influence the body-mass index, since a high body-mass index remained associated with an elevated risk of heart disease even after diet and exercise had been taken into account. This independent association of the body-mass index with the risk of heart disease suggests that some factor other than diet and exercise causes both a high body-mass index and an elevated risk of heart disease.

The article by Stampfer et al. provides useful information, as long as one bears in mind that the body-mass index is not a lifestyle factor and that therefore the stated contribution of behavior to heart disease has probably been overestimated. Of course, even when looking at actual behavior, one should consider that genetics may play a part in how much hunger a person experiences, how energetic that person feels, and how susceptible he or she is to nicotine addiction — but that is a different problem.

Judith S. Shapiro, Ph.D.
Madonna University, Livonia, MI 48150-1173

1 References

1
Stampfer MJ, Hu FB, Manson JE, Rimm EB, Willett WC. Primary prevention of coronary heart disease in women through diet and lifestyle. N Engl J Med 2000;343:16-22
Full Text | Web of Science | Medline

To the Editor:

Stampfer et al. present comprehensive data supporting the association of five low-risk dietary and lifestyle characteristics with an impressively low relative risk of coronary heart disease of 0.17. However, before we discard the statistical dictum that correlation does not prove causation and undertake the unenviable task of changing the lifestyle of the remaining 97 percent of this population, we need to reflect on a few issues.

The risk of death (296 deaths among 84,129 women over a period of 14 years) is extremely low, at 0.025 percent annually. Therefore, if — for the remainder of the cohort — one could achieve a relative risk of 0.20, the absolute reduction in risk would be a mere 0.020 percent, and the number needed to treat would be challenging, at 5000. Assuming an average remaining life expectancy at the time of death of 45 years, 761 years of life could be saved annually (80 percent × 296 deaths × 45 years saved ÷ 14 years of observation), an impressive achievement from the epidemiologic perspective. However, for an individual person, the chances of living in any year (assuming an annual rate of mortality from all causes of 1.1 percent in this group) would change only from 98.90 percent to 98.92 percent.

To achieve this goal, up to five changes in diet and lifestyle would be required, including a total of 1748 years of exercise annually in the cohort (3.5 hours × 52 weeks × 84,129 women ÷ 8760 hours per year). It would take 7.58 days of exercise for a person to gain 3.30 days of life. Although any death in this group is premature and tragic by definition, it seems prudent to apply these correlative data with due caution.

Kenneth M. Kessler, M.D.
University of Miami School of Medicine, Miami, FL 33101

Author/Editor Response

The authors reply:

To the Editor: As Dr. Shapiro correctly points out, genetic factors clearly play a part in influencing the body-mass index. However, genetics cannot be invoked to explain the epidemic in obesity in the United States over the past couple of decades. Moreover, even among persons with a genetic predisposition to obesity, the body-mass index is still — by the inexorable laws of physics — a function of diet (calories taken in) and activity (calories expended). The finding of an independent association between the body-mass index and the risk of heart disease does not suggest that some other factor is causing both a high body-mass index and a high risk of heart disease. The simplest interpretation is that an elevated body-mass index confers an excess risk of heart disease, even after adjustment for physical activity and a limited set of dietary factors. Indeed, in randomized clinical trials of weight loss through either caloric restriction or increased physical activity, loss of adipose mass has been associated with several metabolic benefits, including improvements in lipid levels, blood pressure, and glucose tolerance.

Dr. Kessler notes, as we did in our article, that causal associations are not proved for all the factors we considered. However, some factors are supported by evidence from clinical trials (e.g., fish oils and the ratio of polyunsaturated to saturated fat), and the others are generally accepted (smoking, alcohol consumption, and physical activity) or are well supported. Moreover, a randomized trial of secondary prevention with a diet that contained many elements common to our guidelines demonstrated a marked reduction in recurrent events.1

Despite his elegant arithmetic, Dr. Kessler misses the point. First, he ignores the 832 nonfatal myocardial infarctions. Second, he considers only deaths that met our criteria for deaths due to coronary heart disease, rather than overall mortality. Many of these factors are related to the risk of diabetes, stroke, cancer at many sites, and other major diseases. Most important, Dr. Kessler does not consider that this cohort of nurses is at relatively low risk because of their age (median age at the start of follow-up, approximately 47 years) and socioeconomic status. We found that 82 percent of major coronary events could be attributed to failure to follow the guidelines we identified. If this finding can be extrapolated to the general population (which is at higher risk), then the health consequences are enormous, affecting hundreds of thousands of Americans each year.

Meir J. Stampfer, M.D.
Frank B. Hu, M.D.
Walter C. Willett, M.D.
Harvard School of Public Health, Boston, MA 02115

1 References

1
de Lorgeril M, Renaud S, Mamelle N, et al. Mediterranean alpha-linolenic acid-rich diet in secondary prevention of coronary heart disease. Lancet 1994;343:1454-1459
CrossRef | Web of Science | Medline

Citing Articles (1)