点击显示 收起
Yale-Griffin Prevention Research Center, Yale School of Medicine, 130 Division Street, Derby, CT 06418, E-mail: katzdl{at}polnet
Dear Sir:
The results of the study by Liu et al (1) regarding glycemic load and risk of coronary heart disease (CHD) among participants of the Nurses' Health Study are both fascinating and timely. The debate over the role of carbohydrate in efforts to promote health and control weight has perhaps never been more intense. This study, suggesting that CHD risk is directly related to the glycemic load of the diet, could be used as evidence that recommendations for high intakes of complex carbohydrate are injudicious. However, the information provided is insufficient for reaching this conclusion.
First, although the study showed an association between glycemic load and CHD risk, no consideration is given to the potential bidirectionality of that association. Specifically, with prevailing recommendations to restrict fat intake in the advent of CHD or its risk factors, women who developed angina, anginal equivalents, or overt CHD risk factors might have subsequently reduced their fat intake and substituted carbohydrate. As a result, high glycemic load might appear to be linked to CHD risk, but the causality would actually be in the opposite direction. If the authors have pertinent information, this issue should be addressed.
Other limitations are worth noting. In citing references concerning the adverse effects of carbohydrate on the lipid profile (2–4), the authors were selective and potentially biased, omitting studies that showed beneficial effects of high carbohydrate intake when fiber content was high (5, 6). In Liu et al's study, fiber intake in proportion to total carbohydrate intake was lower in the higher quintiles of glycemic load, as shown in Table 1; this fact is inadequately discussed. Limited discussion is devoted to the neutral effects of dietary starch displayed in Table 3. Finally, the authors imply in their discussion that fruit and vegetables, because of their content of simple sugars, are sources of simple rather than complex carbohydrate. This questionable classification biases the study results against complex carbohydrate and in favor of the conclusions reached by the authors.
As is often true in nutritional epidemiology, this report raises more questions than it answers. That said, one might question whether it is prudent and justified to use the results of this observational study to challenge dietary guidelines that the bulk of evidence suggests would, if implemented, reduce the population burden of CHD.
REFERENCES