点击显示 收起
Brigham and Women's Hospital, Harvard Medical School, Division of Preventive Medicine, 900 Commonwealth Avenue East, Boston, MA 02215-1204
Dear Sir:
We appreciate Schneeman's comments regarding our article. As noted, women with high dietary glycemic loads tended to be a health-conscious group (1). However, eating large amounts of low-quality carbohydrate, as reflected by a high dietary glycemic load, appeared to increase the risk of coronary heart disease (CHD) in these women, independent of healthy choices such as smoking less and consuming more dietary fiber and vitamins.
Any carbohydrate-containing food can induce plasma glucose responses, and dietary glycemic load (the amount of carbohydrate multiplied by its glycemic index) represents the quality and quantity of carbohydrate and the interaction between the 2. The interaction implies that carbohydrate quality, represented by glycemic index, should have a greater biological effect when the amount of carbohydrate consumed is large than when the amount is small. Also, because a common standard referent food—white bread—was used to standardize all carbohydrate-containing foods, we essentially compared the relative associations of different glycemic responses from these foods with CHD risk. Because a higher intake of dietary fiber, vitamin E, or folate was each independently associated with a lower risk of CHD (2), it is important to evaluate the association with glycemic load or glycemic index with adjustment for these other dietary factors. In relation to glycemic index, the Spearman correlation coefficients were –0.20 for dietary fiber, -0.23 for dietary folate, and –0.32 for vitamin E, indicating that the overall glycemic index was inversely associated with the intakes of these micronutrients that are thought to be protective against CHD. Thus, in a multivariate model that included dietary fiber, folate, and vitamin E, the independent association between glycemic load and CHD risk was even stronger (Table 2 of our article).
Cooking methods can have some influence on glycemic index; thus, we used average values for the ways that foods are usually consumed (eg, potatoes are eaten cooked and apples are eaten raw). Meal patterns may affect the absolute glycemic response but do not affect the relative differences between foods (3–5). Metabolic studies using standardized methods indicated that the correlation between the glycemic index of mixed meals and the average glycemic indexes of individual component foods ranges from 0.84 to 0.99 (5–7). Even though the total quantity of the glycemic and insulinemic effects of foods may not be fully captured by dietary glycemic load, these measurement errors were likely to have been modest and unrelated to CHD because diets were assessed before disease occurred. Recently, in a random sample of 185 postmenopausal women in the Nurses' Health Study who provided fasting blood samples, we found a strong positive relation between dietary glycemic load assessed by a food-frequency questionnaire and fasting triacylglycerol concentrations (8), a well-established relation from metabolic studies (9).
Metabolic experiments suggested that the adverse metabolic responses to a high dietary glycemic load, including hyperinsulinemia, hypertriglyceridemia, and low HDL-cholesterol concentrations, are strongly related to an individual's underlying degree of insulin resistance (10). Thus, our observation of a stronger positive association between dietary glycemic load and CHD risk in overweight women highlights the importance of considering the physiologic effects of carbohydrate quality in the context of other metabolic variables. Judged by its abilities to predict physiologic responses as well as clinical endpoints, glycemic index appears to represent a more informative means than the conventional simple versus complex approach in classifying carbohydrates.
REFERENCES