We conducted a study among 15 714 Dutch middle-aged women consuming modest-glycemic-load diets. Results showed that high dietary glycemic load and glycemic index increased the risk of cardiovascular disease. This association was particularly evident among overweight women. Recommendations to follow a high-carbohydrate diet may therefore not be optimal in the prevention of cardiovascular diseases.
Studies showing that the risk of cardiovascular disease is increased among individuals on a high-saturated-fat diet have resulted in recommendations to follow a low-total-fat and low-saturated-fat diet. This consequently has led to an increase in the intake of carbohydrates.1 Glucose and lipid metabolism, however, may be negatively affected by this type of diet.2 Based on the idea that saccharide chain length determines the rate of digestion and absorption, carbohydrates were classified as either simple or complex until recently.3 Different blood glucose and insulin responses result from dietary carbohydrates with different structures.4 Carbohydrates that are absorbed quickly and cause high glucose and insulin levels after eating have a high glycemic index. Glycemic load is calculated as the product of the glycemic index of a specific food and its carbohydrate content, and represents both quantity and quality of carbohydrates.5 Depending on its carbohydrate content, a food can thus have a low glycemic load but a high glycemic index.
Diabetes,5 hyperlipidemia,1 and perhaps being overweight6 are associated with dietary glycemic load; therefore, glycemic load may be associated with cardiovascular disease. Few studies have been conducted investigating this relationship, especially among individuals consuming moderate-glycemic-load diets. An increased risk of coronary heart disease with increasing glycemic load, especially among women who were overweight, was shown in one study,7 although another study by van Dam and colleagues did not show such an association.8 In another study, an association was shown between the risk of stroke and glycemic load among overweight women only.9 We conducted a study among women eating a moderate-glycemic-load diet to determine the association between dietary glycemic load and cardiovascular disease, and whether body mass index (BMI) influenced this relationship.
Subjects and methods
A total of 17,357 women were recruited for the European Prospective Investigation into Cancer and Nutrition (Prospect-EPIC) study.10 The women were being screened for breast cancer and were aged between 49 and 70 years. For the current analysis, we included 15,714 women who were free of coronary heart disease, cerebrovascular disease, and diabetes at baseline. A 10% random sample was taken from the original cohort, using the same inclusion criteria, resulting in a total of 1417 subjects. Cardiac biomarkers were assessed among these women.
The women completed a questionnaire on demographic characteristics and the existence of chronic diseases and risk factors, as well as a food-frequency questionnaire. A physical examination to obtain anthropometric measurements and blood pressure was performed. To determine the occurrence of cardiovascular events, we examined the National Medical Registry for hospital discharge diagnoses and the National Death Registry for causes of death. We used the food-frequency questionnaire to assess dietary glycemic load and glycemic index of the participants. The Cox proportional hazards model was used to estimate the associations of quartiles of glycemic load and glycemic index with cardiovascular disease, coronary heart disease, and cerebrovascular events. These models were adjusted for hypercholesterolemia, smoking, BMI, age, alcohol use, hypertension, total physical activity, systolic blood pressure, menopausal status, hormone replacement therapy, oral contraceptive use, intake of total energy and multivitamins, and energy-adjusted intake of folate, fiber, protein, and vitamin E, as well as saturated, monounsaturated, and polyunsaturated fat. The association of dietary glycemic load and glycemic index with markers of cardiovascular disease was established in secondary analyses using linear regression.
A total of 556 incident cases of fatal or nonfatal coronary heart disease and 243 incident cases of a fatal or nonfatal cerebrovascular event occurred over a follow-up period of 9 years. Table 1 shows the associations of glycemic load and glycemic index with cardiovascular disease. High dietary glycemic load increased the risk of cardiovascular disease after adjusting for cardiovascular disease risk factors and nutritional factors, including fat intake, with a hazard ratio (HR) for the highest compared with the lowest quartile of glycemic load of 1.47 (95% confidence interval [CI], 1.04-2.09; P for trend = .03). There was also a relationship between dietary glycemic index and cardiovascular disease (HR = 1.33; 95% CI, 1.07-1.67) for the highest versus the lowest quartile (P for trend = .02).
Comparable relationships were shown for coronary heart disease (HRQ4 vs Q1 = 1.44; 95% CI, 0.95-2.19; P for trend = .14) and cerebrovascular events (HRQ4 vs Q1 = 1.55; 95% CI, 0.81-2.97; P for trend = .10) in separate analyses, after adjusting for the full model. There was a greater relationship between glycemic index and coronary heart disease (HRQ4 vs Q1 = 1.44; 95% CI, 1.10-1.89; P for trend = .01) than between glycemic index and cerebrovascular events (HRQ4 vs Q1 = 1.12; 95% CI, 0.75-1.69; P for trend = .61).
No association was shown between cardiovascular disease risk and total carbohydrates, monosaccharides, and disaccharides (data not shown). As shown in Table 2, BMI was shown to possibly alter the association between glycemic load and cardiovascular disease in analyses of lean and overweight women adjusted for the full model; this was also shown to a lesser extent with glycemic index. Among women of normal weight, no association was shown between cardiovascular risk and glycemic load (HRQ4 vs Q1 = 1.14; 95% CI, 0.67-1.93; P for trend = .43). The association was greatest among women who were overweight (HRQ4 vs Q1 = 1.78; 95% CI, 1.11-2.85; P for trend = .04), with a P value for interaction of .19. Body mass index did not have an effect on the relationship between cardiovascular risk and glycemic index.
There was an inverse relationship between high-density lipoprotein (HDL) cholesterol levels and glycemic load (ß = -0.005 ± 0.001; P < .001) and between HDL cholesterol levels and glycemic index (ß = -0.83 ± 0.30; P = .006), whereas low-density lipoprotein (LDL) cholesterol levels were associated with glycemic load (ß = 0.005 ± 0.003; P = .12). No important relationships were shown for C-reactive protein, total cholesterol, and glucose levels.
A high dietary glycemic load and high glycemic index were shown to be independently associated with an increased risk of cardiovascular disease among Dutch women who ate a moderate-glycemic-load diet in our study. The same association was also shown for coronary heart disease and cerebrovascular events. Glycemic index was more strongly associated with coronary heart disease. Among women who were overweight, glycemic load was especially related to cardiovascular disease.
Our study was the first to show an increased risk of cardiovascular disease among women on a moderate-glycemic-load diet. In one study, the risk of coronary heart disease was doubled among US women in the highest glycemic load quartile.7 An increase in coronary heart disease was also shown for overall dietary glycemic index. A positive association was shown between total stroke and glycemic load only for overweight US women in another study, but this association did not occur with glycemic index.9 The findings of our study concur with those of other studies that included analyses for coronary heart disease and cerebrovascular events. Among US women, however, the mean energy-adjusted dietary glycemic load ranged from about 100 in the lowest quartile to 200 in the highest quartile, whereas glycemic load ranged from 79 in the lowest quartile to 122 in the highest quartile among Dutch women. Even when corrected for the use of white bread as a reference in the United States, the figures still suggested lower glycemic loads of the diets in the Netherlands than in the United States. Also in these lower ranges, glycemic load and glycemic index were both associated with cardiovascular risk, although not to the same degree.
In the Zutphen Elderly Study, no association was shown between glycemic index and cardiovascular risk.8 Furthermore, no increased cardiovascular risk was shown with high dietary glycemic load in a case-control study.11 In that study, however, relative risks were not adjusted for fat intake, which may have affected the findings, and the odds ratio of 1.38 (95% CI, 0.95-2.00; P for trend = .10) for cardiovascular disease was similar to our findings.
Several previous studies have shown more pronounced associations between glycemic load and risk of coronary heart disease or stroke among overweight women.7,9,11 Although the interaction effect in our study was not statistically significant, the results were consistent.
It is not known how glycemic load increases cardiovascular risk; however, blood lipid profile and insulin resistance most likely play a role. Carbohydrates with high glycemic loads cause significant increases in blood glucose levels and insulin responses after eating. The increase in insulin production causes the blood glucose level to decrease, resulting in a state of hunger.12 This results in further consumption of carbohydrates and increased postprandial glycemia. These reactions may play a role in obesity and insulin resistance over time.13 Obesity and insulin resistance result in decreased HDL cholesterol and increased LDL cholesterol and triglyceride levels. Some intervention studies have shown decreased LDL cholesterol and triglyceride levels among individuals following a low-glycemic-index diet.14 A recent study showed that both high-protein and high-carbohydrate diets increased body fat loss, but a low-glycemic-index diet also optimized blood lipid profile.15 Results from our study confirm these relations with blood lipid profile.
High dietary glycemic load and glycemic index were shown to increase the risk of cardiovascular disease in our study of 15,714 Dutch middle-aged women. This relationship was particularly strong among overweight women. These results suggest that recommendations to follow a high-carbohydrate diet may not be ideal for the prevention of cardiovascular disease.