Does eating a lot of sugar "cause" diabetes?
2-4 DIETARY FIBER, GLYCEMIC LOAD AND RISK OF NON-INSULIN-DEPENDENT DIABETES MELLITUS IN WOMEN
Attention has focused on the hypothesis that dietary factors which increase insulin resistance or insulin demand would, over the long term, influence the risk of NIDDM. This prospective study examined the relationship between glycemic diets, low fiber intake, and risk of NIDDM.
Conclusion: The study supported the hypothesis that diets with a high glycemic load and a low cereal fiber content increase risk of NIDDM.

STUDY

1. A longitudinal study of diet and life style factors in relation to chronic diseases (the Nurses Health Study) enrolled over 65,000 women age 40-65 in 1986. All were free of cardiovascular disease, cancer, and diabetes at baseline.
2. All completed a detailed dietary questionnaire from which the investigators calculated usual intake of dietary fiber, glycemic load, and the dietary glycemic index.
3. The glycemic index is a quantitative assessment of foods based on the incremental glucose response and insulin demand they produce for a given amount of carbohydrate. (Insulin output may vary between foods containing the same amount of carbohydrate.)
4. Follow-up—6 years. Main outcome measure—incidence of NIDDM.

RESULTS

1. Over 900 incident cases of diabetes were documented over 6 years.
2. After adjustment for multiple compounding factors, the dietary glycemic index was positively associated with risk of diabetes. Comparing the highest with the lowest quintile, the relative risk (RR) of NIDDM was 1.4.
3. The glycemic load (an indicator of global dietary insulin demand) was also positively associated with diabetes (RR = 1.5).
4. Cereal fiber intake was inversely related to risk of diabetes—RR = 0.7 when comparing the extreme quintiles.
5. When compared with a low glycemic load and a high cereal fiber intake, the combination of a high glycemic load and a low cereal fiber intake further increased risk of diabetes. (RR = 2.5).
6. There were significant inverse associations with cold breakfast cereal and yogurt, and significant positive association with cola beverages, white bread, white rice, and potatoes.

DISCUSSION

1. "In this large-scale prospective study we observed that diets with high glycemic load and low cereal fiber content were positively associated with risk of NIDDM, independent of other dietary factors and currently known risk factors."
2. Previous studies have consistently found little relationship between total carbohydrate intake and risk of NIDDM. Using total carbohydrate intake, however, does not take into account the glycemic effect or insulin demand of various forms of carbohydrates. Metabolic studies have documented differences in insulin demand generated by various foods containing the same amount of carbohydrate, depending largely on the type or degree of digestibility of the starch content. Foods with a higher carbohydrate digestibility generate a higher insulin demand. Thus, the quality as well as the quantity of carbohydrate must be considered.
3. The glycemic index of some foods: white bread, 100; mashed potatoes, 104; cola beverages, 89; apples, 65; orange juice 65; broccoli, 45; peanut butter, 40; dark bread, 58-70; yogurt, 35.
4. Diets with a high glycemic load as well as a low cereal fiber content are likely to lead to a chronic high demand for insulin. As long as the pancreas is able to augment insulin secretion to meet the extra demand, glucose tolerance remains normal. If the pancreas fails to respond adequately, (relative insulin deficiency) glucose intolerance ensues, leading to NIDDM.

CONCLUSION
The study supports the hypothesis that diets with a high glycemic load and a low cereal fiber content increase risk of NIDDM in women. Grains should be consumed in a minimally refined form to reduce the incidence of diabetes.
JAMA February 12, 1997; 277: 472-77 Original investigation, first author from Harvard School of Public Health, Boston, MA.

[index]