There is some consensus that high fat diets can contribute to the development of obesity and type 2 diabetes in humans and animals. An increase in dietary fat has been shown to produce obesity and diabetes in mice; such diet-induced diabetes can be reversed by reducing the fat in the diet. In humans, there is some evidence that low-fat diets can produce acute improvements in blood sugar control in type 2 diabetes-even in the absence of weight loss. In most human studies, however, dietary fat reduction has been accompanied by a reduction in total calorie intake. It is thus not possible to separate the effects of these 2 metabolic changes. The purpose of this study is to gather preliminary information on the effect of a very-low-fat diet on blood metabolism in persons with type 2 diabetes. The design incorporates controlled feeding procedures, and 30 men and women with type 2 diabetes will be given all foods for 4 weeks--a 2-week diet standardization period (diet composition: 35% fat, 15% protein, 50% carbohydrate), followed by a 2-week experimental diet period. The experimental diet conditions are A) continuation of the moderately-high-fat standardization diet, or B) a very-low-fat diet composed of 10% fat, 15% protein, 75% carbohydrate. Outcomes will be measured after the standardization and the experimental periods. The primary outcome variable is fasting plasma glucose; secondary outcomes are fasting insulin, carbohydrate (meal) tolerance, insulin secretion and blood lipids. In addition, we will gather descriptive data on the potential acceptability and utility of a very-low-fat diet constructed using the fat substitute, olestra (sucrose polyester). There are no results yet.