Relationship of insulin secretion and glycemic response to dietary intervention in non-insulin-dependent diabetes

Forty-two obese subjects with non-insulin-dependent diabetes mellitus had their plasma insulin, C peptide, and glucose levels measured after an overnight fast and in response to a 75-g oral glucose loading. Subjects were then prospectively followed up with dietary treatment, and the same measurements were repeated at 1 year. Although insulin values tended to be lower with greater fasting hyperglycemia at baseline, no correlation was observed among three parameters. However, near-normalization of glycemia (measured as the level of hemoglobin A₁) was associated with significantly higher fasting and stimulated plasma insulin concentrations. Sixteen subjects were matched to each other for equivalent baseline hyperglycemia (by glycosylated hemoglobin) and divided into group 1 (normalization of the hemoglobin A₁ value to 7.0% ± 0.3% [mean ± SE]) and group 2 (persistent hyperglycemia) (hemoglobin A₁ value, 10.7% ± 0.7% [mean ± SE]). Before dietary therapy, the plasma insulin concentrations were twofold to threefold higher in group 1, and despite similar degrees of weight loss, group 2 failed to demonstrate improved glycemia. We concluded that the outcome of diet therapy for non-insulin-dependent diabetes mellitus is dependent on the duration of diabetes and endogenous insulin secretory reserve. There is a subgroup of patients with non-insulin-dependent diabetes mellitus in whom delayed dietary intervention may have a beneficial effect.