Effect of diet on incidence of diabetes in nonobese diabetic mice.

Nonobese diabetic/Lt mice exhibit a diabetes incidence greater than 70% in females at 30 wk of age. In studies designed to see whether increased dietary carbohydrate, fat, or protein influenced the severity or age at onset of the syndrome, we fed semipurified AIN-76 diet adulterated with increased amounts of these ingredients. Surprisingly, all AIN-76-based diets greatly reduced the expected incidence of diabetes at 30 wk. In addition, a hypoallergenic infant formula, Pregestimil, containing casein hydrolysate in place of protein, completely prevented diabetes up to 1 yr of age. To assess how dietary components might modulate the diabetes incidence, we adulterated standard AIN-76 diet with skim milk, gluten, brewer's yeast, or a natural-ingredient rodent open-formula mouse diet (Old Guilford 96 [OG96]. No increase in diabetes incidence was seen with skim milk (10%) or wheat gluten (10%), whereas brewer's yeast (10%) and OG96 (25%) added to AIN-76 increased the incidence compared to mice fed OG96 only. The diabetogenic factor or factors in OG96 could be extracted by chloroform plus methanol (2:1), leaving little activity in the residue. We conclude that diet is a critical factor in diabetes development and that unknown chloroform-methanol-soluble substances in natural-ingredient chow not found in semipurified diets can enhance the development of diabetes in genetically susceptible mice.