Diet modifies the development of insulin-dependent diabetes mellitus i
n animals and in humans. We examined female non-obese-diabetic (NOD) m
ice, a diabetes-prone mouse strain with 70% spontaneous diabetes incid
ence and metabolic abnormalities in non-overtly diabetic litters. They
were fed a diet containing 55% (n=27) or 15% (n=26) protein, respecti
vely, after weaning. At an age of 30 weeks, non-diabetic NOD mice were
submitted to an intravenous glucose tolerance test (0.5 g/kg body wei
ght; blood samples were taken after 2, 4, 8, 10, 15, 20 and 30 min) an
d to perfusion of the pancreas (stimulation media were Krebs-Ringer-He
pes buffer with 5 mmol/l glucose, 30 mmol/l glucose and 5 mmol/l gluco
se plus 19 mmol/l arginine). Diabetic mice were removed from the exper
iment. Serum glucose concentration and body weight were monitored week
ly. Food ingestion was checked at an age of 11 weeks. On average, the
onset of diabetes was diagnosed in mice on a high-protein diet (19.7+/
-1.3 weeks) 4 weeks earlier than in mice on a low-protein diet (23.5+/
-1.1 weeks; P<0.05). Non-diabetic NOD mice on a high-protein diet show
ed significantly better glucose tolerance (as determined by the glucos
e disappearance rate) and mean insulin secretion (at 30 mmol/l glucose
), No difference in the serum glucose concentration between non-diabet
ic mice on the low-protein diet or high-protein diet could be proved.
In non-diabetic mice on the high-protein diet the body weight and food
ingestion exceeded those of mice on the low-protein diet (P<0.05). Hi
gh insulin secretion and glucose tolerance in non-diabetic mice may re
flect the capacity of beta-cells to adapt; however, beta-cells tend to
be destroyed under such circumstances. Thus, a high-protein diet prom
oted the onset of diabetes, but it did not increase significantly the
incidence of the disease.