Metabolic Abnormalities in Starvation Diabetes *

Credit for having first described the phenomenon which was later to be termed "starvation diabetes" or "hunger diabetes" is properly ascribed to Claude Bernard. In his Lefons sur les proprietes physiologiques et les alte'rations pathologiques des liquides de l'organisme, published in 1859, he writes:' "It is possible to make an animal diabetic by making it absorb carbohydrate under certain circumstances: namely if sugar is given to it after 24 or 36 hours fast." His evidence for this statement is still older, since the report of one of his experiments showing glycosuria in a rabbit fed carrots after several days of fasting, is dated 1846. In 1873, and again in 1877, in his Lefons sur le diabete he stresses this observation:" "We can say that we are all potentially diabetic ... If a man or an animal is fasted for some time, and then given a good meal with an abundance of carbohydrate, glucose will appear in the urine." It had been known for a century or more that mild glycosuria of short duration is occasionally encountered in otherwise healthy persons. But Bernard was probably the first to recognize as a specific entity the glycosuria appearing when a period of starvation is broken by carbohydrate intake. Lehmann,44 as a preliminary to an investigation of the effect of arsenate on diabetes mellitus, performed some animal experiments on glycogen formation in the liver. Upon injecting glucose into a mesenteric vein of rabbits starved for 48 hours, he found to his surprise that no increase in liver glycogen occurred, and that glucose appeared in the urine. This result was contrary to the result of experiments by Claude Bernard"1 and by Schopffer.70 When the experiment was repeated on animals not starved, however, he found that in this condition no glycosuria appeared, while a slight increase in liver glycogen could be noted. Curiously enough, he did not relate his observation to Bernard's demonstration of starvation diabetes, although he seems to have been well acquainted with Bernard's work. Lehmann's result, published as