102 METABOLISM IN SEVERE DIABETES. 



It is safe to say, therefore, that sudden and rapid fluctuations in body-weight 

 of several pounds may invariably be ascribed to changes in the water-content 

 of the body. Persistent and regular losses in body-weight may be logically 

 considered as conclusive evidence of insufficient available calories in the diet. 

 This insufficiency may be due to a non-assimilation of carbohydrates, or to a 

 defective oxidation of fat resulting in the formation of /3-oxybutyric and allied 

 acids. In any event, such a persistent and regular loss shows that the patient 

 is not getting sufficient food for daily maintenance and drafts are being made 

 upon body-material to supply the deficiency. 



Nitrogen Excretion of Diabetics without Food. 



Inasmuch as certain forms of artificially induced diabetes, such, for 

 example, as pancreatic diabetes in dogs, are accompanied by an increase in the 

 total disintegration of nitrogenous tissue, it is of interest to note to what extent 

 nitrogenous material is broken down and excreted in the body of severe dia- 

 betics when food is not being taken. Accordingly, during the period of every 

 experiment without food, the urine was collected usually from the first hour of 

 rising in the morning until the end of the experimental period some time in the 

 middle of the forenoon or later. This sample of urine, which thus represented 

 the urine excreted during the experimental period, was analyzed both for nitro- 

 gen and for sugar, and it is evident that the total excretion of nitrogen here may 

 be taken as a measure of the normal excretion of nitrogen of the individual 

 when not actively engaged in processes of digestion. 



The determinations of the nitrogen on the basis of milligrams per hour per 

 kilogram of body-weight are given for all the later experiments in table 117. 

 The average value is 8.4 mg. per kilogram per hour. In the first series of 

 experiments, the average for all cases was 9.4 mg. per hour. The average for 

 all the experiments from 1908 to 1912 gave, with diabetics both light and severe, 

 8.9 mg. of nitrogen per kilogram of body-weight per hour. 



In the earlier publication the average for the relatively few normal sub- 

 jects there used gave 6.85 mg. of nitrogen per hour, and it was explained that 

 the difference between the normal value and that found with diabetics could 

 not by any means explain the enormous increase in metabolism found with 

 these subjects. Since this time a large number of other individuals have been 

 studied and the results obtained with 14 of these, which are given in table 118, 

 show on the average a nitrogen excretion of 6.8 mg. per hour per kilogram of 

 body-weight, a value identical with that given in the earlier report. 



As an examination of the various charts will show, the diabetic subjects 

 were not given excessively rich nitrogenous diets, nor, indeed, was the total 

 nitrogen in the urine excessively large in any case. In conformity with the 

 advice of Cantani, Naunyn, and others, we have purposely kept our patients 

 upon a low-protein diet. This has been done hoping thereby to render the 

 task more simple of keeping the urine free from sugar, or at least of keeping the 

 sugar at a low level. The low protein was prescribed not because of the pre- 

 formed sugar in the protein molecules, but because of the tendency for sugar to 



