41 8 Relation of Salts to Carbohydrate Metabolism 



continue long after all evidence of ketosis had disappeared and after a simpli- 

 fied diet with a low K content had been substituted for the ordinary diabetic 

 diet. The possibility that adrenal insufficiency might account for the salt crav- 

 ing was entertained, but no supporting clinical evidence for this complication 

 was obtained. 



After a preliminary period of observation had demonstrated that the degree 

 of glycosuria in this case, as in the previous one, tended to be inversely pro- 

 portional to the quantity of NaCl ingested, this phase of the problem was 

 investigated more extensively under well-controlled conditions. For the sake 

 of comparison, similar studies were also made on four other juvenile diabetics, 

 who manifested no abnormal taste for salt. 



Studies were carried out on but two patients at a time. The latter were kept 

 in a small metabolic ward under the constant supervision of special nurses, 

 who collected the urine quantitatively and made certain that meals, salt 

 allowances, and insulin were received by the experimental subjects every six 

 hours precisely as ordered. For the sake of uniformity and accuracy, the basic 

 diet with low Na and K content was made up from powdered whole milk, 

 egg white, egg yolk, unsalted butter, cane sugar, fresh lemon juice (for vitamin 

 C), powdered yeast (for vitamins of the B complex), and water. The mineral 

 constituents of a typical day's diet, containing protein 64, fat 96 and carbo- 

 hydrate 132 gm. were as follows: Na, 1.03; K, 1.34; Ca, 0.68; Mg, 0.16; CI, 1.83; 

 P, 0.93 and S, 0.76 gm. 



When used, insulin was given at 6-hour intervals in doses sufficiently small 

 to permit some degie of glycosuria in each subdivision of the day, when the 

 patient was on the basic diet. A significant change in the amount of sugar 

 excreted in the mine during an experiment was then regarded as a reliable 

 criterion of an effect produced by the extra salt ingested. The latter was given 

 with or immediately after each of the four meals every day in the form of a 

 1 per cent solution or in gelatin capsules. Water was given ad libitum. 



The results of this study may be summarized briefly as follows: Ingestion of 

 excessive amounts (1 to 2 gm. per kilogram of body weight) of NaCl or other 

 sodium salts repeatedly reduced the total output of sugar in the urine of all five 

 patients. As illustrated in figure 1, glycosuria was at times reduced to as little 

 as one-fifth of what it was during control periods. Such an effect was either not 

 apparent or was much less marked, however, when but small amounts of NaCl 

 were ingested or when the ordinary diabetic diet was substituted for the basic 

 experimental diet (low in K). The fasting blood-sugar level was distinctly 

 lower and the respiratory quotient was somewhat higher during periods of 

 excessive Na intake. In the severely diabetic patients, ketosis was found to 

 develop earlier after withdrawal of insulin therapy during periods of ex- 

 tremely low NaCl intake than during periods of excessively high intake. 

 Glycosuria increased rapidly in severe diabetics upon withdrawal of insulin 

 in spite of the high intake of sodium. However, unless the insulin dosage was 

 adjusted to a level to allow hyperglycemia when the low-Na basic diet was 



