The glucose tolerance tests in routine clinical use are predominantly of the oral type, in which the sugar is administered by mouth. Such tests, however, suffer from the disadvantage that the rate of intestinal absorption has an effect on the results, and its variations may entirely overshadow any possible changes in carbohydrate metabolism. This difficulty may, of course, be obviated by administering the sugar by vein, and intravenous tolerance tests have actually been used by many workers (1)(2)(3)(4)(5)(6)(7)(8) 2This is not meant to be a complete bibliography. Tunbridge and Allibone (3) list no less than 37 publications dealing with intravenous glucose tolerance tests. sion rates ranged from 155 to 630 mg. of glucose per minute in different experiments. In a typical experiment the infusion lasted 60 minutes. The blood sugar of the subjects was determined in samples of capillary blood by means of the Hagedorn-Jensen method (9), and the samples were generally drawn 10,20, 30, 40, 50, and 60 minutes after the start of the infusion and in about onehalf of the experiments also 30 and 60 minutes after the end of the infusion. Urine samples were collected at the end of the infusion and often also 30 and 60 minutes later. The subjects remained recumbent during the whole experiment.
RESULTS
Mean blood glucose curve during the infusionThe above-described glucose tolerance tests were carried out in some thirty patients in all. After the exclusion of diabetics and a few other patients suffering from diseases known to affect carbohydrate metabolism, twenty-four patients remained whose carbohydrate metabolism was presumably normal. Their results were averaged, and the course of the mean blood glucose value during the infusion is shown in Figure 1. The mean fasting blood glucose prior to the infusion was 84.9 mg. per dl.8 This initial basal value was used as the zero level, and the curve shows the mean increases above this basal value after specified time intervals. It is seen that the blood glucose initially rises rapidly, then more and more slowly and seems to be approaching a plateau at the end of the infusion. When the plateau is reached, the rate of elimination obviously equals the rate of infusion.It is of considerable interest to examine whether the blood glucose curve shown in Figure 1 could be given some relatively simple mathematical interpretation. It is known that certain substances are removed from the blood and other compartments of the body through which they are distributed at a rate defined by a first-order reaction. 8 dl. = deciliter = 100 ml.
452