Tohoku J. exp. Med, 1969, 99, 309-313

Glucose Tolerance Test Performed in the Afternoon

Takayoshi Toyota and Yoshio Goto

Department of Medicine (Prof. S. Yamagata), Tohoku University School of Medicine, Sendai

The availability of afternoon tolerance test for the diagnosis of mellitus was investigated. Fifty grams glucose tolerance test was performed twice in 52 patients of various diseases, i.e., one test was made in the morning and another was done in the afternoon after taking an ordinary break fast alone. No significant difference was seen between the two tests. Hence, we can use the glucose tolerance test for the diagnosis of mild diabetes in the afternoon instead of the test in the early morning without any change of criteria. There was also no significant difference in the plasma response to glucose ingestion when it was given in the morning or in the afternoon.

The most sensitive tool to detect a mild diabetic state today is the glucose tolerance test. The test is made usually in the morning after an overnight fast. In the clinical practice, however, the performance of the test in the afternoon is sometimes more convenient for outpatients not only for the medical but also for the social reasons. Similar situation must be considered also in diabetes mass survey. For these reasons, the present study was performed to see whether the glucose tolerance test would give different results when it was made in the morning or in the afternoon and further to see whether the plasma insulin response to the glucose ingestion is different under these conditions.

SUBJECTS AND METHODS

Fifty-two patients with various diseases were selected for this study. They consisted of 46 males and 6 females, and their ages ranged from 22 to 78 years. Two kinds of glucose tolerance tests were performed in all of them, i.e., one test was made in the morning and the other was done in the afternoon. The test procedure was as following.

Morning glucose tolerance test. After overnight fasting, 50g of glucose in 250ml of water were given orally. Capillary was collected from the ear lobe in the fasting state and every thirty minutes for two hours after the glucose ingestion. Blood sugar was determined by the micromethod of Fujita and Iwatake.1 Afternoon glucose tolerance test. About five hours after an ordinary breakfast, 50g glucose tolerance test was applied in the same way. This test was made within a week after the performance of the morning glucose tolerance test. In six cases, two-dose glucose tolerance test was made in the morning instead of one dose morning test. In this test, 50g of glucose were given twice with an interval of two hours, and blood specimens were collected every thirty minutes for four hours. The two-

Received for publication, August 4, 1969. 309 310 T. Toyota and Y. Goto

dose glucose tolerance test was previously made to see the Hamman-Hirschman phenome non2 or Staub effect,3 although it is rarely used in Japan since our observations.4,5 The plasma insulin content was estimated in five cases and changes of plasma insulin content after the glucose ingestion was compared between the morning and the afternoon tests. The specimens used for insulin assay were obtained by antecubital venipuncture prior to, and 30, 60 and 120 minutes after the glucose ingestion. Plasma insulin was extracted from 2ml of plasma by the modified method of Jephcott6 and determined by the radioimmunoassay method of Morgan and Lazarow.7 Our criteria for the capillary blood sugar of glucose tolerance test are: normal, showing values at fasting not exceeding 120mg/dl, peak values not exceeding 170mg/dl, values at two hours after glucose ingestion not exceeding 120mg/dl (normal zone), and showing not glycosuria at any time; diabetic, showing the peak values more than 200mg/dl and the two hour values more than 140mg/dl (diabetic zone); borderline, showing the patterns not satisfying the above two criteria. If a peak value was in the range from 170 to 200mg/dl, or if a two-hour value fell between 120 and 140mg/dl, we regarded these values to be in a borderline zone. Therefore, seven kinds of the borderline pattern can be classified accord ing to the combination of the peak values and the two-hour values, as shown in Table 1. In this table is represented as DN. The details of our criteria for the glucose tolerance test was described in a previous paper.8 TABLE 1. Classification of glucose tolerance curves according to combination of the peak values and two-hour values

RESULTS

Fifteen of the eighteen cases showing a diabetic blood sugar curve in the morning glucose tolerance test, showed also a diabetic curve in the afternoon glucose tolerance test as shown in Table 2. Three cases exhibited a borderline pattern but none showed a normal pattern in the afternoon test. TABLE 2. Correlation beween the results of the glucose tolerance tests Afternoon Glucose Tolerance Test 311

TABLE 3. Mean blood sugar values during the glucose tolerance test

(Mean•}SE, mg/dl) TABLE 4. Two dose glucose tolerance test and afternoon glucose tolerance test

Of the twenty-seven cases showing a borderline curve in the morning test, thirteen cases gave a borderline curve, nine cases a diabetic curve, and five cases showed a normal curve in the afternoon test. Of the seven cases whose morning tests were normal, two were normal and five were in a borderline in the afternoon test. There was no case whose test changed from normal to diabetic or from diabetic to normal, whether the test was made in the morning or in the afternoon. Of the twenty-seven cases whose morning tests were revealed to be in a borderline, nine cases showed a diabetic and three cases had a diabetic morning test. This figure seems to indicate that the afternoon test is more convenient to unmask a latent diabetes. Statistical analysis, however, showed no significant difference between the two tests (Table 3). It is generally accepted that in normal subjects, glucose is assimilated more easily after a certain period of carbohydrate intake than in the fasting state. This phenomenon is well known as Hamman-Hirschman phenomenon or as Staub effect and the mechanism of this phenomenon was clarified by one of the authors. Hence, a question arises as to whether the afternoon test has a similar meaning as that of the second dose curve in the two-dose glucose tolerance test. Then, the two-dose glucose tolerance test was performed in six cases. The results are summarized in Table 4. In Case 3 the morning test (first dose) was in a borderline and the Staub effect was positive, but her afternoon test showed a diabetic pattern. Case 6 exhibited a negative Staub effect but his afternoon test was slightly improved as compared with the morning test. Therefore no clear relation ship was seen between the Staub effect and the change in the pattern of the 312 T. Toyota and Y. Goto

TABLE 5. Changes in plasma insulin and blood sugar during the glucose tolerance test

tolerance curve in the afternoon test. The levels of plasma insulin during the glucose tolerance tests in the morning and the afternoon were compared as summarized in Table 5. In four of the cases (No. 1-4), the increase in plasma insulin after the glucose ingestion was greater in the morning than in the afternoon, even though no definite difference in blood sugar curve was seen between the two tests. Therefore, no discrimination in diagnostic value was seen between the glucose tolerance test made in the after noon and that made in a fasting state in the early morning.

DISCUSSION Roberts9 emphasized that an afternoon glucose tolerance test was a simple and rational test to detect a slight metabolic aberration such as seen in early diabetes mellitus. He recommended this test from the following points: In early diabetic subjects, insulin reserve is limited, so that their insulin deficiency becomes manifest in the afternoon. Postprandial , one of the symptoms of early diabetes, becomes intensified during the afternoon and early evening. The present study showed that there is no essential difference between the two tests. None of those showing a normal morning test was diagnosed as diabetic by the afternoon test and vice verse. Our study, however, failed to demonstrate that the afternoon test is more useful and more sensitive than the morning test in the diagnosis of mild diabetes. In spite of numerous reports on blood insulin level of diabetic patients, our knowledge on diurnal rhythm of plasma insulin is still incomplete. The present study could not provide any conclusion as to whether the would secret more insulin in the afternoon than in the morning in response to glucose ingestion .

References

1) Fujita, A. & Iwatake, D. Bestimmung des echten Blutzuckers ohne Hefe. Biochem . Z., 1931, 242, 43-60. 2) Hamman, L. & Hirschman, I. I. Studies on blood sugar: IV. Effects upon the blood sugar of the repeated ingestion of glucose. Bull. Johns Hopk . Hosp., 1919, 30, 306 - 310. 3) Staub, H. Bahnung im intermediaren Zuckerstoffwechsel . Biochem. Z., 1921, 118, 93- Afternoon Glucose Tolerance Test 313

102. 4) Goto, Y. Significance of the two-dose glucose tolerance test . Criticism of the Staub effect. Metabolism, 1955, 4, 323-332. 5) Goto, Y. Significance of intravenous double sugar loading test. Criticism of the Staub effect. Tohoku J. exp. Med., 1956, 64, 199-208. 6) Jephcott, C.M. Extraction of insulin and stability of various preparations . Trans. roy. Soc. Can., sect V, 1931, 25, 183-186. 7) Morgan, C. R. & Lazarow, A. Immunoassay of insulin using a two antibody system . Proc. Soc. exp. Biol. Med., 1962, 110, 29-32. 8) Goto, Y. Mild diabetes and borderline cases. J. Jap. diabetic Soc. (Jap.), 1964, 7, 75-82. 9) Roberts, H. J. Afternoon glucose tolerance testing: A key to the pathoge_iesis, early diagnosis and prognosis of diabetogenie hyperinsulinism. J. Amer. Gerixt. Soc., 1964, 12, 423-471.