The current explanation of this phenomenon (Macleod) is that the first dose of glucose sensitizes the insulin-secreting mechanism, so that in response to the second dose the islet cells secrete insulin more readily and more abundantly at a lower level of hyperglycaemia. On the basis of this explanation Sweeney, in 1927, attempted to explain the variations in sugar tolerance found in normal subjects on different diets. Using the ordinary glucose tolerance test as a guide, he investigated the sugar tolerance of healthy individuals during starvation, on a fat diet, on a protein diet, and on a carbohydrate diet. He found that protein had little effect; that fat diets and starvation diminished sugar tolerance; and that carbohydrate diets improved sugar tolerance. Sweeney considered that the diminished sugar tolerance was due to the impaired sensitivity of the insulin-secreting apparatus, consequent upon the absence of the stimulus of carbohydrate ingestion, and that the improved tolerance was the result of the increased sensitivity of this mechanism, owing to greater stimulation.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2444943/pdf/brmedj07161-0009.pdf

December 1927

DIETARY FACTORS THAT INFLUENCE THE DEXTROSE TOLERANCE TEST - A PRELIMINARY STUDY

J. SHIRLEY SWEENEY, M.D.

Abstract

The dextrose tolerance test is now being extensively employed as a diagnostic procedure. It is most beneficially used in the differentiation of mild diabetes mellitus and renal diabetes. It is also being used, and is believed to be of diagnostic value, in many pathologic conditions, such as encephalitis, malignant tumor, pituitary and thyroid dysfunctions and nephritis.

 Although it is definitely established as a diagnostic procedure, there is some diversity of opinion concerning what constitutes a normal response to the oral administration of dextrose. Some writers state that in a healthy person there may be a postprandial rise in blood sugar of from 14 to 16 per cent and a return to the normal within two hours. There are other writers who consider a postprandial hyperglycemia of 20 per cent within normal limits. It is generally believed that the persistence of the postprandial hyperglycemia is of more diagnostic significance than the degree of hyperglycemia. In early cases of diabetes the blood sugar curve rises higher, stays up for a longer time and does not return to normal for several hours. Macleod says that "slight deviations from the normal must not be given too much weight in diagnosis, since they may occur in other diseases or even in perfectly normal persons." All who have studied dextrose tolerance curves have noted the variability exhibited by normal persons, to say nothing of those who are diseased. These variations have been discussed and explained in different ways. 

It occurred to me that perhaps the character of the food and the amount of water that a person had been consuming for a few days prior to the time the tolerance test was made might be factors that would influence the dextrose tolerance curve. If these factors should prove to be capable of altering a tolerance curve, they could be controlled. This would eliminate some of the confusing variability that is so frequently observed. It was these thoughts that lead to the following experiments. 

Young, healthy, male medical students were used to study the effect of different preceding diets. Four groups were formed. The subjects in one group were given a protein diet, those in another a fat diet, those in a third a rich carbohydrate diet, and those in the fourth group were not given any food—the starvation group. Those on the protein diet received only lean meat and the whites of eggs. The students on the fat diet received only olive oil, butter, mayonnaise made with egg yolk, and 20 per cent cream. Those in the group fed on carbohydrates were allowed sugar, candy, pastry, white bread, baked potatoes, syrup, bananas, rice and oatmeal. These diets were followed for two days. Meals were taken at the usual hours, and eating between meals was allowed, provided the diets were followed. Those in the starvation group did without food for two days. 

On the morning of the third day, each student was given by mouth 1.75 Gm. of dextrose per kilogram of body weight, on an empty stomach. Determinations of blood sugar were made from samples of venous blood removed immediately before the dextrose was given, and at 30, 60 and 120 minute intervals following its administration. I made all determinations of blood sugar by the Folin-Wu method.

A better comparison of these groups is obtained by examining table 5 and chart 5 in which are contained the average or type curves of each group. It will be noted that those students who were on the carbohydrate diet exhibited a marked increase in sugar tolerance and those on a protein diet a slight decrease in tolerance, while those who were placed on the fat diet and those who were starved manifested a definite decrease in sugar tolerance. The differences in the average fasting blood sugars are noteworthy. The blood sugar in those of the protein and starvation groups was distinctly lower than that of the members of the fat and carbohydrate groups. 

Because of the great difference in these groups, those students on the fat diet and those in the starvation group who showed the most extreme responses were placed on the carbohydrate diet. Similiarly, those in the carbohydrate group who showed an extreme response were placed on starvation restriction. This was obviously done to determine whether the curve of a person could be changed significantly by diet. The results are presented in table 6 and in charts 6 and 7. 

Comparison of the curves of these five students is striking. The curves of all who had been placed on carbohydrate diets manifested a definite increase in their sugar tolerance. When three of these (the three most extreme) were placed on starvation restrictions, the curves were notably abnormal ; there was a marked postprandial hyperglycemia. which persisted at the end of two hours ; in other words, what was an increased sugar tolerance following the carbohydrate diet became a definitely decreased tolerance following two days of starvation. The remaining two persons who were placed on the fat diet showed a similar decreased tolerance. It should be stated that an interval of at least one week was allowed between the tolerance tests performed on the same subject.