Access to all articles, new health classes, discounts in our store, and more!
Biochemical Standards of Health: I.V. Three Hour Oral Glucose Tolerance Test
Published in Acta Diabetologica Latina, Vol. IV, No. 3, August-September 1967.
* * *
Introduction
The increasing recognition of the importance of carbohydrate metabolism in seemingly nondiabetic disease states such as peptic ulcer,5 carcinomatosis,4,10,18,24 multiple sclerosis,8,13,15,23 atherosclerosis,1,11,26 ischemic heart disease,2,25 gout,16 and mental illness23 as well as diabetes mellitus and so-called prediabetes3 necessitates a new look at blood glucose parameters.
An attempt will be made in this report to analyze first by traditional techniques the classical three-hour oral glucose tolerance pattern in a group of presumably healthy persons, the type usually employed in deriving a “normal range.” The same data will then be reexamined by a method which permits the identification of the blood glucose scores in a relatively healthier subgroup.
Methods
One hundred and forty-seven dental practitioners and their wives (members of the Southern Academy of Clinical Nutrition) shared in this experiment. The age and sex distribution in table 1 shows a predominance in the fourth decade. Without a preparatory diet, the three-hour oral glucose tolerance test was performed after a 100 g glucose load. The blood glucose determinations (fasting, one-, two- and three hours) were made by the Autoanalyzer method. The procedure utilizes the potassium ferricyanide-potassium ferrocyanide oxidation reduction reaction. Additionally, each subject completed the Cornell Medical Index Health Questionnaire (CMI).6 This 195-question form is self-administered and the subject responded to each question by encircling a negative or positive designation. Positive replies ranged from a low of zero to a high of 53 (table 2).
Results
The means and standard deviations for the number of positive answers on the questionnaire and the glucose tolerance pattern are listed (table 3). For the entire sample (n = 147), there were 15.0 ± 11.4 affirmative replies. The blood glucose findings fasting (80.0 ± 11.5), one-hour (128.5 ± 40.1), two-hours (94.8 ± 37.5), and three-hours (76.2 ± 30.9) are within the recognized conventional limits for health. The means (figure 1) and standard deviations (figure 2) are pictorially portrayed.
On the assumption that the fewer the symptoms and signs, the greater the likelihood of health, the 11 subjects with the fewest (0 to 2) positive CMI replies were studied. Table 3 shows 1.9 ± 0.3 affirmative CMI answers in this select group. This is obviously statistically different from the findings for the entire sample. The means and standard deviations for blood glucose are also listed (table 3) and the means (figure 1) and standard deviations pictured (figure 2). Four points warrant special mention. First, the means for the entire sample and the relatively healthier group are very similar (figure 1).
Fig. 1–The mean glucose tolerance scores (interrupted line) for the entire sample (n = 147) and the smaller subgroup (n = 11) with less than three positive CMI responses (continuous line).
Second, in only one instance (fasting) the means are statistically significantly different (P<0.001). Third, the standard deviations differ from 0.5 mg per cent (11.5 versus 11.0) under fasting conditions to 20.5 mg per cent (17.0 versus 37.5) at two hours (figure 2).
Fig. 2–The range (one standard deviation) for the glucose tolerance pattern for the entire sample (n = 147) represented by the interrupted lines and the smaller subgroup (n = 11) with less than three positive CMI replies pictured by the continuous lines.
Lastly, the variances are significantly different at both the two- and three-hour intervals (table 3).
Discussion
The techniques presently employed for deriving physiologic parameters possess two common characteristics. First: the definition of health is very arbitrarily set. For example, Mosenthal and Barry17 describe their selection of healthy persons as “…the present effort concerns itself solely with normal individuals. Fifty tests were carried out. They were divided roughly into three age groups…only ambulant persons were used.” Camerini-Davalos and his colleagues7 utilized a somewhat different specification, “…nondiabetics were selected on the basis of normal glucose tolerance tests initially in individuals with no history of diabetes in relatives.” It is clear from these representative citations that health is very grossly defined.
Second: the specific values for the physiologic glucose tolerance have been established on a very arbitrary basis. Thus, Mosenthal and Barry17 concluded, from mean values only, that “we believe that the criteria for true blood sugar in venous blood are: 100 mg per cent or less in fasting, a limit of 150 for the height of the curve, and 100 or less for the two-hour period.” Additionally, the point should be made that only maximal values are reported. The presumption, obviously, is that there are no minimal physiologic limits. On the other hand, others derive standards from means and two standard deviations. Camerini-Davalos and his co-workers7 stated, “for venous blood (Somogyi-Nelson technic) glucose values considered the upper limit of normal were the mean, plus 2 S.D. of values obtained in fifty normal persons: fasting, mg 100; 0.5 h, mg 160; 1.0 h, mg 145; 1.5 hrs mg 125; 2.0 hrs, mg 110; and 3.0 hrs, mg 105.” The presumption in the latter instance is that 95 per cent (two standard deviations) of a given sample of individuals without diabetic relatives provides the physiologic range for the glucose tolerance test pattern. On this basis, it must be agreed that dental caries is physiologic since 95 per cent of a random sample of the population possesses dental decay. Also, the concept of two standard deviations ignores the fact that diabetes mellitus in particular and chronic disease in general exists in a spectrum.
The hypothesis set forth here is that the range of the glucose tolerance test pattern is more narrow (shown by the continuous lines in figure 2) for a group characterized by relatively greater health as judged by relatively fewer symptoms and signs. Other studies of carbohydrate metabolism, utilizing different techniques for arriving at relatively greater health have also indicated that the physiologic spread is much more narrow than held by traditional techniques for deriving health parameters. In a group of 100 routine dental patients, the fasting blood glucose range was found to progressively shrink as one eliminated evidence of oral pathosis such as oral symptoms (xerostomia, stomatodynia, and gingival tenderness), gingival pathosis, edentulousness, clinical tooth mobility, and alveolar bone loss.9 The range of fasting blood glucose in a larger sample of 290 patients, utilizing both systemic and oral findings, was also found to progressively constrict to within a relatively narrow range.21 By the method of developing the symptomless and sign free individual, the ranges for both the classical19 and the cortisone20 glucose tolerance test patterns were observed to be more constricted than generally recognized by the usual methods of standardization.
The hypothesis set forth in these studies suggests that the physiologic range for the most popular measure of carbohydrate metabolism, the glucose tolerance test, is more narrow than that presently employed. As a hypothesis, it should, of course, be tested. In this regard, limited published reports are available to test the thesis. For example, data obtained from 401 glucose tolerance tests at the Mayo Clinic12 were analyzed with special reference to the correlation between the fasting level of true blood glucose and the frequency of normal, equivocal, and abnormal (diabetic) glucose tolerance curves. It was demonstrated that the frequency of abnormal test curves increased with small but increasing levels of fasting blood glucose.
Of the many who have addressed themselves to this problem, none has pinpointed the issue better than Krall14 with the following statement: “The detection of diabetes can be compared to fishing with a small mesh net that increases the catch of fish but also seines some non-fish or the wrong variety of fish, as opposed to using a larger mesh which would be more specific for the size and type of fish sought but bring a smaller yield. For research and documentation purposes, the Joslin Clinic prefers to fish with as sensitive a net as possible, fully aware that all hyperglycemia is not necessarily diabetes.”
Conclusion
1) The following quote14 underscores how arbitrary are the existing criteria for the glucose tolerance test. In 1947 a group of consultants to the U.S. Public Health Service Diabetes Program collaborated in determining diagnostic levels. As with most committees, the ultimate result was a compromise hybrid which depended on a point system for diagnosis. The situation is still confused as shown in a recent public symposium on the subject “Diagnostic Criteria in Diabetes,” which emphasized the fact that it is nearly impossible to find any significant number of AA. who use similar criteria or who are even talking about the same thing;
2) it is the purpose of this report to attempt to evolve less arbitrary standards through a study of the relatively symptomless and sign-free individual;
3) the findings, within the limits of this study, suggest that the physiologic oral glucose tolerance test pattern range is more narrow than generally held. Specifically, the variance is more restricted at the two- and three-hour intervals;
4) the observations noted in this study confirm the findings of the limited number of investigators who have studied the problem.
Summary
The present physiologic blood glucose ranges for the three-hour oral glucose tolerance test are based upon the means and two standard deviations of a presumably healthy population. An attempt is made in this report to analyze the blood glucose status of a group of dental practitioners and their wives according to traditional methods. These same data are reexamined by a technique which permits the identification of the blood glucose scores in a relatively healthier subgroup of the entire sample. The findings suggest that, in the relatively symptomless and sign-free individual, the blood glucose range is more narrow than conventionally held.
Note: The Italian version of this article can be accessed via the pdf.
References Cited:
- Aleksandrow D., Ciswicka-Sznajderman M., Ignatowska H., Wocial B.: “Studies of Carbohydrate Metabolism in Atherosclerosis.” Atheroscler. Res. 2, 171, 1962.
- Antar M. A., Ohlson M. A., Hodges R. E.: “Changes in Retail Market Food Supplies in the United States in the Last Seventy Years in Relation to the Incidence of Coronary Heart Disease, with Special Reference to Dietary Carbohydrates and Essential Fatty Acids.” J. Clin. Nutr. 14, 169, 1964.
- Balodimos M. C., Hurxthal L. M.: “The Remote Prediabetic State.” Geriatrics 21, 119, 1966.
- Benjamin F., Romney S. L.: “Disturbed Carbohydrate Metabolism in Endometrial Carcinoma.” Cancer 17, 386, 1964.
- Berry M.: “Is Hypoglycemia an Etiologic Factor in Ulcer?” Rec. Med. 170, 53, 1957.
- Brodman K., Erdmann A. J., Wolff H. G.: Cornell Medical Index Health Questionnaire: Manual. Cornell University Medical College – New York 1949.
- Camerini-Davalos R. A., Caulfield J. B., Rees S. B., Lozano-Castaneda O., Naldjian S., Marble A.: “Preliminary Observations on Subjects with Prediabetes.” Diabetes 12, 508, 1963.
- Cerkez C., Chandler J. H., Chandler D.: “Glucose Metabolism in Multiple Sclerosis.” Nerv. Syst. 23, 377, 1962.
- Cheraskin E., Ringsdorf W. M. Jr.: “Physiologic Fasting Blood Glucose: Range or Point?” Dent. Med. 16, 96, 1961.
- Cheraskin E., Ringsdorf W. M. Jr., Hutchins K., Setyaadmadja A. T. S. H., Wideman G. L.: “Carbohydrate Metabolism and Cervical (Uterine) Carcinoma.” J. Obstet. Gynec. 97, 817, 1967.
- Cohen A. M.: “Fats and Carbohydrates as Factors in Atherosclerosis and Diabetes in Yemenite Jews.” Heart J. 65, 291, 1963.
- Frethem A. A.: “Clinics on Endocrine and Metabolic Diseases. 10. Relation of Fasting Blood Glucose Level to Oral Glucose Tolerance Curve.” Mayo Clin. 38, 110, 1963.
- Jones H. H., Jones H. H. Jr., Buncii L. D.: “Biochemical Studies in Multiple Sclerosis.” Intern. Med. 33, 831, 1950.
- Krall L. P.: “When is Diabetes?” Clin. N. Amer. 49, 893, 1965.
- Mcardle B., Mackenzie I. C. K., Webster G. R.: “Studies on Intermediate Carbohydrate Metabolism in Multiple Sclerosis.” Neurol. Neurosurg. Psychiat. 23, 127, 1960.
- McKechnie J. K.: “Gout, Hyperuricemia and Carbohydrate Metabolism.” Afr. Med. J. 38, 182, 1964.
- Mosenthal H. O., Barry E.: “Criteria for an Interpretation of Normal Glucose Tolerance Tests.” Intern. Med. 33, 1175, 1950.
- Pelner L.: “Host-Tumor Antagonism. 27. Nutrition and Cancer.” Amer. Geriat. Soc. 10, 701, 1962.
- Ringsdorf W. M. Jr., Cheraskin E.: “Physiologic Glucose Tolerance Test.” Prog. 2, 281, 1962.
- Ringsdorf W. M. Jr., Cheraskin E.: “Physiologic Cortisone Glucose Tolerance Test.” Med. Ass. Ala. 31, 359, 1962.
- Ringsdorf W. M. Jr., Cheraskin E.: “Biochemical Standards of Health: III. Fasting Blood Glucose.” Y. J. Dent. 33, 185, 1963.
- Salzer H. M.: “Relative Hypoglycemia as a Cause of Neuropsychiatric Illness.” Nat. Med. Ass. (N.Y.) 58, 12, 1966.
- Sawyer G. T.: “Treatment of Multiple Sclerosis with Tolbutamide.” Amer. Med. Ass. 174, 470, 1960.
- Tannenbaum A.: “Dietary Factors in Carcinogenesis.” Acta Un. Int. Cancr. 10, 117, 1954.
- Turner N. C., Gertler M. M., Cady L. D. Jr.: “Some Aspects of Carbohydrate Metabolism and Coronary Heart Disease; Amyloclastic Action of Saliva in Persons Prone to Coronary Heart Disease and in Normal Controls.” Geriatrics 17, 20, 1962.
- Waddell W. R., Field R. A.: “Carbohydrate Metabolism in Atherosclerosis.” Metabolism 9, 800, 1960.