The Oral Glucose Tolerance Test (OGTT):

Transcription

1 he Oral Glucose olerance est (OG): Effect of Rate of Ingestion of Carbohydrate and Different Carbohydrate Preparations ROBER J. HEINE, M.D., IAN HANNING, B.Sc, LINDA MORGAN!, Ph.D., AND K. GEORGE M. M. ALBERI, D.Phil. he glucose load of the oral glucose tolerance test (OG) is well standardized. However, recommendations on rate of ingestion and nature of the load are vague. In this study the effect on blood glucose, serum insulin, C-peptide, and plasma gastric inhibitory polypeptide (GIP) of giving 75 g glucose in 300 ml over 1 and 10 min (Gl and G10) was investigated in six subjects. In five an isocaloric amount of partially hydrolyzed starch (Hycal) was also used (HI and H10). he fast glucose intake, compared with the slow ingestion, resulted in an earlier rise in blood glucose levels, accompanied by a faster serum insulin and C-peptide response. Between 90 and 135 min blood glucose concentrations were significantly higher after the 10-min glucose intake. At 1 min blood glucose levels were 5.5 ± 0.5 and 4-7 ± 0.5 mmol/l, respectively, for G10 and Gl (P < 0.05). In the first half hour after slow and fast Hycal intake no differences were seen in blood glucose, serum insulin, and C-peptide levels. Between 45 and 1 min blood glucose levels were significantly higher after the 10-min Hycal intake. At 1 min blood glucose levels were 5.3 ±0.2 and 4-4 ± 0.1 mmol/l, respectively, for H10 and HI (P < 0.01). Except for a faster rise in glucose and insulin levels after glucose loading in 1 min, no further differences were found, when compared with Hycal. No significant differences were seen in the GIP responses. hus differences in rates of ingestion can cause significant differences in later results. A standard time for glucose ingestion should be specified, DIABEES CARE 6: , SEPEMBER-OCOBER Considerable effort has been expended on standardizing the glucose load of the oral glucose tolerance test (OG) and the diagnostic criteria for its interpretation. 1 " 5 However, recommendations on rate of ingestion and nature of the load have been variable or vague. he World Health Organization advises a glucose load of 75 g in ml water, which should be consumed in 5-15 min 5 whereas the Australian Association of Clinical Biochemists suggests the same glucose load but ingested within 5 min. 6 he latter, in practice, often results in ingestion in 1-2 min. o our knowledge the effect of different glucose ingestion rates on glucose and insulin levels during an OG has not been investigated. here is thus little evidence to support these recommendations. Other carbohydrates such as partially hydrolyzed starch or maltose have been accepted instead of glucose as the carbohydrate load for the OG. 45 ' 7 ' 8 he advantage of maltose and hydrolyzed starch over glucose is the less frequency of side effects such as nausea and vomiting. 7 he aim of our study was to compare two different ingestion rates, one the fastest reasonable rate and the other at the midpoint of the WHO suggested range, of both a glucose and a partially hydrolyzed starch load on blood glucose, serum insulin, C-peptide, as well as on plasma gastric inhibitory polypeptide (GIP) levels, in the OG. MAERIALS AND MEHODS Subjects. Six normal volunteers were recruited, three men and three women yr old. Ideal body weight ranged from 91% to 118% (Metropolitan Life Insurance ables, 1959). None were taking any medication or known to be suffering from any disease. All volunteers maintained their usual diet before the investigations. he studies were performed at 8:30 a.m. after an overnight fast. Protocol Subjects participated in a random order in the study comparing fast (in 1 min, Gl) and slow (in 10 min, G10) intake of 75 g of glucose, dissolved in tap water to a volume of 300 ml. In five of these volunteers the study was repeated using an isocaloric amount of Hycal (Beecham Prod- DIABEES CARE, VOL. 6 NO. 5, SEPEMBER-OCOBER

2 OG: EFFEC OF RAE OF INGES1ON AND YPE OF CARBOHYDRAE/R. J. HEINE AND ASSOCIAES ABLE 1 Hycal: carbohydrate constituents imol l.min 0-15 min 0-30 min glucose area Anhydrous D glucose Disaccharides risaccharides etrasaccharides Pentasaccharides Hexa- and higher saccharides % (wt/wt) JLr i ucts, Brentford, United Kingdom) (partially hydrolyzed starch; able 1), dissolved to a volume of 300 ml, again ingested in 1 min (HI) and 10 min (H10). Free flowing venous blood samples were taken from an antecubital vein before and for 240 min after ingestion of the carbohydrate load. Sampling was every 5 min during the first half hour and quarter-hourly afterward. ests were performed at least 1 wk apart. Nausea was not felt by any subject. glucose insulin area nmol l.min 25 C-peptide area mu l insulin G1G10 H1 H10 G1 G10 H1 H10 C-peptide FIG. 2. Area under blood glucose, serum insulin, and C-peptide curves from 0 to 15 and 0 to 30 min after glucose and Hycal loading in I and 10 min: Gl, GIO, HI, and HIO. x: P < 0.05; xx: P < 0.02; xxx: P<0.0l ^==^s FIG. I. Blood glucose, serum insulin, and C'peptide response to 75'g glucose loading in I and 10 min. Gl: ---; GIO: ; x: P < 0.05; xx: P < 0.02; xxx: P <0.0l. Glucose and hormone estimations. Blood for glucose assay was immediately deproteinized in 5 ml ice-cold 5% (vol/vol) perchloric acid and estimated by a standard fluorimetric method. 9 Serum insulin was measured by double-antibody radioimmunoassay. 10 Serum C-peptide was measured by radioimmunoassay with ethanol precipitation." Plasma GIP levels were measured by radioimmunoassay. 12 Statistical analysis. Results are expressed as mean and SEM. he two-tailed Student paired t-test was used to assess sta- 442 DIABEES CARE, VOL. 6 NO. 5, SEPEMBER-OCOBER 1983

3 pg ml 1250 r GIP Comparison between 75 g glucose ingestion in 1 and 10 min. he fast glucose intake resulted in an earlier rise in blood glucose levels (Figure 1). he area under the glucose curve between 0 and 15 min was significantly greater with Gl than GlO (6.1 ± 1.4 versus 2.6 ± 0.6mmol/L min, P < 0.05) (Figure 2). his was accompanied by faster serum insulin and C-peptide response in the Gl study (Figures 1 and 2). he areas under the insulin and C-peptide curves between both 0 and 15 min and 0 and 30 min were significantly greater after Gl. he insulin areas were 173 ± 39 and 71 ± 25 mu/l min (P < 0.05) and 653 ± 99 and 469 ± 78 mu/l min (P < 0.05) at 0-15 and 0-30 min, respectively, while the corresponding values for C-peptide were 3.8 ± 0.9 versus 1.2 ± 0.3 nmol/l min (P < 0.02) and 16.2 ± 2.5 versus 11.1 ± 1.5 nmol/l min (P < 0.05). Plasma GIP levels were not significantly different with either rate of ingestion (Figure 3). Between 90 and 135 min blood glucose concentrations were significantly higher after GlO (Figure 1). At 1 min the glucose levels were 5.5 ± 0.5 and 4.7 ± 0.5 mmol/l, respectively, for GlO and Gl (P < 0.05). he areas under the glucose curves between 0 and 1 min and between 0 and 240 min were not significantly different for Gl and GlO. he insulin response, however, was significantly greater after GlO, calculated over 240 min: 4211 ± 771 versus 3654 ± 655 mu/l min (P < 0.02). otal C-peptide response (0-240) also tended to be greater after GlO, but the difference did not achieve statistical significance (Figure 4). Plasma GIP responses were similar on both occasions. Comparison between ingestion of 1 ml Hycal in 1 and 10 min. In the first 30 min no differences were seen in blood glucose, serum insulin, C-peptide, or plasma GIP responses. 0-1 min min 250 mmol l.min 300 glucose area minutes FIG. 3. Plasma GIP responses to glucose and Hycal loading in 1 and 10 min. Gl: ; GlO: ; HI: ; HJO: tistical significance of differences. Areas under the glucose and insulin curves were calculated with the formula: Area = (X n+i - X n ) (Y n + Y n+1 ) 2 where X n = sampling time in minutes at time point n and Y n = concentration of measured substance at time point n. RESULS -100 L mull.min 6000 r L insulin area C-peptide area G1 G10 H1 H10 G1G10 H1H10 FIG. 4. Area under blood glucose, serum insulin, and C-peptide curves from 0 to 1 and 0 to 240 min after glucose and Hycal loading in 1 and 10 min: Gl, GlO, HI, and H10. x: P < 0.05; xx: P < 0.02; xxx: P < 0.0J. DIABEES CARE, VOL. 6 NO. 5, SEPEMBER-OCOBER

4 mmol I glucose both from 0 to 15 and 0 to 30 min were significantly greater for Gl compared with HI (P < 0.01 for both). he insulin, C-peptide, and GIP responses over 1 and 240 min were not significantly different for glucose and Hycal loading. Between 45 and 1 min the glucose levels were signifi' cantly higher at four time points after slow Hycal intake (Figure 5). At 1 min the blood glucose levels were 5.3 ± 0.2 and 4-4 ± 0.1 mmol/l, respectively, for HIO and HI (P< 0.01). he area under the glucose curve from 0 to 1 min was significantly smaller after HI: 65.4 ± 23.8 versus ±51.6 mmol/l min for HI and HIO (P < 0.05) (Figure 4). he insulin, C-peptide, and GIP responses were not significantly different, either from 0 to 1 min or from 0 to 240 min (Figures 3 and 4). Comparison between glucose and Hycal. he area under the glucose curve between 0 and 15 min was significantly greater for Gl compared with HI (6.1 ± 1.7 and 2.4 ± 1.6 mmol/ L min, P < 0.02). No significant differences were found between G10 and H10 (Figure 2). he area under the insulin curve between 0 and 15 min was significantly greater for Gl compared with HI (168 ± 47 and 118 ± 45 mu/l min, respectively, P < 0.05). he areas under the C-peptide curves DISCUSSION he results of this study indicate that the rate of intake of the carbohydrate load and to a lesser degree the nature of the load influence the results of the OG. he earlier glucose response after fast glucose intake might be explained by the time needed for digestion of poly- and disaccharides in the intestinal lumen and cellular brush border, in contrast to the direct glucose absorption by a special carrier mechanism Except insulin for the faster rise in glucose and insulin levels after glucose loading in 1 min, no further differences were found when compared with Hycal. GIP responses were similar. GIP has been recognized as an important insulinotropic hormone. 15 However, it seems only to act as such when there is slight hyperglycemia. 16 he fast insulin response after the 1-min glucose intake seems to be unrelated to GIP but dependent entirely on the rate of glucose rise. he 2-h glucose value, crucial now as a diagnostic criterion, was strongly influenced by the rate of intake of the carbohydrate load. he lower 2-h glucose value after the fast glucose intake might partially be explained by the earlier insulin response; however, no differences in insulin response were seen when slow and fast Hycal intake were compared. he area under the glucose curve from 0 to 1 min was 0.4- significantly greater after the 10-min Hycal intake. With a similar insulin response it may be assumed that the amount minutes of glucose absorption was greater after slow Hycal ingestion. his suggests that the fast intake of an hyperosmolar fluid, FIG. 5. Blood glucose, serum insulin, and C-peptide response to 1mainly consisting of polysaccharides, impairs glucose absorp- ml Hycal in I and 10 min. HI: ; HIO:. x: P < 0.05; xx: P < 0.02; xxx: P <0.0l. tion, possibly by increasing the rate of intestinal transit. 13 he areas under the glucose curves after glucose intake were not significantly different, perhaps because of the faster absorption of glucose. he 2-h glucose value is thus influenced by the rate of intake of the carbohydrate load, possibly by the early insulin response, as seen with fast glucose loading and/or by the influence on glucose absorption with fast Hycal ingestion. hus we have shown that the results of the OG can be influenced by the rate of ingestion of the carbohydrate load. We chose to compare the fastest practical rate of ingestion, 1 min, with the midpoint of the WHO suggested range, 10 min. It is possible that if 5 min had been compared with 15 min smaller differences would have been found. he effect of different rates of ingestion in diabetic subjects was also not studied. However, it seems reasonable to conclude that a standard time for the rate of ingestion of the carbohydrate load should be specified and used uniformly, both for diagnostic purposes and in epidemiologic surveys. From the Department of Clinical Biochemistry and Metabolic Medicine, Royal Victoria Infirmary, Newcastle-upon-yne, NEl 444 DIABEES CARE, VOL. 6 NO. 5, SEPEMBER-OCOBER 1983

5 4LP United Kingdom (R.J.H., I.H., K.G.M.M.A.), and the Department of Clinical Biochemistry, University of Surrey, Guildford, United Kingdom (L.M.). REFERENCES 1 Nobel, E. de, and Laar, A. van't: he size of the loading dose as an important determinant of the results of the oral glucose tolerance test. A study in subjects with slightly impaired glucose tolerance. Diabetes 1978; 27: Forster, H., Haslbeck, M., and Melmest, H.: Metabolic studies following the oral ingestion of different doses of glucose. Diabetes 1972; 21: Peterson, D.., and Reaven, G. M.: Evidence that glucose load is an important determinant of plasma insulin response in normal subjects. Diabetes 1971; : National Diabetes Data Group: Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 1979; 28: W.H.O. Expert Committee on Diabetes Mellitus: W.H.O. echnical Report Series 646. World Health Organization, Geneva, Garcia-Webb, P., and Bonser, A. M.: Glucose or glucose monohydrate for glucose tolerance tests? Letter to the editor. Diabetologia 1981; 21: Harano, Y., Sakamoto, A., Izumi, K., Shimizu, Y., Hoski, M., Schichizi, M., Shigeta, Y., Ohgaku, S., and Abe, H.: Usefulness of maltose for testing glucose tolerance. Am. J. Clin. Nutr. 1977; 30: Chandalia, H. B., and Boshell, B. K.: Diagnoses of diabetes. he size and nature of carbohydrate load. Diabetes 1970; 19: Lloyd, B., Burrin, J. ( Smythe, P., and Alberti, K. G. M. M.: Enzymatic fluorimetric continuous flow assays for blood glucose, lactate, pyruvate, alanine, glycerol and 3-hydroxybutyrate. Clin. Chem. 1978; 24: Soeldner, J. S., and Slone, D.: Critical variables in the radioimmunoassay of serum insulin using the double antibody technic. Diabetes 1965; 14: Heding, L. G.: Radioimmunological determination of human C-peptide in serum. Diabetologia 1975; 11: Ekins, R. P.: Radioimmunoassay design. In Radioimmunoassay in Clinical Biochemistry. Pasternak, C. A., Ed. London, Heyden, 1975: Losowsky, M. S., Walker, B. E., and Kelleher, J.: Physiology and biochemistry of intestinal absorption. In Malabsorption in Clinical Practice. London, Churchill Livingstone, 1974: Gray, G. M.: Carbohydrate digestion and absorption. Gastroenterology 1970; 58: Dupre, J., Ross, S. A., Watson, D., and Brown, J. C: Stimulation of insulin secretion by gastric inhibitory polypeptide in man. J. Clin. Endocrinol. Metab. 1973; 37: Anderson, D. K., Elahi, D., Brown, J. C, obin, J. D., and Andres, R.: Oral glucose augmentation of insulin secretion. Interactions of gastric inhibitory polypeptide with ambient glucose and insulin levels. J. Clin. Invest. 1978; 62: DIABEES CARE, VOL. 6 NO. 5, SEPEMBER-OCOBER