[Kinetic analysis in the evaluation of glucose tolerance disorders. Evaluation and use of a new method in the diagnosis of glucose tolerance in clinical practice].
A kinetic model of oGTT has been used. This model is characterized by 7 glycaemia collections (0, 30, 45, 60, 120, 180 min). In some cases this is supplemented by determination of C-peptide and insulin values (0, 60, 120 min). This method is very simple and highly useful in clinical practice because it gives information about physiological stimulation by enterohormones, the first glucose passage through liver and next glucose utilization. The first part of the study deals with random error of the kinetic model of oGTT found by repeated calculations (> 10,000 repeatings) of identical initial glycaemias in individual groups (DM, PGT, N). Random error of glucose clearance (the most suitable parameter) ranged within 0.2 -2.1% in individual deviations. Then the identical calculations were made but loaded with glycaemia and with certain error in individual collection intervals (glycaemia 0.1, 0.2, 0.5 mmol/l; collection with 30, 60, 300 s). Random error of the method increased significantly with dispersion variance of glycaemias (maximum 17% with glucose clearance at glycaemia dispersion variance of 0.5 mmol/l); changed time intervals less affected random error quantity. As acceptable and frequent deviation in practice was determined glycaemia dispersion variance of 0.2 mmol/l (corresponding to a total analytical error of glycaemia measurement) and dispersion variance of time intervals of 60 s. At these values, random error of the method increased maximally to 10.3%. Glucose has also a biological variability (not published yet for individual time intervals), the value of random error may be higher but will not achieve half of values of random error obtained at the classical oGTT. The second part of the study deals with comparison of both the tests (classical oGTT, kinetic curve of oGTT) in 126 probands examined at metabolic and diabetologic out-patient department of the Hospital in Sternberk. It can be concluded that using a classical oGTT, 60% of patients were classified into incorrect groups as defined by dynamic results obtained by analysis of the kinetic model. Interesting enough is the fact that almost half of persons who cannot be classified by the classical oGTT had impaired glucose tolerance. Out of them, 15.6% had diabetes mellitus; almost 20% of normal patients also had impaired glucose tolerance. Over 1/5 of persons with impaired tolerance for glucose according to the classical oGTT was found by the kinetic model to have diabetes mellitus. On the contrary, 1/10 of diabetic patients diagnosed by oGTT had normal glucose clearance, over 1/2 of patients had only impaired glucose tolerance. Then in 126 persons random error of the method was again calculated (> 2,000 repeated calculations for each proband) amounting 6.6%.