A-038 Correlation of the insulin resistance index (HOMA IR) with oral glucose tolerance test (OGTT) and glycated hemoglobin (A1c) in subjects without diagnosed diabetes mellitus

Background Insulin resistance (IR) is due to an impairment in insulin action, and is the main factor related to diabetes mellitus (DM). The homeostasis model assessment of insulin resistance index (HOMA-IR) is used in clinical practice for IR evaluation, as it has a good correlation with the euglycemic clamp. Glycated hemoglobin (A1c) and the oral glucose tolerance test (OGTT) are used routinely to diagnose or to evaluate increased risk for DM. However, A1c and OGTT were not defined on the basis of the principal pathophysiological abnormalities of type 2 DM, i.e. impaired tissue sensitivity to insulin and failure of pancreatic b-cells to compensate IR. To gain a deeper insight into the matter, we correlate both, OGTT and A1c to HOMA-IR index, in non-diabetic subjects. Methods The authors studied the correlation of HOMA-IR with A1c and glucose after 2 hours of overload with 75 g of dextrosol (glucose OGTT), using a clinical laboratory database. Were included patients aged 20-65 years, both gender. Glucose, insulin, glucose OGTT and A1c were measured in the same visit. Were excluded subjects: hospitalized, using medications to treat DM, dyslipidemia, or medications that could interfere in the measurements of glucose, insulin ou A1c; with AST, ALT, or creatinine above the reference range for gender and age. Glucose was measured by enzymatic colorimetric method, insulin by electrochemiluminescence, and A1c by immunoturbidimetry. Calculation formula of HOMA-IR was [glucose (in mg/dL) x insulin (in mcIU/mL) / 405]. For the statistical study, the Shapiro Wilk test was performed to assess whether the analytes had a normal distribution. To evaluate the correlation with HOMA-IR, non-parametric Spearman*s correlation index (r value) was used. Confidence intervals (CI) for r value were calculated for a better analysis of the correlation significance. The statistical significance level was p < 0.05. Results After applying the exclusion criteria and subtracting outliers, the bank with all the data of interest resulted in 328,345 records, with 38,20 ± 8,96 years, 65,94% female, IMC = 27,14 ± 6,78 kg/m2. As the analytes did not present a Gaussian distribution, they were log-transformed. Results, in mean and standard deviations are: glucose = 88,32 ± 8,2 mg/dL; insulin = 8,78 ± 5,01 mcIU/mL; glucose OGTT = 115,69 ± 44,8 mg/dL; and A1c = 5,44 ± 0,61%. Both parameters showed a positive correlation, and both with a narrow CI, with HOMA-IR. The r value, CI and p obtained were: HOMA-IR x Glucose OGTT: r = 0,4 - IC = 0,4 – 0,44 – p < 0,001 HOMA-IR X A1c: r = 0,3 – IC = 0,24 – 0,3 – p < 0,001 Conclusion Since patients had A1c and OGTT performed on the same day and time, reflecting the same metabolic state, our data show that, although both parameters showed a positive correlation with HOMA-IR, this was more relevant with OGTT. In this large database, was a better indirect indicator of IR, comparing with HOMA-IR.