Accuracy of HOMA-IR in Clinical Practice

Abstract

Purpose: Insulin resistance (IR) is closely associated with NASH and cardiovascular mortality in the general population. There is also data suggesting that patients with hepatitis C and IR have lower response rates to the treatment with Pegylated Interferon and Ribavirin. HOMA-IR (Homeostasis Model Assessment of Insulin Resistance) is an attractive screening tool for insulin resistance in clinical practice. The formula includes fasting insulin (mU/L) multiplied by fasting glucose (mg/dL) then divided by 405. Values of HOMA-IR> 2 are strongly correlating with the presence of IR. This observational study evaluates the accuracy of HOMA-IR in our hospital clinical lab. Methods: 32 subjects with high risk for IR were screened with OGTT and HOMA-IR for a diet and exercise study. Blood was drawn after 12 hours overnight fast and was sent to the main hospital clinical lab for measurement of insulin and glucose (speckled top tube). At the same time plasma was immediately centrifuged and glucose was measured in real time using Glucose Analyzer (Beckman, CA). The rest of separated plasma was stored at −70°C until later analysis in our obesity research clinical lab (ORC). HOMA-IR was calculated using same lab values and the results were split in two groups: the ORC data and the clinical lab data. Results: The clinical lab data identified only 14 (43.75%) subjects with HOMA> 2 compared with the ORC lab data, which had 28 (87.5%). The range of HOMA-IR measured by clinical lab was 0.34-6.19 (mean=2.32) and that measured by ORC lab was 1.38-10.92 (mean=5.09). The average difference between insulin and glucose values of the two groups was 8.44 mU/L and 20.56 mg/dL respectively. Five subjects had fasting glucose levels over 120 mg/dL in the ORC lab data compared with none over 106 mg/dL in the clinical lab data. If only clinical lab data had been used, 14 (50%) subjects would have been screened out for IR and new onset diabetes would have been missed in five subjects. Conclusion: Insulin and glucose degradation is likely the explanation for the observed differences between the two groups. Inaccuracies in the measurement of fasting insulin and glucose in clinical practice will likely underestimate HOMA-IR. There is data in literature suggesting that in a red top tube glucose concentration may decrease with 25% over 3 hours due to hemolysis. As a result, we conclude that only special gray top tubes (with Fluoride-Oxalate as anti-hemolytic) should be used for fasting glucose measurement. Also, standardized methods for insulin measurement should be developed (i.e. colect in lavender top tube with EDTA, refrigerate until processed and centrifuge sample as soon as possible) in order to prevent insulin degradation and increase the accuracy of the results.