HOMA‐IR and QUICKI: decide on a general standard instead of making further comparisons

Abstract

Aim: To limit further comparisons between the two fasting indices Homeostasis Model Assessment for Insulin Resistance (HOMA‐IR) and Quantitative Insulin Sensitivity Check Index (QUICKI), and to examine their robustness in assessing insulin sensitivity. Methods: A total of 191 obese children and adolescents (age 13.9 ± 2.9 years, BMI SDS 6.1 ± 1.6), who had undergone a Frequently Sampled Intravenous Glucose Tolerance Test (FSIVGTT), were included. Receiver operating characteristic curve (ROC) analysis was used to compare indices in detecting insulin resistance and Bland–Altman plots to investigate agreement between three consecutive fasting samples when compared to using single samples. Results: ROC analysis showed that the diagnostic accuracy was identical for QUICKI and HOMA‐IR [area under the curve (AUC) boys 0.80, 95%CI 0.70–0.89; girls 0.80, 0.71–0.88], while insulin had a nonsignificantly lower AUC (boys 0.76, 0.66–0.87; girls 0.75, 0.66–0.84). Glucose did not perform better than chance as a diagnostic test (boys 0.47, 0.34–0.60; girls 0.57, 0.46–0.68). Indices varied with consecutive sampling, mainly attributable to fasting insulin variations (mean maximum difference in HOMA‐IR −0.8; −0.9 to −0.7). Conclusions: Using both HOMA‐IR and QUICKI in further studies is superfluous as these indices function equally well as predictors of the FSIVGTT sensitivity index. Focus should be on establishing a general standard for research and clinical purposes.