Evaluation of HOMA and QUICKI as measures of insulin sensitivity in prepubertal children

Abstract: Background: Simple fasting sample methods to measure insulin sensitivity (S_I) such as homeostasis model assessment (HOMA) and quantitative insulin‐sensitivity check index (QUICKI) have been widely promoted in adult studies but have not been formally evaluated in children. The aim of this study was to compare HOMA and QUICKI to the minimal model as measures of S_I in prepubertal children. Method: The study population consisted of twins (n = 44), premature (n = 17), small for gestational age (SGA) (15), and normal (n = 3) prepubertal children. The insulin‐sensitivity index derived by the minimal model (S_IMM) was calculated by the minimal model with plasma glucose and insulin data from a 90‐min frequently sampled intravenous glucose test with tolbutamide. The HOMA resistance index (R_HOMA) and QUICKI were calculated from fasting plasma glucose and insulin values. Results: The correlation between R_HOMA and S_IMM (r = ?0.4, p < 0.001) was no better than that between fasting insulin and S_IMM (r = ?0.4, p < 0.001). QUICKI was poorly correlated with S_IMM (r = 0.2, p = 0.02). The correlation between S_IMM and R_HOMA is largely confined to low S_I values (<10 × 10^?4/min µU/mL). In seven SGA subjects, the introduction of growth hormone treatment led to an expected fall in S_IMM by 8.2 ± 2.8 × 10^?4/min µU/mL (p = 0.02) that was not detected by either R_HOMA (p = 0.1) or QUICKI (p = 0.2). Similarly, S_IMM values were lower in obese (n = 9) compared to non‐obese subjects (p = 0.04); however, no difference was found between these two groups with either R_HOMA (p = 0.21) or QUICKI (p = 0.8). Conclusion: As measures of S_I in prepubertal children, R_HOMA is no better than fasting insulin and QUICKI, a poor measure. Neither R_HOMA nor QUICKI was able to detect changes in S_I induced by either obesity or growth hormone therapy.