Evaluating a pragmatic estimate of insulin sensitivity in Latino youth with obesity

The whole-body insulin sensitivity index (WBISI) is a widely used surrogate of insulin sensitivity estimated from glucose and insulin concentrations during an oral glucose tolerance test (OGTT). The original WBISI uses five timepoints but reduced sampling models have been proposed. These reduced models have not been applied to the paediatric population. The purpose of this study is to compare cross-sectional agreement and changes in response to lifestyle intervention between the original WBISI and the WBISI₁₂₀ using fasting and 2-hour glucose and insulin concentrations from OGTT among Latino adolescents with obesity. We also examined the cost-differential between the two measures. Secondary analyses were conducted with data from Latino adolescents (14-16 years) with obesity (BMI ≥ 95th percentile) who were recruited for a randomized controlled lifestyle intervention trial. Baseline data in youth (n = 148) who underwent a 2-hour OGTT was used to analyse agreement between WBISI and WBISI₁₂₀. Data from 65 participants who completed a 12-week lifestyle intervention were used to assess changes and the rate of change between WBISI and WBISI₁₂₀. Research costs to determine both measures were compared. WBISI₁₂₀ showed good cross-sectional agreement (ICC_agreement = 0.88) with the full WBISI. Following intervention, WBISI₁₂₀ increased 62.5% (M ± SD, 1.6 ± 1.2 to 2.6 ± 1.7, P < .001) while WBISI increased by 25.0% (1.6 ± 1.0 to 2.0 ± 1.0, P < .001) but the rate of change for WBISI and WBISI₁₂₀ was not significantly different (P = .11). WBISI₁₂₀ costs ~70% less than WBISI. WBISI₁₂₀ may offer a cost-effective surrogate estimate of insulin sensitivity in Latino youth with obesity.