Methods for quantifying adipose tissue insulin resistance in overweight/obese humans

BACKGROUND/OBJECTIVES: Insulin resistance of adipose tissue is an important feature of obesity-related metabolic disease. However, assessment of lipolysis in humans requires labor-intensive and expensive methods, and there is limited validation of simplified measurement methods. We aimed to validate simplified methods for the quantification of adipose tissue insulin resistance against the assessment of insulin sensitivity of lipolysis suppression during hyperinsulinemic-euglycemic clamp studies. SUBJECTS/METHODS: We assessed the insulin-mediated suppression of lipolysis by tracer-dilution of [1,1,2,3,3-H-2(5)] glycerol during hyperinsulinemic-euglycemic clamp studies in 125 overweight or obese adults (85 men, 40 women; age 50 +/- 11 years; body mass index 38 +/- 7 kg m(-2)). Seven indices of adipose tissue insulin resistance were validated against the reference measurement method. RESULTS: Low-dose insulin infusion resulted in suppression of the glycerol rate of appearance ranging from 4% (most resistant) to 85% (most sensitive), indicating a good range of adipose tissue insulin sensitivity in the study population. The reference method correlated with (1) insulin-mediated suppression of plasma glycerol concentrations (r = 0.960, P < 0.001), (2) suppression of plasma non-esterified fatty acid (NEFA) concentrations (r = 0.899, P < 0.001), (3) the Adipose tissue Insulin Resistance (Adipo-IR) index (fasting plasma insulin-NEFA product; r = -0.526, P < 0.001), (4) the fasting plasma insulin-glycerol product (r = -0.467, P < 0.001), (5) the Adipose Tissue Insulin Resistance Index (fasting plasma insulin-basal lipolysis product; r = 0.460, P < 0.001), (6) the Quantitative Insulin Sensitivity Check Index (QUICKI)-NEFA index (r = 0.621, P < 0.001), and (7) the QUICKI-glycerol index (r = 0.671, P < 0.001). Bland-Altman plots showed no systematic errors for the suppression indices but proportional errors for all fasting indices. Receiver-operator characteristic curves confirmed that all indices were able to detect adipose tissue insulin resistance (area under the curve >= 0.801, P < 0.001). CONCLUSIONS: Adipose tissue insulin sensitivity (that is, the antilipolytic action of insulin) can be reliably quantified in overweight and obese humans by simplified index methods. The sensitivity and specificity of the Adipo-IR index and the fasting plasma insulin-glycerol product, combined with their simplicity and acceptable agreement, suggest that these may be most useful in clinical practice.