ABSTRACT
BACKGROUND: The major adverse consequences of obesity are associated with the development of insulin resistance (IR) and adiposopathy. The Homeostasis Model Assessment-Adiponectin (HOMA-AD) was proposed as a modified version of the HOMA1-IR, which incorporates adiponectin in the denominator of the index. OBJECTIVES: To evaluate the performance of the HOMA-AD index compared with the HOMA1-IR index as a surrogate marker of IR in women, and to establish the cutoff value of the HOMA-AD. SUBJECTS/METHODS: The Brazilian Metabolic Syndrome Study (BRAMS) is a cross-sectional multicenter survey. The data from 1,061 subjects met the desired criteria: 18-65 years old, BMI: 18.5-49.9 Kg/m² and without diabetes. The IR was assessed by the indexes HOMA1-IR and HOMA-AD (total sample) and by the hyperglycemic clamp (n = 49). Metabolic syndrome was defined using the IDF criteria. RESULTS: For the IR assessed by the clamp, the HOMA-AD demonstrated a stronger coefficient of correlation (r = -0.64) compared with the HOMA1-IR (r = -0.56); p < 0.0001. In the ROC analysis, compared with the HOMA1-IR, the HOMA-AD showed higher values of the AUC for the identification of IR based on the clamp test (AUC: 0.844 vs. AUC: 0.804) and on the metabolic syndrome (AUC: 0.703 vs. AUC: 0.689), respectively; p < 0.001 for all. However, the pairwise comparison did not show evidence of superiority for the HOMA-AD in comparison with the HOMA1-IR in the diagnosis of IR and metabolic syndrome (p > 0.05). The optimal cutoff identified for the HOMA-AD for the diagnosis of IR was 0.95. CONCLUSIONS: The HOMA-AD index was demonstrated to be a useful surrogate marker for detecting IR among adult women and presented a similar performance compared with the HOMA1-IR index. These results may assist physicians and researchers in determining which method to use to evaluate IR in light of the available facilities.