Body Muscle-to-Fat Ratio, Rather Than Fat-to-Muscle Ratio, Significantly Correlates With Measured Insulin Resistance in Patients With Type 2 Diabetes Mellitus

BACKGROUND: Insulin resistance (IR) assessment is important in treating type 2 diabetes mellitus (T2DM). We thus compared body muscle-to-fat ratio (BMFR) and fat-to-muscle ratio (FMR) values against M/I values as clinical index of IR. METHODS: Subject included 118 untreated T2DM patients. Hyperinsulinemic-euglycemic clamp examination was performed to calculate the M/I as index of IR. Body composition was measured by impedance analysis using InBody770. RESULTS: Simple linear regression analyses confirmed correlations between M/I and BMFR (B: 0.756 (P < 0.01), coefficients of determination (R(2)): 0.572, mean absolute error (MAE): 3.19, and root mean squared error (RMSE): 4.14), and between M/I and FMR (B: -0.601 (P < 0.01), R(2): 0.362, MAE: 3.97, and RMSE: 5.05). Against the M/I values, BMFR also showed better goodness-of-fit than did FMR. In comparing correlation coefficients, the BMFR absolute B value was significantly larger than that of FMR (P = 0.027). CONCLUSIONS: BMFR is more useful than FMR in quantifying IR in patients with T2DM because the correlation between BMFR and the insulin sensitivity index M/I is significantly greater than that between FMR and M/I.