Table 4.
Regression coefficients from robust multiple linear models for the prediction of HbA1c, fasting insulin resistance (HOMA2-IR) and OGTT-derived insulin sensitivity (Matsuda Index) at follow-up
| HbA1c | Insulin resistance (HOMA2-IR) | Insulin sensitivity (Matsuda index) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| β | 95% CI | p value | R 2 | β | 95% CI | p value | R 2 | β | 95% CI | p value | R 2 | |
| Body fat mass (kg) | ||||||||||||
| Baseline fat mass (kg) | 0.01 | 0.00–0.01 | 0.008 | 0.07* | 0.03 | 0.01–0.05 | 0.002 | 0.07* | −0.07 | −0.12—0.02 | 0.005 | 0.04* |
| Change fat mass (kg) | 0.00 | −0.01–0.01 | 0.868 | 0.02 | −0.01–0.05 | 0.280 | −0.06 | −0.15–0.02 | 0.155 | |||
| Trunk FM (kg) | ||||||||||||
| Baseline trunk FM (kg) | 0.06 | 0.02–0.10 | 0.008 | 0.11* | 0.25 | 0.12–0.38 | <0.0001 | 0.15* | −0.63 | −0.97—0.29 | <0.0001 | 0.10* |
| Change trunk FM (kg) | 0.01 | −0.01–0.03 | 0.332 | 0.05 | −0.01–0.10 | 0.083 | −0.15 | −0.29—0.01 | 0.034 | |||
| Arm FM (kg) | ||||||||||||
| Baseline arm FM (kg) | −0.02 | −0.14–0.09 | 0.662 | 0.08 | 0.10 | −0.25–0.44 | 0.583 | 0.08* | −0.38 | −1.37–0.611 | 0.448 | 0.07* |
| Change arm FM (kg) | 0.05 | −0.03–0.12 | 0.203 | 0.15 | −0.08–0.37 | 0.206 | −0.72 | −1.37—0.07 | 0.031 | |||
| Leg FM (kg) | ||||||||||||
| Baseline leg FM (kg) | −0.04 | −0.08–0.00 | 0.065 | 0.09* | −0.19 | −0.31—0.07 | 0.002 | 0.12* | 0.47 | 0.15–0.78 | 0.004 | 0.07* |
| Change leg FM (kg) | −0.00 | −0.03–0.03 | 0.983 | 0.04 | −0.04–0.12 | 0.367 | −0.10 | −0.30–0.10 | 0.331 | |||
| VAT area (10 cm2) | ||||||||||||
| Baseline VAT (10 cm2) | 0.02 | −0.00–0.03 | 0.093 | 0.08* | 0.11 | 0.06–0.16 | <0.0001 | 0.18* | −0.23 | −0.38—0.09 | 0.001 | 0.11* |
| Change VAT (10 cm2) | 0.01 | −0.01–0.02 | 0.454 | 0.08 | 0.04–0.12 | <0.0001 | −0.20 | −0.31—0.09 | 0.001 | |||
| SAT area (10 cm2) | ||||||||||||
| Baseline SAT (10 cm2) | 0.00 | −0.01–0.01 | 0.385 | 0.08* | 0.01 | −0.02–0.05 | 0.372 | 0.08* | −0.08 | −0.18–0.01 | 0.093 | 0.09* |
| Change SAT (10 cm2) | 0.00 | −0.00–0.01 | 0.356 | 0.02 | −0.01–0.04 | 0.186 | −0.09 | −0.16—0.02 | 0.013 | |||
| VAT are (10 cm2) and Leg FM (kg) | ||||||||||||
| Baseline VAT (10 cm2) | 0.01 | −0.01–0.03 | 0.353 | 0.10* | 0.09 | 0.04–0.15 | 0.002 | 0.18* | −0.19 | −0.36—0.02 | 0.027 | 0.12* |
| Change VAT (10 cm2) | 0.00 | −0.01–0.02 | 0.593 | 0.08 | 0.03–0.12 | 0.002 | −0.16 | −0.22—0.54 | 0.007 | |||
| Baseline leg FM (kg) | −0.03 | −0.07–0.02 | 0.269 | −0.07 | −0.20–0.06 | 0.306 | 0.16 | −0.22–0.54 | 0.409 | |||
| Change leg FM (kg) | −0.00 | −0.03–0.03 | 0.784 | −0.03 | −0.11–0.06 | 0.556 | 0.02 | −0.21–0.24 | 0.863 | |||
Data are presented as β-coefficients, 95% confidence intervals, R2 for each model and p values. Each model includes baseline body fat and fat distribution measures, the change in the body fat and fat distribution measures as predictor variables and HbA1c, fasting insulin resistance (HOMA2-IR) and OGTT-derived insulin sensitivity (Matsuda Index) measures at follow-up as the outcome variables. All the body fat and fat distribution models were adjusted for the potential effects of age at baseline, and regional FM, VAT and SAT models were also adjusted for baseline body fat mass. FM, fat mass. VAT, visceral adipose tissue area; SAT, subcutaneous adipose tissue area. *P value for the model < 0.05