Monitoring changes in total fat mass and abdominal adiposity are important in understanding the impact of different types of weight loss interventions on health risks. Our objective was to assess the usefulness of anthropometry and bioelectrical impedance analysis (BIA) in predicting fat mass changes during moderate weight loss.

Fat mass changes were assessed in 34 overweight adults (24 females, 10 males) after a 12-week supervised weight loss induced by caloric restriction (−30% of requirement) using BIA and DXA. Agreement between BIA and DXA measurements was assessed by Bland–Altman plots. Linear regression modeling was used to predict body and truncal fat mass from anthropometric measures.

Diet intervention resulted in a significant decrease in body weight (−7.86±2.87kg), body mass index (BMI −2.69±0.98kg/m²), total body fat (−5.22±2.32kg), truncal fat (−2.80±1.94kg) and waist circumference (−5.52±3.57cm). BMI and body weight were highly correlated with body fat (0.83 and 0.92 in females and 0.94 and 0.92 in males, respectively) and truncal fat (0.75 and 0.87 in females; 0.90 and 0.84 in males, respectively) during weight loss. Waist circumference was more correlated with truncal fat in males than females (0.94 vs. 0.85 in females). Compared to DXA, BIA underestimated total body fat changes in males (−8.8kg; p<0.001) and overestimated total body fat changes in females (+2.1kg; p<0.001).

Body mass index, body weight, and waist circumference provide simple and more accurate than BIA estimates of relative changes in total and truncal fat during moderate weight loss in adults.