Estimation of body fat in adults using a portable A-mode ultrasound.

OBJECTIVE This study aimed to develop and validate equations to estimate body fat based on anthropometric measurements of subcutaneous fat thickness (SFT) and muscle thickness (MT) measured by A-mode ultrasound (BodyMetrix) in Brazilian adults. METHODS Individuals (n = 206) underwent air-displacement plethysmography for body composition assessment. Arm, thigh, and calf circumferences were also obtained. SFT from triceps, biceps, subscapular, abdominal, thigh, and calf regions and MT from triceps, biceps, thigh, and calf regions were measured by BodyMetrix. Prediction equations were developed by stepwise multiple linear regression using the circumferences, weight, height, SFT, and MT. Lin's concordance correlation coefficient, mean difference, and 95% limits of agreement (95% LOA) were assessed in apparent and internal validity. RESULTS The prediction equation for whole-body fat for men included thigh circumference, triceps and thigh SFT, biceps MT, weight, and height. The equation for women included age, calf circumference, abdominal and calf SFT, weight, and height. The prediction equation overestimated men's whole-body fat by 0.5 percentual points, in average, and the lower and upper 95% LOA were -6.8% and 7.7%, respectively. For women, the prediction equation overestimated whole-body fat by 0.1 percentual points, in average. Lower and upper 95% LOA were -6.5% and 6.7%, respectively. Optimism-adjusted results using 500 repetitions with same size samples have shown similar results. Body fat extremes did not influence the whole-body fat estimation. CONCLUSIONS BodyMetrix A-mode ultrasound, in association with selected conventional anthropometric measurements, proved to be a reliable tool for the estimation of body fat percentage.