Exploratory approach to speculate on body composition models for elite teenage basketball players

The study of human body composition covers many fields, from clinical assessment to sports performance. Humans manifest different traits that set limitations in model estimation. This investigation aimed to (i) provide sport-specific densities for reducing biological variation in body composition assessment, and (ii) develop new regression models for body volume and density estimation by field applications. Thirty elite basketball players (16.84 ± 1.15 years; 188.48 ± 6.74 cm; 82.66 ± 12.23 kg) with a mean of 10.48 years of sport experience underwent anthropometric, bio-impedance (BIA) and air-displacement plethysmography (ADP) evaluations. Many models accounting for body fat percentage (BF) used the measured body mass (BM), volume (BV) and density (d) to analyse adipose and fat-free tissues. A new model payable for all players encompassed BF, with low error propagation (3.4% of BM). In addition, two new methods estimated BV by anthropometry (R² = 0.96, RMSE = 2.9%) or BIA (R² = 0.81, RMSE = 4.56%), with a high degree of precision (97.8 and 86.8%) and accuracy (100 and 99.6%). Body composition requires rigorous speculation, advanced instruments and high costs that could lead investigators to omit relevant concepts and produce estimation biases. Specific and easier methods may enhance its applicability.