Marcus Vinicius de Oliveira Cattem Josely Correa Koury ^*

• Nutrition Institute, Rio de Janeiro State University, Rio de Janeiro, Brazil

Skeletal muscle mass, body cell mass, total body water (TBW), and bone mineral mass (BMC) are components of fat-free mass (FFM), which conducts electrical energy due to its high water and electrolyte content. Multifrequency bioelectrical impedance analysis (MF-BIA) has been used to predict FFM, and studies have explored its application for quantifying BMC, a subset of FFM. However, the accuracy of the BMC predicted using MF-BIA depends on the methodological rigor of the frequency selection. This study examined the relationships between BMC and raw MF-BIA data at different frequencies in male adolescent soccer players (n=149; 15.6 ± 0.6 years). The MF-BIA device obtained raw bioelectrical data at 5, 50, and 500 kHz. BMC was quantified using dualenergy X-ray absorptiometry (DXA). Multiple linear regression models and bioelectrical impedance vector analysis (BIVA) were applied to evaluate whole-body and segmental BMC relationships. Whole-body BMC (R²= 0.522), and upper and lower limb BMC (R²=0.349) were best predicted at 5 kHz, while trunk BMC (R²= 0.301) was best predicted at 50 kHz. BIVA revealed a leftward vector shift in participants with higher BMC quartiles. The calculated phase angle (PhA) was significantly higher in the highest BMC quartile for 5, 50, and 500 kHz in both upper and lower limbs (p< 0.05). These findings indicate that MF-BIA could be used as a supplementary tool for studying BMC in adolescent athletes. However, its utility is constrained by prediction and interpretation errors, emphasizing the importance of careful frequency selection.