# Establishing percentiles for phase angle, skeletal muscle mass index, and muscle-to-fat ratio in Brazilian adolescents: a bioelectrical impedance analysis study with 59,000 participants

## TL;DR

This study establishes national reference percentiles for body composition metrics in Brazilian adolescents using bioelectrical impedance data from over 59,000 participants.

## Contribution

The study provides the first large-scale, nationally representative reference curves for phase angle, skeletal muscle mass index, and muscle-to-fat ratio in Brazilian adolescents.

## Key findings

- Reference percentiles for phase angle, skeletal muscle mass index, and muscle-to-fat ratio were generated for Brazilian adolescents.
- Percentile spreads for muscle-related indices increased with age, reflecting greater interindividual variability.
- Device-dependent body fat percentage percentiles showed expected age- and sex-specific patterns but require validation.

## Abstract

Reliable normative references for adolescent body composition are essential for screening and surveillance, yet Brazilian data at national scale remain scarce and heterogeneous across methods and devices.

To generate age- and sex-specific reference curves for phase angle (PhA), skeletal muscle mass index (SMMI), and muscle-to-fat ratio (MFR) in Brazilian adolescents, and to present device- dependent body fat percentage (%BF) percentiles as secondary, interpretation- cautioned outputs.

We analyzed a frozen, de-identified extract of routine multi- frequency bioimpedance assessments (10.00–19.99 years) spanning all five Brazilian macro-regions and multiple recruitment contexts (schools, clinics/primary care, private practices, gyms/fitness centers, community programs). The extract is produced by a vendor privacy/QA pipeline compliant with LGPD; upstream cleaning logs are not accessible to the authors. Prespecified plausibility and deduplication rules (anthropometry and impedance thresholds; robust BIVA on R/H and Xc/H) were re- applied and yielded no additional exclusions. Sex-specific age curves were modeled with GAMLSS (LMS) to estimate smoothed percentiles (P3, P10, P25, P50, P75, P90, P97). %BF, derived from a proprietary embedded equation, is reported as secondary.

The analytic cohort comprised 59,000 adolescents from all macro-regions and diverse contexts. Median PhA, SMMI, and MFR increased with age in boys and showed the expected attenuated or plateauing patterns in girls across mid- to late adolescence. Percentile spreads widened with age for muscle-related indices, indicating growing interindividual variability. Device-dependent %BF percentiles exhibited age- and sex- specific trajectories consistent with physiological expectations but should be interpreted with caution pending external validation.

We provide national reference percentiles for PhA, SMMI, and MFR that are immediately useful for clinical and public- health applications and less sensitive to device-specific assumptions. %BF curves are offered as secondary and require independent validation against a criterion method before routine use. Future work should confirm regional representativeness and cross-device portability.