Prediction of Device-Derived Body Age Using Bioelectrical Impedance Analysis and Graded Treadmill Performance: An Interpretable Machine Learning Study
Abstract
Background: Body Age, a simplified indicator of health and fitness, is increasingly available as a value derived by commercial bioelectrical impedance analysis devices. The purpose of this study was to train and internally validate machine learning models to predict device-derived Body Age throughout a leakage-aware set of anthropometric, body-composition, metabolic, and graded treadmill-performance variables. Methods: This exploratory study used a secondary analysis of 55 de-identified body-composition baseline records and graded treadmill test records. To reduce target leakage, 20 independent predictor variables were selected after removing proprietary control parameters and composite evaluation outputs. Elastic Net regression, Support Vector Regression (SVR), and Gradient Boosting Regression (GBR) were developed and internally validated using nested cross-validation, with five outer folds for performance estimation and four inner folds for hyperparameter selection. Missing values were imputed with the median within each validation pipeline, and feature standardization was performed when applicable. Model performance was evaluated using mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R2). Predictor importance was determined using standardized Elastic Net coefficients. Results: Device-derived Body Age ranged from 15 to 30 years (19.35 ± 3.00 years). Elastic Net demonstrated the best predictive performance, achieving an MAE of 0.68 ± 0.33 years, RMSE of 1.09 ± 0.80 years, and R2 of 0.811 ± 0.157, outperforming SVR and GBR. Visceral fat index was the strongest predictor, followed by fat percentage, body fat mass, waist-to-hip ratio, body weight, protein mass, body mass index, and skeletal muscle. Exercise-derived variables contributed relatively little to model performance. Conclusion: An interpretable regularized linear model accurately reconstructed proprietary device-derived Body Age and outperformed more complex nonlinear algorithms. Body-composition measurements, particularly visceral adiposity and lean tissue, were the strongest predictors of this proprietary device-derived measure of this consumer-oriented measure, whereas graded treadmill performance provided only limited additional predictive value.
Related
- Agreement of Bioelectrical Impedance Analysis Prediction Equations with Dual-Energy X-Ray Absorptiometry for Tracking Resistance Training-Induced Changes in Fat-Free Mass in Older Women.
- Missing Link between Glycemic Control and Bioelectrical Impedance Analysis
- Bioelectrical impedance-derived body composition and systemic therapy outcomes in non-small cell lung cancer
- Toward Clinically Feasible Assessment of Muscle Mass: Validation of a Seated Bioelectrical Impedance Device Against Dual-Energy X-Ray Absorptiometry (DXA) in Older Adults
- Association of electrical properties of bioelectrical impedance spectroscopy with sedentary time, physical activity and muscle function in older adults
BibTeX
@article{bekir2026prediction,
title = {{Prediction of Device-Derived Body Age Using Bioelectrical Impedance Analysis and Graded Treadmill Performance: An Interpretable Machine Learning Study}},
author = {Wafaa Mohmoud Abdellatif Bekir and Fatma Hassan Abd Elbasset Mourgan and Manaf AlMatar},
journal = {Biomedical and Biotechnology Research Journal},
year = {2026},
month = jul,
doi = {10.4103/bbrj.bbrj_101_26},
url = {https://doi.org/10.4103/bbrj.bbrj_101_26},
note = {CC BY-NC-SA},
}
RIS
TY - JOUR
TI - Prediction of Device-Derived Body Age Using Bioelectrical Impedance Analysis and Graded Treadmill Performance: An Interpretable Machine Learning Study
AU - Wafaa Mohmoud Abdellatif Bekir
AU - Fatma Hassan Abd Elbasset Mourgan
AU - Manaf AlMatar
PY - 2026
DA - 2026/07/01
JO - Biomedical and Biotechnology Research Journal
DO - 10.4103/bbrj.bbrj_101_26
UR - https://doi.org/10.4103/bbrj.bbrj_101_26
AB - Background: Body Age, a simplified indicator of health and fitness, is increasingly available as a value derived by commercial bioelectrical impedance analysis devices. The purpose of this study was to train and internally validate machine learning models to predict device-derived Body Age throughout a leakage-aware set of anthropometric, body-composition, metabolic, and graded treadmill-performance variables. Methods: This exploratory study used a secondary analysis of 55 de-identified body-composition baseline records and graded treadmill test records. To reduce target leakage, 20 independent predictor variables were selected after removing proprietary control parameters and composite evaluation outputs. Elastic Net regression, Support Vector Regression (SVR), and Gradient Boosting Regression (GBR) were developed and internally validated using nested cross-validation, with five outer folds for performance estimation and four inner folds for hyperparameter selection. Missing values were imputed with the median within each validation pipeline, and feature standardization was performed when applicable. Model performance was evaluated using mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R2). Predictor importance was determined using standardized Elastic Net coefficients. Results: Device-derived Body Age ranged from 15 to 30 years (19.35 ± 3.00 years). Elastic Net demonstrated the best predictive performance, achieving an MAE of 0.68 ± 0.33 years, RMSE of 1.09 ± 0.80 years, and R2 of 0.811 ± 0.157, outperforming SVR and GBR. Visceral fat index was the strongest predictor, followed by fat percentage, body fat mass, waist-to-hip ratio, body weight, protein mass, body mass index, and skeletal muscle. Exercise-derived variables contributed relatively little to model performance. Conclusion: An interpretable regularized linear model accurately reconstructed proprietary device-derived Body Age and outperformed more complex nonlinear algorithms. Body-composition measurements, particularly visceral adiposity and lean tissue, were the strongest predictors of this proprietary device-derived measure of this consumer-oriented measure, whereas graded treadmill performance provided only limited additional predictive value.
SN - 2588-9834
SN - 2588-9842
ER -