Annals of Clinical Epidemiology
Online ISSN : 2434-4338
Development and validation of population-specific prediction equations for lean body mass in older Japanese adults
Chikako Iwai, Akira Okada, Bo-Kyung Son, Yuta Suzuki, Tatsuhiko Azegami, Kazuki Aoyama, Hidehiro Kaneko, Weida Lyu, Tomoki Tanaka, Hideo Yasunaga, Katsuya Iijima
Author information
JOURNAL OPEN ACCESS Advance online publication

Article ID: 27016

Details
Abstract

BACKGROUND

Lean body mass (LBM) is an important component of sarcopenia, yet commonly used prediction equations developed in middle-aged American populations may not accurately estimate LBM in older Japanese adults. This study aimed to develop LBM prediction equations specific to older Japanese adults by updating existing models.

METHODS

We analyzed data from the Kashiwa Study, a prospective cohort of community-dwelling older Japanese adults initiated in 2012, wherein LBM was measured using bioelectrical impedance analysis. We evaluated the previously published US equations and updated them using intercept recalibration and slope adjustment. Calibration, Bland–Altman agreement, and discrimination for low handgrip strength and gait speed <1 m/s were assessed in the validation cohort. Repeated measurements within individuals were accounted for using participant-level bootstrap confidence intervals clustered at the individual level.

RESULTS

Among 2,019 participants (6,492 measurements), 1,989 baseline measurements comprised the development cohort, and 4,503 follow-up measurements formed the validation cohort. The original equations systematically underestimated LBM in both sexes, whereas the intercept-recalibrated and slope-adjusted equations achieved near-optimal calibration. The slope-adjusted equations demonstrated slightly higher c-statistics for discriminating low handgrip strength than the original equations (men: 0.769 vs. 0.757; women: 0.759 vs. 0.750). Similar improvements were observed for gait speed <1 m/s.

CONCLUSIONS

Widely used LBM prediction equations exhibit systematic bias when applied to older Japanese adults. Population-specific model updating substantially improved calibration, although external validation is warranted. Population- and age-appropriate LBM prediction may strengthen epidemiological research on frailty, disability, and adverse health outcomes in aging societies.

Content from these authors
© 2026 Society for Clinical Epidemiology

This article is licensed under a Creative Commons [Attribution-NonCommercial-NoDerivatives 4.0 International] license.
https://creativecommons.org/licenses/by-nc-nd/4.0/
Previous article Next article
feedback
Top