Annals of Geriatric Medicine and Research · Nov 2, 2026

Height, Weight, and BMI Muscle Mass Adjustment for Sarcopenic Obesity: Differential Associations with Functional Outcomes among Healthy Asian Older Adults

By Koh JEJ, Lim JP, Chia JQ et al.

BACKGROUND: The optimal method of adjusting appendicular lean mass (ALM) for sarcopenic obesity (SO) remains uncertain. Height²-adjusted ALM reflects absolute muscle deficit, whereas weight- and body mass index (BMI)-adjusted approaches capture relative muscle insufficiency in the context of adiposity. We examined whether associations between SO and functional outcomes differ according to ALM adjustment method. METHODS: We conducted a cross-sectional analysis of 300 community-dwelling older adults from the GeriLABS-2 cohort in Singapore (mean age 67.4±7.1 years; 68.7% female). Body composition was assessed using dual-energy X-ray absorptiometry. SO was defined using ALM adjusted by height², weight, and BMI (Foundation for the National Institutes of Health [FNIH] and Asian Working Group for Sarcopenia 2025 cutoffs). Exploratory multivariable logistic regression evaluated associations with Short Physical Performance Battery (SPPB), gait speed, falls risk, social activity, and self-rated health. RESULTS: SO prevalence ranged from 5.3% to 7.3% across definitions. Overall four-group agreement was very good (κ=0.87-0.95), although SO-specific agreement was lower (κ=0.61-0.90). SO was associated with low SPPB across all definitions. Slower gait speed was associated with SO for height² and BMI-FNIH definitions. Increased falls risk was observed only with BMI-based definitions (FNIH: OR=3.54, 95% CI:1.05-11.90; AWGS: OR=3.34, 95% CI: 1.02-10.97). CONCLUSION: ALM adjustment methods identify overlapping but not identical SO phenotypes and show different patterns of association with functional outcomes. These exploratory findings suggest that ALM adjustment may influence the clinical phenotype identified as SO, but do not establish the superiority of any single method and require confirmation in larger longitudinal studies.

doi.org/10.4235/agmr.26.0100Open accessPubMed 42968444Biomarkers