Agreement between bioelectrical impedance analysis and equation-derived estimates of appendicular skeletal muscle in acutely hospitalized multiethnic older patients.
Liu Y, Tan L, Merchant R • Aging clinical and experimental research • 2026
Furushima's equation aligns better with BIA-measured ASM than COCONUT and may complement handgrip strength in sarcopenia screening, though subgroup-specific biases require caution.
Key Findings
Results
Furushima's equation showed stronger overall concordance with BIA-measured ASM than the COCONUT equation.
Lin's concordance correlation coefficient (CCC) was 0.696 for Furushima versus 0.538 for COCONUT.
Furushima's equation had a smaller underestimate bias (1.56 kg) compared to COCONUT (3.03 kg).
The study included 295 hospitalized, multiethnic older adults in Singapore.
ASM was measured using multi-frequency BIA S10 as the reference method.
Results
Furushima's equation achieved moderate diagnostic agreement for low appendicular skeletal muscle mass index (ASMI), while COCONUT showed poor agreement.
Furushima achieved κ = 0.48 and AUC = 0.78 for detecting low ASMI.
COCONUT showed poor diagnostic agreement with κ = 0.17 and AUC = 0.62.
Low ASMI was defined using AWGS 2019 cut-offs.
Agreement was assessed using Cohen's κ and area under the receiver operating characteristic curve (AUC).
Results
COCONUT performed particularly poorly in women, while Furushima improved performance in women but remained low overall.
Among women, COCONUT had a CCC of 0.192 and a bias of 4.40 kg.
Furushima's equation improved performance among women but CCC remained low overall (CCC = 0.371).
Analyses were stratified by sex and obesity status.
Results
Among obese patients, Furushima's equation demonstrated the highest concordance with BIA-measured ASM.
In obese patients, Furushima achieved a CCC of 0.862 and a bias of only 0.42 kg.
This represented the strongest subgroup performance observed for either equation.
Analyses were stratified by obesity status alongside sex.
Methods
The study used a cross-sectional design in a multiethnic hospitalized population to evaluate equation-derived ASM estimates against BIA.
The sample comprised 295 hospitalized older adults in Singapore.
Agreement was assessed by Lin's concordance correlation coefficient (CCC) and Bland-Altman analysis.
The COCONUT equation uses calf circumference, age, and sex (NHANES-based), while Furushima's equation uses age, anthropometry, and handgrip strength (HGS).
The study population was described as multiethnic, conducted in Singapore.
Liu Y, Tan L, Merchant R. (2026). Agreement between bioelectrical impedance analysis and equation-derived estimates of appendicular skeletal muscle in acutely hospitalized multiethnic older patients.. Aging clinical and experimental research. https://doi.org/10.1007/s40520-025-03283-2