Related Experiment Video
Updated: May 11, 2026
Segmentation and Linear Measurement for Body Composition Analysis using Slice-O-Matic and Horos
Published on: March 21, 2021
Development of a practical screening tool to predict low muscle mass using NHANES 1999-2004
Michael J Goodman1, Sameer R Ghate, Panagiotis Mavros
1Department of Pharmacotherapy, University of Utah, 30 South 2000 East Room 258, Salt Lake City, UT, 84112-5820, USA, michael.goodman@pharm.utah.edu.
Background:
Skeletal muscle mass declines after the age of 50. Loss of skeletal muscle mass is associated with increased morbidity and mortality.
Objective:
This study aims to identify predictors of low skeletal muscle mass in older adults toward development of a practical clinical assessment tool for use by clinicians to identify patients requiring dual-energy X-ray absorptiometry (DXA) screening for muscle mass.
Methods:
Data were drawn from the National Health and Nutrition Examination Surveys (NHANES) from 1999 to 2004. Appendicular skeletal mass (ASM) was calculated based on DXA scans. Skeletal muscle mass index (SMI) was defined as the ratio of ASM divided by height in square centimeters. Elderly participants were classified as having low muscle mass if the SMI was 1 standard deviation (SD) below the mean SMI of young adults (20-40 years old). Logistic regression was conducted separately in males and females age ≥65 years of age to examine the relationship between patients identified as having low muscle mass and health behavior characteristics, adjusting for comorbid conditions. The model was validated on a separate sample of 200 patients.
Results:
Among the NHANES study population, 551 (39.7 %) males and 374 (27.5 %) females had a SMI below the 1 SD cutoff point. NHANES study subjects with a low SMI were older (mean age, 76.2 vs. 72.7 for male; 76.0 vs. 73.7 for female; and both p < 0.0001) and had a lower body mass index (mean BMI, 24.1 vs. 29.4 for male; 22.9 vs. 29.7 for female; p < 0.0001). In adjusted logistic regression analyses, age (for males) and BMI (for both males and females) remained statistically significant. A parsimonious logistic regression model adjusting for age and BMI only had a C statistic of 0.89 for both males and females. The discriminatory power of the parsimonious model increased to 0.93 for males and 0.95 for females when the cutoff defining low SMI was set to 2 SD below the SMI of young adults. In the validation sample, the sensitivity was 81.6 % for males and 90.6 % for females. The specificity was 66.2 % for males and females.
Conclusions:
BMI was strongly associated with a low SMI and may be an informative predictor in the primary care setting. The predictive model worked well in a validation sample.
