Novel MRI-Guided Ultrasound Equations for Whole-Body Muscle Mass in Caucasian Adults

Background: Muscle mass is a critical indicator of health and functionality, yet its accurate measurement remains challenging. Ultrasound offers a promising alternative, providing cost-effective, non-invasive assessments of surrogates of muscle mass: muscle thickness and cross-sectional area (CSA). This study aims to develop and validate ultrasound-derived equations for estimating whole-body muscle mass in healthy Caucasian adults using magnetic resonance imaging (MRI) as the reference standard, incorporating CSA measurements to enhance the accuracy and applicability of these equations.

Methods: We enrolled 211 healthy Caucasian adults (age: 42.0 years [29.0-58.0], 52% female) for whole-body MRI and ultrasound examinations, assessing eight muscle thicknesses and seven CSAs across the right arm, trunk and leg. The sample was divided into a development group (two-thirds) and a cross-validation group (one-third). Stepwise multiple regression established ultrasound equations in the development group and cross-validation group. After successful cross-validation, the full sample was used to create the most accurate and most practical equations.

Results: The most accurate equation for estimating whole-body muscle mass included ultrasound muscle thickness measurements of the forearm extensor, rectus abdominis, rectus femoris, biceps femoris and tibialis anterior muscles, CSA measurements of the triceps brachii and tibialis anterior muscles and sex, weight and BMI. This model achieved an adjusted R2 of 0.942 and a standard error of estimate (SEE) of 1.7 kg. A more practical equation, requiring fewer measurements, focused on ultrasound muscle thickness and CSA of select arm, abdominal and leg muscles, combined with sex and height. This simplified model showed an adjusted R2 of 0.927 and an SEE of 2.0 kg, offering a good balance between accuracy and measurement burden.

Conclusions: The equations developed in this study enable accurate estimation of whole-body muscle mass in a Caucasian population using ultrasound. A practical, time-efficient equation with fewer variables is available alongside a more detailed equation that provides higher accuracy. Importantly, models that combine muscle thickness and CSA measurements demonstrated improved prediction accuracy (higher adjusted R2 and lower SEE) compared to those using muscle thickness alone, supporting the inclusion of CSA in future applications.