Measuring muscle and subcutaneous adipose tissue: agreement and reliability between dietetic-led ultrasound and magnetic resonance imaging

Fuente: PubMed "essential OR oil extract"
Nutrition. 2025 Dec 14;144:113068. doi: 10.1016/j.nut.2025.113068. Online ahead of print.ABSTRACTBACKGROUND: Low muscle mass is prevalent and has detrimental effects on morbidity and mortality. Body composition assessment is essential for early detection of muscle loss and malnutrition, yet reference methods such as magnetic resonance imaging (MRI) are costly and routinely inaccessible in clinical practice. Ultrasound has emerged as a promising noninvasive, bedside tool for body composition assessment.AIM: To assess the agreement between ultrasound and MRI-measured muscle thickness and cross-sectional area (CSA) and subcutaneous adipose thickness at the bicep and quadriceps. To investigate inter- and intra-operator reliability of ultrasound measurements.DESIGN: A cross-sectional study was conducted with 35 healthy participants. Ultrasound images were taken by two researchers. An MRI scan was conducted during the same research visit. Intraclass correlation coefficient (ICC) and Bland-Altman analyses were performed to test agreement of ultrasound against MRI, as well as inter- and intra-operator reliability.RESULTS: The mean difference between measures was 0.04 mm (SD = 2.68 mm) at the right quadricep 2/3 point, showing minimal bias. The mean differences for quadricep muscle thickness ranged from -1.41 to 0.04 mm and CSA -0.07 to 0.05 cm2. High ICC of >0.90 were obtained for muscle and subcutaneous adipose thickness at all 5 sites when comparing ultrasound with MRI, and for the inter- and intra-operator reliability of ultrasound.CONCLUSIONS: In healthy participants, ultrasound measures of muscle and subcutaneous adipose thickness had a high level of agreement compared with MRI measures. This highlights the potential to use ultrasound for body composition assessment in clinical practice and research.PMID:41505810 | DOI:10.1016/j.nut.2025.113068