Identification of ultrasound predictors and development of a diagnostic scoring system for sarcopenia using LASSO regression: a cross-sectional study

Ultrasound-based muscle assessment has emerged as a practical and noninvasive tool for evaluating sarcopenia; however, the relative diagnostic value of individual ultrasound-derived muscle parameters and their integration into a simple scoring system remain unclear. This cross-sectional study included 145 community-dwelling women aged 65-80 years. Eight ultrasound-derived quadriceps muscle parameters were evaluated: rectus femoris thickness, cross-sectional area, and echo intensity; vastus intermedius thickness; vastus lateralis thickness, fascicle length, and pennation angle; and quadriceps femoris thickness, defined as the combined thickness of the rectus femoris and vastus intermedius. Sarcopenia was diagnosed in 8 participants (5.5%). Most ultrasound parameters demonstrated significant weak-to-moderate correlations with handgrip strength and skeletal muscle mass index. Least absolute shrinkage and selection operator regression identified quadriceps femoris thickness, vastus lateralis thickness, and rectus femoris cross-sectional area as the most informative predictors. Based on these variables, a simple preliminary ultrasound-based scoring system was developed using parameter-specific cutoff values. A cutoff score of ≥ 3 demonstrated high diagnostic performance, with an area under the receiver operating characteristic curve of 0.914, sensitivity of 0.875, and specificity of 0.861. This exploratory ultrasound-based scoring system may suggest a feasible approach for sarcopenia screening in older women, although further validation in larger populations is warranted.