Premium
Influence of Methods Used to Establish Sarcopenia Cutoff Values for Skeletal Muscle Measures Using Unenhanced and Contrast‐Enhanced Computed Tomography Images
Author(s) -
Feng Zhichao,
Rong Pengfei,
Luo Muqing,
Sun Xianxi,
Wang Wei
Publication year - 2019
Publication title -
journal of parenteral and enteral nutrition
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.935
H-Index - 98
eISSN - 1941-2444
pISSN - 0148-6071
DOI - 10.1002/jpen.1519
Subject(s) - sarcopenia , medicine , cutoff , skeletal muscle , mcnemar's test , percentile , nuclear medicine , concordance , lumbar , hounsfield scale , radiology , computed tomography , mathematics , physics , quantum mechanics , statistics
Background Multiple cutoff values of computed tomography (CT)‐based skeletal muscle measures have been proposed, but there is currently no consensus used to identify sarcopenia. We aimed to evaluate the influence of statistical methods used to establish sarcopenia cutoff values and to examine the impact of contrast enhancement on the skeletal muscle measures. Methods The skeletal muscle area (SMA) and muscle radiation attenuation (MRA) of 316 healthy individuals were measured on unenhanced CT images at the third lumbar vertebra level, and the skeletal muscle index (SMI) was SMA divided by height squared. Possible cutoff values were established using 2 methods: 5 th percentile of individuals aged 20‐60 years or mean − 2 × SD of individuals aged 20‐50 years. The concordance was assessed using Cohen's κ coefficients and McNemar test. The skeletal muscle parameters on 3 phases from 30 CT examinations were compared. Results The concordance between the 2 methods was almost perfect (κ coefficients: 0.830‐0.849) for low MRA but slight to moderate (κ coefficients: 0.189‐0.591) for low SMI, especially in the men ( P < 0.01). Compared with the unenhanced images, the mean SMA, SMI, and MRA on the contrast‐enhanced images increased by 0.8%‐1.7%, 0.8%‐1.8%, and 14.8%‐21.6% (all P < 0.001), respectively, and only the changes in MRA were clinically significant. Conclusions The methods for establishing cutoff values and contrast enhancement influence the identification of low SMI and low MRA, respectively. Thus, the definition of sarcopenia should include the standardized method for establishing cutoff values and the phase of CT for analysis.