Psoas Muscle Cross-sectional Area as a Measure of Whole-body Lean Muscle Mass in Maintenance Hemodialysis Patients

Abstract

Objective We investigate whether psoas or paraspinous muscle area measured on a single L4–5 image is a useful measure of whole lean body mass compared to dedicated mid-thigh magnetic resonance imaging (MRI). Design Observational study. Setting Outpatient dialysis units and a research clinic. Subjects 105 adult participants on maintenance hemodialysis. No control group was used. Exposure variables Psoas muscle area, paraspinous muscle area, and mid-thigh muscle area (MTMA) were measured by MRI. Main outcome measure Lean body mass was measured by dual-energy absorptiometry (DEXA) scan. Results In separate multivariable linear regression models, psoas, paraspinous, and mid-thigh muscle area were associated with increase in lean body mass. In separate multivariate logistic regression models, c-statistics for diagnosis of sarcopenia (defined as < 25th percentile of lean body mass) were 0.69 for paraspinous muscle area, 0.81 for psoas muscle area, and 0.89 for mid-thigh muscle area. With sarcopenia defined as < 10th percentile of lean body mass, the corresponding c-statistics were 0.71, 0.92, and 0.94. Conclusions We conclude that psoas muscle area provides a good measure of whole body muscle mass, better than paraspinous muscle area but slightly inferior to mid thigh measurement. Hence, in body composition studies a single axial MR image at the L4–L5 level can be used to provide information on both fat and muscle and may eliminate the need for time-consuming measurement of muscle area in the thigh.

MRI‐Derived Biomarkers Related to Sarcopenia: A Systematic Review

“…Indeed, the study performed by Yang et al showed that a single assessment of mid‐thigh CSA performed at 50% of femur length by MRI correlates well with subcutaneous fat ( r = 0.98, P < 0.001), intermuscular fat ( r = 0.97, P < 0.001), and thigh muscle volume ( r = 0.97, P < 0.001) measured by MRI, allowing speeding up volume quantity MRI assessment. These results were confirmed by the study performed by Morrell et al, who showed that lean body weight values estimated by dual‐energy absorptiometry significantly correlates with CSA of psoas muscle ( r = 0.74, P < 0.001), paraspinous muscle ( r = 0.58, P < 0.001), and mid‐thigh muscles ( r = 0.81, P < 0.001) measured on a single MRI slice.…”

“…Of 67 records identified through search query and two identified through other sources, 45 were excluded from title and abstract, while the remaining 24 were downloaded for a deeper assessment. Finally, a total of 18 articles matched the topic at issue …”

“…Body regions analyzed were the trunk in 6/18 articles (33%), thigh in 7/18 (39%), calf in 2/18 (11%), back muscles in 1/18 (6%), and mixed combinations of body regions in 2/18 (11%) …”

“…In addition to previous studies, Morrell et al propose psoas CSA as a valuable substitute of mid‐thigh measurement, to be performed during routine abdominal MRI thanks to their high correlation ( r = 0.83, P = 0.002). Moreover, psoas CSA measured at the vertebral level L4–L5 has also been suggested as an accurate predictor of sarcopenia.…”

“…Moreover, psoas CSA measured at the vertebral level L4–L5 has also been suggested as an accurate predictor of sarcopenia. The area under the receiver operating characteristic curve for sarcopenia diagnosis was 0.81 (with sarcopenia defined as <25th percentile of LBM) and 0.92 (with sarcopenia defined as <10th percentile of LBM) …”