Relative muscle indices and healthy reference values for sarcopenia assessment using T10 through L5 computed tomography skeletal muscle area

Brian A. Derstine,Sven A. Holcombe,Nicholas C. Wang,Brian E. Ross,June A. Sullivan,Stewart C. Wang,Grace L. Su

DOI: https://doi.org/10.1038/s41598-024-71613-x

IF: 4.6

2024-09-25

Scientific Reports

Abstract:Sarcopenia is the age-related loss of skeletal muscle mass and function. Computed tomography (CT) assessments of sarcopenia utilize measurements of skeletal muscle cross-sectional area (SMA), radiation attenuation (SMRA), and intramuscular adipose tissue (IMAT). Unadjusted SMA is strongly correlated with both height and body mass index (BMI); therefore, SMA must be adjusted for body size to assess sarcopenic low muscle mass fairly in individuals of different heights and BMI. SMA / height (rather than ) provides optimal height adjustment, and vertebra-specific relative muscle index (RMI) equations optimally adjust for both height and BMI. Since L3 measurement is not available in all CT scans, sarcopenic low muscle mass may be assessed using other levels. Both a mid-vertebral slice and an inferior slice have been used to define 'L3 SMA', but the effect of vertebral slice location on SMA measurements is unexplored. Healthy reference values for skeletal muscle measures at mid- and inferior vertebra slices between T10 and L5, have not yet been reported. We extracted T10 through L5 SMA, SMRA, and IMAT at a mid-vertebral and inferior slice using non-contrast-enhanced CT scans from healthy, adult kidney donor candidates between age 18 and 73. We compared paired differences in SMA between the mid-vertebral slice versus the inferior slice. We calculated the skeletal muscle gauge as . We used allometric analysis to find the optimal height scaling power for SMA. To enable comparisons with other published reference cohorts, we computed two height-adjusted measures; (optimal) and (traditional). Using the young, healthy reference cohort, we utilized multiple linear regression to calculate relative muscle index z-scores ( , ), which adjust for both height and BMI, at each vertebra level. We assessed Pearson correlations of each muscle area measure versus age, height, weight, and BMI separately by sex and vertebra number. We assessed the differences in means between age 18–40 versus 20–40 as the healthy, young adult reference group. We reported means, standard deviations, and sarcopenia cutpoints (mean-2SD and 5th percentile) by sex and age group for all measures. Sex-specific allometric analysis showed that height to the power of one was the optimal adjustment for SMA in both men and women at all vertebra levels. Differences between mid-vertebra and inferior slice SMA were statistically significant at each vertebra level, except for T10 in men. was uncorrelated with height, whereas was negatively correlated with height at all vertebra levels. Both and were positively correlated with BMI at all vertebra levels. was uncorrelated with BMI, weight, and height (minimal positive correlation in women at , , and ) whereas was uncorrelated with BMI, but negatively correlated with height and weight at all levels. There were no significant differences in SMA between 18–40 versus 20–40 age groups. Healthy reference values and sarcopenic cutpoints are reported stratified by sex, vertebra level, and age group for each measure. Height to the power of one ( SMA / height ) is the optimal height adjustment factor for SMA at all levels between through . The use of should be discontinued as it retains a significant negative correlation with height and is therefore biased towards identifying sarcopenia in taller individuals. Measurement of SMA at a mid-vertebral slice is significantly different from measurement of SMA at an inferior aspect slice. Reference values should be used for the appropriate slice. We report sarcopenic healthy reference values for skeletal muscle measures at the mid-vertebral and inferior aspect slice for T10 through L5 vertebra levels. Relative muscle index (RMI) equations developed here minimize correlation with both height and BMI, producing unbiased assessments of relative muscle mass across the full range of body sizes. We recommend the use of these RMI equations in other cohorts.

multidisciplinary sciences