Muscle mass evaluation in patients with intestinal failure or insufficiency: Comparison of methods

Reduced muscle mass (MM) is commonly seen in patients suffering from malnutrition and sarcopenia. Both predispose for increased mortality as well as increased length- and cost of hospital-stays. Calf circumference (CC) has been demonstrated as a simple tool for MM estimates. Measuring CC is more available compared to the methods bioelectrical impedance analysis (BIA) and dual-energy x-ray absorptiometry. Like arm muscle circumference (AMC), the accuracy of CC has been evaluated in a few studies, but its specificity and sensitivity need further investigation. Therefore, the aim of this study is to investigate if CC, CC corrected with body mass index (BMI) (CC-corr), AMC and fat free mass index (FFMI) can be used as a proxy measurement for MM compared to skeletal muscle (SM) in a cohort of 300 patients suffering from intestinal failure (IF) or intestinal insufficiency (INS). Further, we aimed to investigate the prevalence of reduced MM with the different methods and between the groups IF and INS.

Anthropometric measurements were collected from patients affiliated to Center for Nutrition and Intestinal Failure with IF or INS between 2021 and 2024 by a medical laboratory scientist (co-author). Descriptive statistics including Pearson correlation coefficients were performed between SM and CC, CC-corr, AMC and FFMI. Furthermore, agreement analysis using Cohens Kappa as well as sensitivity and specificity were calculated to compare the diagnostic accuracy between SM and CC, CC-corr, AMC and FFMI.

Reduced MM in IF and INS is common in both groups. Based on AMC (7.0 %), FFMI (42.7 %), CC-corr (57.3 %), CC (52.7 %) and SM (58.0 %) low MM were seen respectively. The best agreement was between SM index (SMI) and FFMI (κ = 0.700), where the agreement between SM and AMC, SM and CC-corr and SM and CC were below the minimal satisfactory agreement of 0.6 (κ = 0.093, κ = 0.276, κ = 0.570). Pearson correlation coefficient was nearest 1 between SMI and FFMI (r > 0.800, p < 0.001). There were significant linear associations between SMI and FFMI as well as SM and CC-corr, CC and AMC both unadjusted and adjusted for gender, age, BMI and the two patient groups (p < 0.05), however only CC-corr differed between the groups.

The findings of this study show limited agreement between CC-corr and AMC on the study population in question, however the agreement was better when using CC without adjusting for BMI. Thus, it is dubious if CC-corr can serve as a proxy measure for MM in IF and INS patients. This study advises for more research on a reference material conducted on a more comparable population (i.e., a Danish standard population adjusted for age, gender, and BMI). Independent of methods, both groups showed a high prevalence of reduced MM.