P435 - COMPARISON OF EQUATIONS FOR MEASURING MUSCLE MASS BY BIOIMPEDANCE ANALYSIS (BIA) IN INTENSIVE CARE UNIT (ICU) PATIENTS: A CROSS-SECTIONAL STUDY

P435

COMPARISON OF EQUATIONS FOR MEASURING MUSCLE MASS BY BIOIMPEDANCE ANALYSIS (BIA) IN INTENSIVE CARE UNIT (ICU) PATIENTS: A CROSS-SECTIONAL STUDY

R. Thibault^1,2,*, E. Gieryga¹, T. Saillenfest¹, Q. Quelven³, Y. Launey⁴, N. Nesseler⁵, J.-M. Tadié³

¹Endocrinology-Diabetology-Nutrition, CHU Rennes, ²Nutrition Metabolims Cancer, NuMeCan, Univ Rennes, INRAE, INSERM, ³Medical ICU, ⁴Surgical ICU, ⁵Cardiovascular ICU, CHU Rennes, Rennes, France

Rationale: No method to assess muscle mass is validated in the ICU. Objectives: to compare body composition by ICU department and gender (primary), assess malnutrition prevalence and correlation/concordance between muscle mass (MM) equations (secondary).

Methods: Prospective observational study. Inclusion criteria: consecutive, expected length of stay >48 hr, abdominal CT scan <72 hr. BIA equations: MM (Janssen (1) if <70 yr) and appendicular muscle mass (AMM) (Kyle (2), Sergi (3) if ≥70 yr). GLIM malnutrition diagnosis: etiology=acute illness; ≥1 phenotypic criterion: weight loss >5% <6 mo or >10% >6 mo, BMI <20 (<70 yr) or 22 (≥70 yr), FFMI<15 (women (F)), <17 (men (M)). Spearman correlation of Kyle vs Sergi AMM equations if ≥70 yr, and Janssen and Kyle MM equations if <70 yr. Bland-Altman agreement ​​between Kyle and Sergi AMM equations if ≥70 yr.

Results: N=75 patients (surgical (S), n=39, medical (M), n=31, cardiovascular (CV) ICUs, n=5); 60.7±15.2 yr, 64% M, shock 28%. Low FFMI, n=10 (13%), BMI, n=10 (13%), weight loss, n=8 (10.4%). Malnutrition: n=17 (22.7%), severe in 6.7%. FFMI (F, 18.7±2.9 vs. 18.5±4.3; M, 21.4±4.4 vs. 21.0±4.7, NS) and phase angle (4.1±1.7 vs. 4.5±2.3°, P=0.46) were not different in S vs MICUs Body fat index higher in SICU (F, 10.3±5.0 vs. 7.6±5.4; M, 5.8±2.5 vs. 4.4±1.6, P=0.02). Strong correlation between the Janssen and Kyle equations (n=48, rho=0.97, P<0.001), and between Sergi and Kyle’s (n=27, rho=1, P<0.001). Good agreement of AMM by Sergi and Kyle equations (n=27) (outliers <4%).

Conclusion: In a selected population of ICU patients, ¼ is malnourished. There is a strong correlation and good agreement between different methods for estimating muscle mass. Their reliability needs confirmation.

References: 1. Janssen I et al. J Appl Physiol 2000;89:465-71; 2. Kyle UG et al. Clin Nutr 2003;22:537-43; 3. Sergi G et al. Clin Nutr. 2015;34:667-73

Disclosure of Interest: None declared