P274 Bioelectrical impedance analysis identifies low skeletal muscle index in patients with Inflammatory Bowel Disease
Nguyen, A.(1);Holt, D.(2);Burns, M.(2);Herath, M.(3);Gibson, P.R.(4);Ebeling, P.(5);Milat, F.(3);Moore, G.(6);
(1)Monash University- Monash Health and Alfred Health, Gastroenterology, Melbourne, Australia;(2)Monash Health, Gastroenterology, Melbourne, Australia;(3)Monash University- Monash Health and Hudson Institute of Medical Research, Endocrinology, Melbourne, Australia;(4)Monash University and Alfred Health, Gastroenterology, Melbourne, Australia;(5)Monash University and Monash Health, Endocrinology, Melbourne, Australia;(6)Monash University and Monash Health, Gastroenterology, Melbourne, Australia;
Background
Low skeletal muscle index (SMI) has been associated with increased need for surgery and post-surgical complications in patients with Inflammatory Bowel Disease (IBD). On dual energy X-ray absorptiometry (DXA), SMI is routinely calculated by dividing limb skeletal muscle mass (kg) by the square of the height (m2). Bioelectrical impedance analysis (BIA) provides a quick, inexpensive and radiation-free method to measure body composition. The aim of this study was to compare BIA with DXA to identify patients with low skeletal muscle mass and to show change over 14 weeks.
Methods
This single-centre prospective study recruited adult patients with IBD and adult controls, matched for gender. Active disease was defined as a faecal calprotectin of ≥150 μg/g. Active IBD patients undergoing biologic induction had repeat body composition measurements at least 14 weeks apart. All patients underwent single-frequency segmental BIA (Tanita BC-418) and DXA (GE Lunar Prodigy). A BIA-predicted SMI was calculated by dividing limb fat-free mass (kg) by the square of the height (m2).
Results
Of 48 patients with IBD, 54% had Crohn’s disease, 56% were male and median age was 29 (IQR 23-40) years. 28 had active disease, 16 inactive disease and 4 had no faecal calprotectin available. Patients with IBD and controls (n=30) had similar total body water, measured by BIA, and differed in albumin levels (39 vs 40.1 g/L, p=0.011). In IBD patients and controls, BIA total fat-free mass correlated strongly with DXA-derived fat-free mass (0.96 and 0.93, both p<0.0001). The BIA-predicted SMI also correlated well with DXA-derived SMI (0.89 and 0.91, both p<0.0001).
The correlation was stronger for Ulcerative Colitis (0.96) compared to Crohn’s disease (0.84). In patients with IBD, the mean difference between BIA-predicted SMI and DXA-derived SMI was 1.1 ± 0.6 kg/m2 with 95% limits of agreement, -0.1 to 2.3 kg/m2 . Controls followed a similar trend with a mean difference of 0.5 ± 0.6 kg/m2 and 95% limits of agreement, -0.5 to 1.6 kg/m2. 
The BIA-predicted SMI demonstrated a strong ability to identify low DXA-derived SMI for women with IBD (AUC: 0.98) and less so for men (AUC: 0.79). In 14 patients undergoing biologic treatment, change in DXA-derived SMI correlated with BIA-derived SMI (0.76, p=0.003).
Conclusion
An accurate estimate of SMI can be obtained with BIA as point-of-care in clinical practice. Further prospective studies with larger sample sizes are needed to determine the accuracy of BIA in monitoring changes of SMI during IBD treatment.