LB038 - COMPARING MULTI-FREQUENCY BIOELECTRICAL IMPEDANCE ANALYSERS IN IBD PATIENTS RECEIVING JAK INHIBITOR THERAPY: AGREEMENT, PROPORTIONAL BIAS, AND CLINICAL INTERCHANGEABILITY
LB038
COMPARING MULTI-FREQUENCY BIOELECTRICAL IMPEDANCE ANALYSERS IN IBD PATIENTS RECEIVING JAK INHIBITOR THERAPY: AGREEMENT, PROPORTIONAL BIAS, AND CLINICAL INTERCHANGEABILITY
D. Vranesic Bender1,*, M. Sabic1, I. Karas1, I. Kodvanj2, I. Kodvanj1, A. Kunovic1, D. Ljubas Kelecic1, A. Horvat1, Z. Racic1, Z. Krznaric1
1Unit of Clinical Nutrition and Intestinal Failure Unit, 2Division of Gastroenterology, Department of Internal Medicine, University Hospital Centre Zagreb, Zagreb, Croatia
Rationale: Body-composition assessment is important in the management of inflammatory bowel disease (IBD), but agreement between commercially available measurement devices has not been well established. We investigated the methodological agreement of two bioelectrical impedance devices in a cohort of patients with IBD receiving JAK inhibitors.
Methods: Thirty-two adult patients with IBD on JAK inhibitors underwent paired body-composition assessment using a Bodystat Multiscan 5000 multi-frequency bioelectrical impedance analyser and a Tanita MC-780 (BC-420MA). Measurements of fat mass (FM), fat percentage, fat-free mass (FFM), total body water (TBW), TBW percentage, extracellular water (ECW), intracellular water (ICW), skeletal muscle mass (SMM), and phase angle were compared. Agreement was assessed using Bland–Altman analysis, Lin's concordance correlation coefficient (CCC), intraclass correlation coefficient (ICC(2,1)), and proportional bias testing.
Results: Agreement was good for FM, fat percentage, FFM, TBW, ECW, and SMM, with CCC ≥0.92, ICC(2,1) ≥0.93, and Bland–Altman biases below 1.5 kg (or 1.5 percentage points). TBW%, ICW, and phase angle showed poor agreement (CCC: 0.861, 0.853, and 0.749; ICC: 0.864, 0.857, and 0.755). For phase angle, the lower bound of the ICC 95% CI crossed zero (−0.011), indicating compatibility with no inter-device reliability. FFM showed statistically significant proportional bias (p = 0.005), with the Bodystat–Tanita difference widening systematically at higher FFM values.
Conclusion: The Bodystat Multiscan 5000 and Tanita MC-780 demonstrated acceptable group-level agreement for most body-composition outcomes, but the devices are not interchangeable, particularly for TBW%, ICW, phase angle, and FFM in patients with higher FFM values.
Disclosure of Interest: None declared