Routine abdominal magnetic resonance imaging can determine psoas muscle area in paediatric Crohn's disease and correlates with bioelectrical impedance spectroscopy measures of lean mass

Background: Paediatric Crohn's disease (CD) has been associated with undernutrition. Accurate and accessible measures of body composition would provide data to personalise nutritional therapy. We assessed feasibility of MRI-derived measures of psoas cross-sectional area (PCSA) in paediatric CD and correlated with anthropometric and bioelectrical impedance spectroscopy (BIS) measures.

Methods: MRI small bowel/pelvis images of patients with CD, aged <18 years, were retrieved. Patients with concurrent anthropometric and BIS measurements were eligible for inclusion. The PCSA at L3 was calculated by two assessors and combined. To assess reproducibility of measures we calculated the coefficient of variation (CoV). Age, height-Z-scores, weight-Z-scores and BIS measures were correlated with PCSA. Using normal paediatric data from CT-scans we derived psoas area Z-scores for our cohort.

Results: 10 patients were included. Mean age at MRI scan was 14.6 years (11.7-16.3). PCSA was calculated for all MRI scans. There was high reproducibility between measurers, mean CoV 0.099. There was a significant positive correlation between PCSA and BIA-derived fat free mass, Pearson correlation coefficient (PCC) 0.831, p = 0.003. Correlation coefficients for PCSA and Height-for-age Z-score, weight-for-age -Z-score and age were PCC 0.343- p = 0.33, PCC = 0.222- p = 0.54, and PCC 0.6034- p = 0.065, respectively. The mean PCSA Z-score was -1.81, with 70% of the patients having a Z-score < -2.0.

Conclusions: These data demonstrate the feasibility of deriving measures of body composition from routine MRI imagine. There was significant positive correlation between PCSA and BIS-derived lean mass. Further studies are required to confirm applicability of normal ranges prior to routine clinical implementation.