Comparison of two bioelectrical impedance analyzers for estimating body composition in a cohort of pediatric oncology patients

Background and Aims: Unfavourable changes in body composition are frequently reported in children with cancer. An easy and affordable method to measure body composition is bioelectrical impedance analysis (BIA). In this study, we compared the Tanita MC780-MA Body Composition Analyzer (Tanita) versus the Bodystat Quadscan 4000 analyzer (Bodystat) to determine whether the Tanita is a suitable analyzer for pediatric oncology patients. Methods: In this study, 84 childhood cancer patients/survivors were included. Per patient 1, 2 or 3 visits were planned during or after treatment. During each visit, BIA measurements were performed first with the Tanita analyzer followed by the Bodystat analyzer. In total, 131 measurements were included. Spearman correlation and concordance correlation coefficients of Lin (CCC) were calculated for fat percentage (Fat%), fat mass (FM) and fat free mass (FFM). Bland-Altman plots were constructed to assess the agreement between both analyzers. Results: Fat%, FM and FFM values differed statistically significantly when comparing the BIA devices (Wilcoxon signed rank test, P<0.001), but absolute differences between medians were small. Spearman correlations and CCC's were high for FM (0.94 and 0.92, respectively) and FFM (0.99 and 0.99, respectively), but moderate for Fat% (0.70 and 0.65, respectively). Bland-Altman plots did not show low differences of means (biases; for Fat%, FM and FFM 2,6%, 0,6 kg and -0,6 kg respectively), but relatively large limits of agreement (-9.77%–15.89%, -5.6 kg–7.2 kg and -5.9 kg–4.6 kg respectively). Conclusions: Based on the good agreement between the Tanita and Bodystat on group level, and the fact that the Tanita is a more patient-friendly device, the Tanita can be used to measure body composition instead of the Bodystat in pediatric oncology patients, making it easier and faster to gain information about body composition.