Body composition assessment in 6-month-old infants: A comparison of two- and three-compartment models using data from the Baby-bod study.

Abstract

An appreciation of infant body composition is helpful to understand the 'quality' of growth in early life. Air displacement plethysmography (ADP) using PEA POD and the deuterium dilution (DD) technique are commonly used body composition approaches in infants. We evaluated the comparability of body composition assessed using both techniques with two-compartment (2C) and three-compartment (3C) models in 6-month-old infants. Infant fat mass (FM) and percent fat mass (%FM) obtained from a 2C model using PEA POD (2C-PP) and a 2C model using the deuterium dilution technique (2C-DD) were compared to those derived from a 3C model, and to each other, using Bland-Altman analysis and Deming regression. Measurements were available from 68 infants (93% Caucasian, 53% male). The mean biases were not significant between any of the method comparisons. However, significant constant and proportional biases were identified in 2C-DD vs 3C and 2C-PP vs 2C-DD, but not in the 2C-PP vs 3C comparison. Furthermore, we observed significant associations between the mean differences and infants' percent total body water (%TBW). While no significant between-method mean differences were found in body composition estimates, some comparisons revealed significant constant and proportional biases and notable associations between the mean differences and %TBW were observed. Our results emphasise the importance of method choice, ensuring methodological uniformity in long-term studies, and carefully considering and regulating multiple pre-analytical variables, such as the hydration status of the participants.