Do anthropometric measures accurately reflect body composition in preterm infants?

Abstract

Recent literature suggests that neonatal adiposity is predictive of later metabolic health, while neonatal lean mass is predictive of later cognitive function amongst preterm infants. Anthropometric indices that accurately reflect neonatal body composition could improve clinical care and aid future research, but their validity has not been systematically tested in preterm infants. To determine the weight/length indices that best reflect neonatal body composition in preterm infants. Weight and length were measured, and body composition (fat-free mass (FFM), fat mass (FM) and percent body fat (%BF)) was assessed using air-displacement plethysmography within 72 h of birth in 218 preterm infants. The best weight/length proxy for FFM, FM and %BF were those with the highest proportion variance explained (R2 ) and lowest root mean square error (RMSE) in linear regression models. Among all of the weight/length indices tested, weight/length2 was the best proxy for %BF, but nonetheless exhibited very low variance explained (R2 = 0.27) and high prediction error (RMSE = 3.5% fat). Body weight unadjusted for length was strongly associated with FFM (R2 = 0.97). No weight/length index accurately reflected %BF. Weight/length indices are not appropriate for assessment of relative adiposity in preterm infants near birth. What's known on this subject: Compared with term infants, preterm infants have increased fat mass and diminished fat-free mass upon hospital discharge. Early adiposity predicts later metabolic health, while early lean mass is predictive of later neurodevelopmental outcomes. Optimal anthropometric proxies for preterm body composition at birth are not established. Weight is an adequate surrogate for lean mass at birth in preterm infants. There are no weight/length indices that accurately reflect neonatal adiposity at birth.