Weight for length measures may not accurately reflect adiposity in preterm infants born appropriate for gestational age during hospitalisation or after discharge from the neonatal intensive care unit

Background: Weight/length (W/L) indices are poor surrogates for adiposity in preterm infants born appropriate for gestational age (AGA) at birth, but whether the association subsequently improves is unknown. Objective: To determine if W/L indices accurately reflect adiposity in premature infants born AGA in later infancy. Methods: Associations between W/L indices and fat mass, fat mass index and percent body fat (%BF) obtained via air displacement plethysmography (ADP) were examined in 260 preterm infants (majority born AGA) at 28 to 63 weeks' postmenstrual age (PMA). Accuracy of W/L indices as indicators of adiposity was assessed by proportion of variance explained (R²) and root mean square error from linear regression of adiposity on W/L indices and proportion of infants misclassified by W/L indices. Accuracy was further compared in term vs preterm infants at term-equivalent age. The impact of early vs late preterm status on associations between W/L indices and %BF was also examined. Results: BMI and W/L were most strongly associated with %BF but yielded poorly fitting models (maximum R² = 0.35; 53% misclassification). A significant interaction of W/L indices and early vs late preterm status on %BF revealed that estimation of %BF differs by status. Accuracy of W/L indices was worse in preterm infants at term-equivalent age. Conclusions: W/L indices were not good indicators of adiposity in preterm infants from 28 to 63 weeks' PMA (born AGA) with all categories of W/L indices combined. Future research should examine whether results are similar in preterm infants born with disproportionate W/L or who experience disproportionate growth postnatally.