Background Weight gain of preterm infants in the UK is assessed using charts constructed from cross-sectional gestation based birth data, the British 1990 reference (B1990R). Cross-sectional weight charts may not reflect the realistic postnatal growth potential of preterm infants. The optimal postnatal weight gain that leads to long-term health is unknown and aggressive nutritional strategies to meet the current standards may not be without risks to short and long-term health.

Aim To compare routinely recorded longitudinal weight data of infants <32 weeks gestation, and cared for in Wales, with the B1990R.

Methods Daily weight and clinical information recorded in the ‘Badgernet’ neonatal database, from birth till discharge, was collected for all eligible infants over a 4-year period (2011–2014). The B1990R was used to compute weekly gender and gestation specific standard deviation scores (z-scores) and analysed by pre-defined gestation groups. Multivariate analysis was undertaken to determine the effect of confounding variables on weight gain.

Results Data from 1427 preterm infants was analysed (Table 1). Despite being similar at birth, weekly z-scores of our cohort were significantly lower than predicted by the 1990 chart throughout hospital stay; z scores of male infants were significantly lower compared to females. This pattern was consistent across all gestation groups. After adjustment for confounding variables, gestation at birth, duration of ventilation and postnatal steroids remained independent factors that significantly affected z-score at discharge.

• Download figure
• Open in new tab
• Download powerpoint

Abstract G404(P) Table 1

Main results

Conclusion The longitudinal weight profile of preterm infants in Wales is significantly below predicted standards as per the B1990R. While suboptimal nutrition is multifactorial with scope for further improvement, we speculate that the observed growth failure is significantly contributed by an unrealistic expectation of postnatal growth. A large population based longitudinal growth chart may be more representative of appropriate growth in this population.