Developing Predictive Models For Postnatal Growth Of Preterm Infants During And After Unimpaired Postnatal Adaptation

Abstract

<p><strong>Background:</strong> Postnatal growth of preterm infants does not match recommended intrauterine growth, due to the initial weight loss that accompanies healthy body composition rearrangements after birth. Thus, optimal postnatal growth for preterm infants is currently unknown.</p> <p><strong>Objectives: </strong>(1)<strong> </strong>Collect longitudinal postnatal growth data of 30–36 week GA preterm infants with unimpaired postnatal adaptation; (2) Develop regressions that predict the growth trajectory such an infant will adjust to by days of life 7/14/21; (3) Extrapolate and validate the regressions downwards to 25 weeks.</p> <p><strong>Methods:</strong> Infants of 30–36 week GA, born/admitted to 1/5 participating centres between 2008–2012, who met pre-specified criteria for unimpaired postnatal adaptation and who had at minimum 14 days of data were included. Day-specific anthropometric data from birth to discharge were abstracted retrospectively. Z-score regressions for days 7/14/21 were developed. Regressions were then extrapolated to 25 weeks and validated using an independent study population.</p> <p><strong>Results:</strong> Of 6203 infants, 665 met the screening criteria. By day 14, infants adjusted to stable growth trajectories that were 84±13% of the recommended weight-for-age. Using the following predictors: GA, z-score at birth and hospital-centre, regressions accurately predicted z-scores at days 7, 14 (n=665; R²=0.939, 0.889) and 21 (n=333; R²=0.841). Validation using 25-29 week GA infants (n=173) suggested models were also accurate within this age-range.</p> <p><strong>Conclusion: </strong>These results provide robust estimates of a hypothesis of healthy postnatal growth for preterm infants. Future steps include assessing long-term outcomes in a randomized control trial and assessing the quality of growth using body composition analyses.</p>