Nutrition in the month after birth and clinical and neurodevelopmental outcomes in extremely low birthweight infants

Abstract

The survival of extremely preterm infants has improved markedly with advances in perinatal care; however, these infants continue to have higher rates of neurodisability than their term-born counterparts. Adequate nutrition is necessary to enhance growth and development outcomes; however, there is little consensus on what nutrition regimens support optimal neurodevelopmental outcomes, particularly in the most susceptible infants: those born of an extremely low birthweight. The primary aim of this thesis was to investigate the associations between nutritional intake in the first 28 days after birth and neurodevelopmental outcome at 2 years corrected age in a large cohort of extremely low birthweight infants, with a secondary aim to assess clinical outcome. This includes a sub-group analysis of nutritional intake by ethnicity. Our research is a secondary cohort analysis of the ProVIDe trial, a multicentre, double-blind, randomised controlled trial of 434 extremely low birthweight infants admitted to 8 neonatal intensive care units in New Zealand and Australia. Prospectively collected de-identified trial data were acquired from the ProVIDe trial investigators. Linear and logistic regression analyses were used to investigate associations between nutritional factors and clinical or neurodevelopmental outcomes, with adjustments for sex, smallness for gestational age, and hospital site. Multinomial regression was used for neurodevelopmental outcomes with more than two categories. The use of breastmilk fortifier and early enteral feeding was positively associated with several neurodevelopmental outcome measures. Extremely low birthweight infants who received breastmilk fortifier were 80% less likely to die or have neurodisability at 2 years’ corrected age (aOR 0.2, 95%CI 0.1, 0.4; p<0.0001) and 70% less likely to have moderate/severe neurodisability (aOR0.3, 95%CI 0.2, 0.5; p=0.0002) than those who did not receive breastmilk fortifier. Breastmilk fortifier use was also positively associated with Bayley-III cognitive and motor scores, although the effect size was small. Children with moderate/severe neurodisability had lower intakes of several nutrients than infants with no neurodisability. Children assessed as having moderate/severe neurodisability took longer to reach full enteral feeds by 3.5 days (95% CI 1.4, 5.6; p=0.001) and children with cerebral palsy 3.3 days (95%CI 0.9, 5.8; p=0.008) compared with children who did not have a disability. Infants who had their first enteral feed on day ≥3 compared with infants who received their first enteral feed on day 1 had 4-fold greater odds (AOR 4.1; 95% CI 1.7, 10.2; p=0.002) of having a Behavior Rating Inventory of Executive Function Global Executive Composite score in the impaired category. There were also several significant associations between nutritional intake and clinical outcomes. Infants with intraventricular haemorrhage or who died before discharge were 10% as likely to receive breastmilk fortifier than those who did not have an intraventricular haemorrhage or who survived to discharge (aOR 0.1, 95% CI 0.0, 0.2; p= 0.0001, aOR 0.1, 95% CI 0.0, 0.2; p<0.0001 respectively). Infants who died before discharge received lower intakes of several nutrients across all 4 weeks plausibly as a result of lower fortification rates or inclusion criteria for data of infants who died. Infants with retinopathy of prematurity and bronchopulmonary dysplasia had significantly lower breastmilk volume intakes in weeks 3 and 4 compared to infants without retinopathy of prematurity and bronchopulmonary dysplasia (retinopathy of prematurity week 3 mean difference -28.0 ml.Kg-1.d.-1 [95% CI -45.4, -10.6; p=0.002], week 4 mean difference -42.5 ml.Kg-1.d.-1 [95% CI -60.5, -24.6; p<0.0001], bronchopulmonary dysplasia week 3 mean difference -23.6 ml.Kg-1.d.-1 [95% CI -38.1, -9.1; p=0.002] and week 4 mean difference -32.7 ml.Kg-1.d.-1 [95% CI -47.8, -17.7; p=<0.0001]). Whilst those who had late onset sepsis, patent ductus arteriosus or necrotising enterocolitis had significantly lower intakes of breastmilk and several nutrients primarily in weeks 2, 3 and 4 compared to their reference groups (patent ductus arteriosus mean difference week 2 -18.8 ml.Kg-1.d.-1 [95% CI -29.5, -8.1; p=0.0006], week 3 -16.1 ml.Kg-1.d.-1 [95% CI -29.6, -2.7; p=0.02], week 4 -19.5 ml.Kg-1.d.-1 [95% CI -33.3, 5.2; p=0.007], necrotising enterocolitis mean difference week 2 -27.0 ml.Kg-1.d.-1 [95% CI -12.0, -42.0; p=0.0005], week 3 -58.6 ml.Kg-1.d.-1 [95% CI -76.7, -40.6; p<0.0001], week 4 -62.4 ml.Kg-1.d.-1[95% CI -82.1, -42.6; p=<0.0001] late onset sepsis mean difference week 2 -27.2 ml.Kg-1.d.-1 [95% CI -37.9, -16.5;p<0.0001], week 3 -45.5 ml.Kg-1.d.-1 [95% CI -58.4, -32.5;p<0.0001], week 4 -43.3 ml.Kg-1.d.-1 [95% CI -57.0, -29.6;p<0.0001]). Our results indicate that infants in this cohort with neurodisability at 2 years’ corrected age or clinical complications of preterm birth received more conservative nutrition and lower nutrient intakes in the 4 weeks after birth than those without neurodisability or clinical complications. The presence of some of these complications, e.g. interventricular haemorrhage, necrotising enterocolitis and sepsis, may directly have impacted nutritional management; for others, the reason for the association is less clear. Further investigation into the causal relationship between nutritional intake and neurodevelopmental outcomes, as well as clinical complications, in preterm infants is crucial for optimising nutrition protocols for extremely low birthweight infants.