Neurodevelopmental Outcomes at School Age After Extremely Preterm Birth

In this issue of Pediatrics, Serenius et al report outcomes at age 12 years from a national cohort of 462 children who were born less than 27 weeks’ gestational age (GA) in Sweden from 2004 to 2007.1 Compared with 373 term-born controls, the children born less than 27 weeks’ GA had higher rates of neurodevelopmental disability (NDD), psychiatric disorders, abnormal motor function, and behavioral problems; 22.1% had multimorbidity compared with 1.1% of controls.1 Of note, 37.4% (173/462) had a moderate-severe NDD (any of an IQ <−2 SDs relative to controls, moderate or severe cerebral palsy Gross Motor Function Classification System levels 2–5, or blindness or deafness). The rate of moderate-severe NDD at 12 years was similar to when the cohort had been assessed at age 6.5 years (33.6% 148/441).2 Serenius et al should be congratulated on obtaining vital follow-up data up to early adolescence on a complete country-wide cohort of the most premature survivors.There are several other regional cohorts of survivors including those born less than 27 weeks’ GA and accompanying term-born controls since the 1990s that have reported rates of moderate-severe NDD at school age, but only from high-income countries.3–6 In all these studies it is possible to identify rates of moderate-severe NDD using criteria that are almost identical to the Swedish study, including IQ less than −2 SDs relative to term-born controls, which is the major contributor to NDD. Extracting data from these studies for those born less than 27 weeks’ GA and less than 26 weeks’ GA, the rates of moderate-severe NDD are similar to those reported by Serenius et al (Table 1). Only the Australian study corrected age for prematurity when the children were assessed, which will have contributed to this study’s low rates of NDD.There are no comparable regional data on moderate-severe NDD at school age reported from the United States. The closest would be from the ELGAN study, which recruited infants born 23–27 weeks’ GA from 14 US hospitals in 2002–2004. The rate of moderate-severe NDD, defined almost identically to the Swedish study, was 22.9% (184/802) at age 10 years.7Because the children in the Swedish study were born 2 decades ago, it is pertinent to consider if there are any perinatal or later interventions that have subsequently been introduced into clinical care in high-income countries that may improve long-term neurodevelopment.Magnesium sulfate given antenatally to women likely to deliver preterm reduces the risk of cerebral palsy in the child8 and has been used in clinical practice in many high-income countries for the past 15 years. Forty-nine of 462 (10.6%) children in the Swedish study had cerebral palsy, which was moderate or severe in one-half, so the rate of cerebral palsy in survivors born less than 27 weeks’ GA, and hence the rate of moderate-severe NDD, might be expected to fall in cohorts born today who are exposed to magnesium sulfate for fetal neuroprotection. A fall in the rate of cerebral palsy in survivors born less than 28 weeks’ GA in the state of Victoria, Australia, has been observed after magnesium sulfate was introduced into clinical practice.9Delayed cord clamping compared with early or immediate cord clamping at birth may reduce mortality to hospital discharge and the rate of death or NDD at 2–3 years in infants born preterm.10 However, data into school age from delayed cord clamping are unavailable. The American College of Obstetricians and Gynecologists has recommended delayed cord clamping for infants born preterm since 2020.11Respiratory support practices at birth and in the neonatal intensive care unit have changed since the mid-2000s. There is now a greater emphasis on noninvasive respiratory support, in particular nasal high flow, which was rare in the mid-2000s but is now common, and a reduction in invasive respiratory support.12 However, regional data on neurodevelopmental benefits at school age related to changes in newborn respiratory support are lacking.Although there have been numerous randomized controlled trials investigating various aspects of neonatal nutrition since the mid-2000s, the only nutritional intervention with evidence of benefit to school age for infants born extremely preterm has been from oral supplementation of feeds with docosahexaenoic acid (DHA). Among 656 children born less than 29 weeks’ gestation and assessed at age 5 years, treatment with 60 mg of DHA per kg per day, from 3 days after starting enteral feeds to 36 weeks’ postmenstrual age, resulted in an improvement in IQ of 3.5 points (95% CI, 0.4–6.5).13 Since the study results were reported in 2022, we are unaware of any widespread implementation of supplementing DHA in the enteral feeds of infants born extremely preterm, which would be required to reduce NDD in whole populations of infants born extremely preterm.Caffeine to prevent apnea in the newborn period in infants less than 1251 g birthweight improves neurodevelopmental outcomes up to age 11 years.14,15 Although it is likely that many children in the Swedish study would have been treated with caffeine in the newborn period between 2004 and 2007, it is almost certain that more infants born less than 27 weeks’ GA today would be treated with caffeine than in the mid-2000s, particularly since the first long-term results from the landmark Caffeine for Apnea of Prematurity study were only reported in 2007.16Developmental interventions started after discharge home improve motor development in early childhood and cognitive scores in children up to age 5 years, but evidence for benefit into later school age is lacking.17 Because some of the developmental interventions are costly and resource-intensive and are unable to be used in everyday clinical practice, current efforts are focusing on developmental interventions that can be delivered more widely and cheaply, such as online.More whole-population studies of long-term outcomes of survivors born less than 27 weeks’ GA, such as that reported by Serenius et al, will be required in the future. These must include not only newer birth cohorts, to keep up with the rapid changes in perinatal and neonatal intensive care, but also studies of existing cohorts into later adult life, as few adults today have been born less than 27 weeks’ GA. In addition to the neurodevelopmental, psychiatric, motor, and behavioral outcomes studied by Serenius et al, other outcomes including physical health, growth, quality of life, academic achievement, and social functioning, as well as parental and family outcomes, could be assessed to obtain as complete a picture of the long-term consequences of being born <27 weeks’ GA as possible. In addition to specific evaluations that are part of the initial study design, any new interventions or cohorts would benefit from the ability to link with existing data on long-term outcomes, not only in health but also in school and in the workplace, to obtain data for as many long-term survivors as possible. Although complete data were available for less than two-thirds of both the preterm and term cohorts in the Swedish study, access to country-wide databases allowed for some data to be obtained and outcomes to be estimated on all known survivors eligible for follow-up at 12 years. Such linkage studies are more established in some countries but are restricted in many other countries for various reasons, including concerns about privacy. Despite the difficulties of conducting long-term studies into later childhood and adulthood, we must persist. If we do not continue to look for evidence of long-term benefit or harm from perinatal and neonatal intensive care, we will not find anything at all and will remain ignorant, to the detriment of infants under our care and their families.