Objective Infants born extremely preterm (EP, <28 weeks’ gestation) exhibit poorer growth and neurodevelopmental impairment in early childhood compared with their term-born peers. Whether poor growth persists and whether associations of growth with neurodevelopmental functioning have changed in the decades since the introduction of surfactant are not well described. This study aims to (1) compare growth from birth to 2 years then 8 years in children born EP between three different eras, and (2) investigate the associations of growth from birth to 2 years then 8 years with cognitive, academic, executive and motor function at 8 years, and if associations have changed over time.
Design Prospective observational cohort studies in the State of Victoria, Australia in three discrete eras: 1991–1992, 1997 and 2005. EP children had weight and head circumference measured at birth, and weight, head circumference and height at 2 and 8 years. Cognitive ability, academic performance, executive function and motor skills were assessed at 8 years, corrected for prematurity.
Results 499/546 (91%) of surviving EP children were fully assessed at 8 years. Growth in children born EP did not differ substantially between eras and associations between growth and neurodevelopment did not change over time. Overall, better weight and head growth from birth to 2 years were associated with improved neurodevelopment at 8 years.
Conclusions Growth of children born EP has not improved in more recent eras. Better early head and weight growth are associated with improved neurodevelopment in mid-childhood.
- extremely preterm
- long-term outcomes
Statistics from Altmetric.com
What is already known on this topic?
Children who are born extremely preterm (EP) have lower anthropometric measurements and lower scores on neurodevelopmental assessments than their term counterparts in early childhood.
What this study adds?
Among EP survivors, growth to mid-childhood is not improving over time. Better growth in the first 2 years is associated with improved neurodevelopment at 8 years.
Approximately 750 000 infants are born extremely preterm (EP, 22–27 completed weeks’ gestation) worldwide each year.1 Survival rates have increased since the introduction of antenatal corticosteroids, surfactant and an increased willingness to offer intensive care; however, improved survival rates have not led to improvements in neurosensory outcomes.2 3 Impaired neurodevelopment following preterm birth can have a life-long impact on the child, their family and society, including negative effects on mental health, employment rates and healthcare costs.4
Children born EP are exposed to many deleterious postnatal stressors during a period of rapid brain growth.5 Smaller brain volumes and abnormal brain microstructure compared with term newborns have been reported as early as term equivalent age.6 However, knowledge is lacking regarding interventions that may mitigate abnormal brain development in the EP population. Optimal brain growth in preterm infants is associated with better nutrition in the first 2 weeks of life, and specifically with ratios of energy, lipids and protein7; however, providing adequate nutrition to EP infants is difficult to achieve. Consequently, EP children grow poorly and lag behind term-born peers in all anthropometric measures through mid-childhood and beyond.8
Some studies have reported associations between postnatal growth impairment and deficits in early childhood neurodevelopment,9 10 while others have not.2 We previously reported that poor early head growth was associated with lower cognitive, academic and motor test scores at 8 years in EP children born in 1991–1992,11 but it is unclear whether growth is improving or if the same relationships of growth with neurodevelopment exist in more recent eras.
The aims of this study were (1) to compare growth from birth to 2 years and 2 to 8 years, and (2) to investigate the associations of growth with cognitive, academic, executive and motor function at 8 years, across three eras from 1991 to 1992, 1997 and 2005. We hypothesised that growth would improve over the eras and that better head growth would be associated with improved neurodevelopment in mid-childhood.
Patients and methods
All EP (22–27 completed weeks’ gestation) survivors in the state of Victoria were recruited in three distinct eras: 1991–1992 (24 months), 1997 (12 months) and 2005 (12 months), as previously described.12 Controls comprised infants born with gestational ages ≥37 weeks or a birth weight ≥2500 g, matched for social variables; controls were necessary to standardise test scores at 8 years for some outcome variables. Participants attended follow-up visits at 2 and 8 years, outcomes for which have been reported elsewhere.2
The studies were approved by the Human Research Ethics Committees at participating centres. Written informed consent was obtained from parents for controls and for the 2005 EP cohort; follow-up was considered routine care for EP children in earlier cohorts.
Perinatal and sociodemographic data
Perinatal and sociodemographic data were collected prospectively. Gestation at birth was confirmed by ultrasound before 20 weeks in 90% of participants, or menstrual history in the remainder. Sociodemographic data included years of maternal education (dichotomised into lower and higher around median years of schooling) and social class (based on the occupation of the major income earner in the family and categorised as lower (unskilled or unemployed) or higher (semiskilled, skilled, or professional)).
Weight and head circumference were measured at birth, and weight, height and head circumference were measured at 2 and 8 years, corrected for prematurity, by trained staff using standardised equipment. Body mass index (BMI) was calculated at 2 and 8 years using the formula: (weight in kilograms)/(height in metres squared). All measurements were converted to z-scores relative to the British Growth Reference data, which adjusts measurements for age and sex.13
Cognitive, academic, executive function and motor assessments
All assessments at 8 years’ corrected age were performed by assessors blinded to the birth weight and gestation of participants. Age was corrected for prematurity to avoid a known bias in cognitive test scores.14 Cognitive function was tested with the Wechsler Intelligence Scales for Children, third edition15 for the 1991–1992 cohort; the fourth edition16 for the 1997 cohort; and the Differential Ability Scales, second edition17 for the 2005 cohort. Academic skills of reading, spelling and arithmetic were assessed using the Wide Range Achievements Test, version 318 for the 1991–1992 and 1997 cohorts, and version 419 for the 2005 cohort. Because different versions of the tests were used across eras, cognitive and academic results were converted to z-scores relative to the mean scores of contemporaneous controls within each era, weighted for the social variables of the EP cohorts, as previously described.2 Children too impaired to complete cognitive or academic testing were assigned scores of −4 SD.
Executive function in everyday life was assessed using the parent-completed Behaviour Rating Inventory of Executive Function questionnaire.20 Executive dysfunction was defined as T-scores >65 on the Global Executive Composite, the Behavioral Regulation Index or the Metacognition Index.
Cerebral palsy was diagnosed in children with abnormal tone or tendon reflexes who had a loss of motor skills. Motor performance was tested with the Movement Assessment Battery for Children (MABC)21 for the 1991–1992 and 1997 cohorts and the MABC-222 for the 2005 cohort. Children with severe motor impairments who could not perform the MABC were given a centile score of 0.1. Poor motor function was defined as either cerebral palsy or motor impairment on the MABC (<5) or MABC-2 (≤5).
Data were analysed using Stata V.15.1 (Stata Corp, College Station, TX). Growth of weight and head circumference from birth to 2 years, and of weight, head circumference, height and BMI from 2 to 8 years were compared between eras using mixed-effects linear regression models, with individual models for each measurement outcome, and a fixed effect of era and an interaction between age and era (Aim 1). Mixed models allow inclusion of individuals with at least one measurement at any time point.
For Aim 2, growth was calculated between birth and 2 years, and between 2 and 8 years as the z-score difference over that time for individuals with data points at both ages. To assess if relationships of neurodevelopment with growth differed between eras, models included an interaction term for era and growth. Relationships of cognitive and academic function with growth were explored using linear regression fitted using generalised estimating equations with robust (sandwich) estimation of standard errors to account for multiple births. Results are reported as regression coefficients (change in SD score for the 8-year outcomes associated with a 1 SD increase in each growth variable per year between the two relevant time points; viz., birth to 2 years or 2–8 years) with 95% CIs. Similar analyses were conducted for dichotomised measures of executive and motor dysfunction using logistic regression with results reported as ORs with 95% CIs. Results were also adjusted for lower maternal education and lower social class. Given the high follow-up rate, we considered multiple imputation for missing data was unnecessary.
Data were interpreted primarily on the strengths and directions of the associations. We sought stronger evidence (p<0.01) for important interaction terms looking for differences between eras, in acknowledgement that we have performed many different tests.
Overall, 503 EP survivors were assessed at 8 years, 499/546 (91%) of whom were fully assessed. Perinatal, sociodemographic and 8-year assessment data are shown in table 1. As previously reported, EP children in 2005 had lower scores for academic achievement, increased executive dysfunction and worse motor function than their 1991–1992 and 1997 counterparts.2 23 24
Anthropometric measurements and growth
All z-scores were approximately normally distributed. The means for all measurements were below zero at each time point across all three eras (figure 1, online supplementary table 1), meaning that on average children were smaller than expected for age and sex. The most negative scores were seen for head circumference at 2 and 8 years, which exceeded more than 1 SD below the expected mean (figure 1, online supplementary table 1). There was some evidence that growth in height between 2 and 8 years was lower in the 2005 than the 1991–1992 cohort, but little evidence for growth differences in the other anthropometric measures across the eras (table 2).
Associations of cognitive, academic, executive function and motor outcomes at 8 years with growth
There was no strong evidence that the relationships of developmental outcomes with growth between birth and 8 years differed between eras (all p values for interactions >0.01), so all data were combined for subsequent analyses.
There was strong evidence that greater head circumference growth between birth and 2 years was associated with higher IQ and reading scores and lower odds of executive dysfunction or poor motor function at 8 years (tables 3 and 4). These relationships were similar when adjusted for lower maternal education and lower social class. There was little evidence of associations between changes in head circumference z-scores from 2 to 8 years and any outcomes at 8 years (tables 3 and 4).
There was evidence that reading scores at 8 years were positively associated with growth in weight between birth and 2 years, but the association was in the opposite direction between 2 and 8 years (table 3). There was little evidence for any other associations of 8-year neurodevelopmental outcomes with changes in weight, height or BMI z-scores between any other time points. After adjustment for maternal education and social class, the strength of the evidence increased enough so that better growth in weight between birth and 2 years was associated with higher IQ and with lower odds of poor motor function.
This population-based study of three sequentially recruited cohorts of EP infants demonstrated ongoing impairment of growth until at least school age, with no signs of improvement over time, contrary to our expectations. Of the growth measurements, there was an increase in head circumference z-scores from birth to 2 years associated with improved cognitive, academic, executive function and motor outcomes at school-age, as hypothesised, and also there were some associations of improved neurodevelopment with better growth in weight between birth and 2 years, which were unexpected. The findings of lower scores for academic achievement and worse executive and motor function in 2005 compared with the earlier eras have been reported previously.2 23 24
That preterm children suffer from poor growth has been reported on many occasions previously.9 25 Scharf et al, who examined 650 children at 2 years from a birth cohort of 920 infants born preterm (<37 weeks), reported that a high proportion had z-scores for head circumference, weight, height and BMI >−2 SD below the norm and Ehrenkranz et al, who followed 450 EP infants, found that those in the lowest quartile for growth during primary hospitalisation had weight and head circumference values below the 10th percentile at 18–22 months.26 27 The persistence of growth impairment in EP children into school-age has also been reported.28 29 In one study, Bracewell reported that infants born before 25 weeks’ gestation remained smaller at 6 years compared with population norms.30 We had anticipated that growth to mid-childhood would have improved in the 2005 compared with the 1991–1992 cohort so the findings of the current study are contrary to our expectations.
Suboptimal head growth has previously been linked with poorer neurodevelopmental outcomes following preterm birth from early childhood to adulthood. For example, infants born before 32 weeks’ gestation who had the smallest increases in head size in the first 3 months post-hospital discharge had lower IQ scores at 5 years,28 British children born EP with poor early head growth performed poorly in tests of motor skills and IQ at 7 years,31 and former very low birth weight infants with lower head growth velocity from birth to term-equivalent age were found to have lower neurocognitive, executive function and visual memory skills when assessed at 25 years.32
The results of the current study confirm our hypothesis that poor early head growth was associated with poorer neurodevelopment across multiple domains in mid-childhood, consistent with other follow-up studies.10 27 29 33 If interventions to improve early head growth can be found, they might also improve important functional school-age outcomes. That we also had some positive associations of improved early growth in weight between birth and 2 years with better neurodevelopment suggests that nutrition to increase weight gain in the early years should be considered an area for further research.
There are many reasons why children born EP are smaller than expected on all anthropometric measurements, including limitations on fluid volumes in the first weeks of life, interruptions to nutrient supply due to competing demands on intravenous lines and feed intolerance, nutrient malabsorption due to gastrointestinal tract immaturity and the mismatch between energy supply and expenditure related to thermoregulation and breathing.34 In addition, diversity in clinician-led nutrition practices, despite the availability of international guidelines for nutrition in preterm infants,35 have been shown to be associated with poorer in-hospital growth.36 Future research should focus on devising and assessing nutritional interventions, such as modifying caloric and nutrient intake or ratios (carbohydrate/fat/protein) to optimise existing standardised feeding protocols. Long-term follow-up studies of children born EP should include data collection on important growth and nutrition outcomes and associations with later development to inform clinical practice.
Recent studies investigating the effects of different nutrition regimes on the brain of preterm infants have suggested a close association between appropriate nutrient intake and brain development.7 37 38 However, uncertainty regarding the optimal approach remains with some studies supportive of high-calorie, protein and lipid regimes for improving outcomes,39 40 and others suggesting that even with improved growth, significant neurodevelopmental impairment may not be reduced and high nutrient intakes may even be harmful.41 In addition, growth and neurodevelopmental impairment in children born EP are multifactorial and better nutrition alone will not ameliorate all of the effects on the brain and other organs that follow preterm birth. However, we recommend routine follow-up of nutrition and growth to identify children with growth failure early so that interventions, such as dietary advice, can be actioned.
The strengths of our study include the high rates of follow-up of sequential geographical cohorts and the ability to determine outcomes to mid-childhood over three distinct time periods. In addition, recruitment based on gestational age at birth rather than weight means we do not have an over-representation of small-for-gestational-age infants. Having repeated anthropometric measurements over time allows for investigation of the association between growth and neurodevelopment from birth to mid-childhood, rather than just at birth or hospital discharge. Pooling of data from three cohorts improved the power to detect associations between growth and neurodevelopment, so that we were able to find associations of early growth in weight with improved neurodevelopment, whereas we could not when we looked previously in the first cohort alone.11 The main limitation is that these cohort studies were not designed to investigate the effects of nutritional interventions on growth and neurodevelopmental outcomes, hence, detailed data on nutritional intake were not available.
EP children remain at risk of poor growth from birth to mid-childhood. Better head and weight growth in early childhood are associated with better cognition, academic performance, executive and motor function. Therefore, it is essential that EP children undergo ongoing close surveillance to identify early growth failure followed by appropriate and timely interventions, to ensure their growth and neurodevelopmental potential are optimised.
Collaborators Members of the Victorian Infant Collaborative Study Group—Convenor: Jeanie Cheong (Neonatal Services, Royal Women’s Hospital, Melbourne, Australia; Victorian Infant Brain Studies, Murdoch Children's Research Institute, Melbourne, Australia; Department of Obstetrics & Gynaecology, University of Melbourne, Melbourne, Australia; Premature Infant Follow-up Program, Royal Women’s Hospital, Melbourne, Australia). Collaborators (in alphabetical order): Peter Anderson (Victorian Infant Brain Studies, Murdoch Children's Research Institute, Melbourne, Australia; Premature Infant Follow-up Program, Royal Women’s Hospital, Melbourne, Australia; Department of Paediatrics, University of Melbourne, Melbourne, Australia), Alice Burnett (Victorian Infant Brain Studies, Murdoch Children's Research Institute, Melbourne, Australia; Premature Infant Follow-up Program, Royal Women’s Hospital, Melbourne, Australia; Department of Paediatrics, University of Melbourne, Melbourne, Australia; Neonatal Medicine, The Royal Children’s Hospital, Melbourne, Australia), Catherine Callanan (Premature Infant Follow-up Program, Royal Women’s Hospital, Melbourne, Australia), Elizabeth Carse (Newborn Services, Monash Medical Centre, Melbourne, Australia), Margaret P Charlton (Newborn Services, Monash Medical Centre, Melbourne, Australia), Noni Davis (Premature Infant Follow-up Program, Royal Women’s Hospital, Melbourne, Australia), Lex W Doyle (Neonatal Services, Royal Women’s Hospital, Melbourne, Australia; Victorian Infant Brain Studies, Murdoch Children's Research Institute, Melbourne, Australia; Department of Obstetrics & Gynaecology, University of Melbourne, Melbourne, Australia; Premature Infant Follow-up Program, Royal Women’s Hospital, Melbourne, Australia; Department of Paediatrics, University of Melbourne, Melbourne, Australia), Julianne Duff (Premature Infant Follow-up Program, Royal Women’s Hospital, Melbourne, Australia), Leah Hickey (Department of Paediatrics, University of Melbourne, Melbourne, Australia; Neonatal Medicine, The Royal Children’s Hospital, Melbourne, Australia), Esther Hutchinson (Premature Infant Follow-up Program, Royal Women’s Hospital, Melbourne, Australia; Neonatal Medicine, The Royal Children’s Hospital, Melbourne, Australia), Marie Hayes (Newborn Services, Monash Medical Centre, Melbourne, Australia), Elaine Kelly (Premature Infant Follow-up Program, Royal Women’s Hospital, Melbourne, Australia; Neonatal Services, Mercy Hospital for Women, Melbourne, Australia), Katherine J Lee (Clinical Epidemiology and Biostatistics, Murdoch Children's Research Institute, Melbourne, Australia), Marion McDonald (Premature Infant Follow-up Program, Royal Women’s Hospital, Melbourne, Australia), Gillian Opie (Neonatal Services, Mercy Hospital for Women, Melbourne, Australia), Gehan Roberts (Victorian Infant Brain Studies, Murdoch Children's Research Institute, Melbourne, Australia; Premature Infant Follow-up Program, Royal Women’s Hospital, Melbourne, Australia; Department of Paediatrics, University of Melbourne, Melbourne, Australia; Centre for Community and Child Health, The Royal Children’s Hospital, Melbourne, Australia), Amanda Williamson (Neonatal Services, Mercy Hospital for Women, Melbourne, Australia).
Contributors LH, JLYC and LWX conceived and designed the study, performed data analysis and interpretation, and drafted and revised the article. KL provided statistical advice and performed data analysis. AB, AJS, GR and PA assisted in drafting and revising the article. All authors approved the final version as submitted and agree to be accountable for all aspects of the work.
Funding Supported by grants from the National Health and Medical Research Council of Australia (Centre of Clinical Research Excellence #546519; Centre of Research Excellence #1060733; Project Grant #108702; Career Development Fellowship #1141354 to JLYC, #1127984 to KL, #1108714 to AJS) and the Victorian Government’s Operational Infrastructure Support Program.
Disclaimer The funding sources had no role in the study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.
Competing interests None declared.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available on reasonable request. Authors can be contacted at the corresponding author contact details to access additional data on reasonable request.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.