Neurodevelopmental outcome of late-preterm infants: A pragmatic review

This article is the fifth in a series on paediatric health. Articles in this series aim to provide information about diagnosis and management of presentations in infants, toddlers and pre-schoolers in general practice.

Across the globe, approximately 15 million babies are born preterm, at <37 weeks of gestation each year.¹ Over recent years the number of preterm births, especially in the late-preterm gestation (34^0/7 to 36^6/7 weeks) category, has increased.²

Infants born preterm are at an increased risk of neurodevelopmental abnormalities.³ Therefore, most neonatal intensive care units provide routine follow-up of at-risk patients for ongoing developmental surveillance. Infants born very preterm at gestational age <32^0/7 weeks and those who had complications during the neonatal period are considered to be at high risk.³ Despite recent evidence of higher developmental risks in late-preterm infants when compared with full-term infants, routine developmental surveillance is not generally offered.⁴

With limited hospital-based resources to monitor the development of late-preterm infants, the responsibility for follow-up shifts to child health nurses and general practitioners (GPs), who need to be aware of the increased neurodevelopmental and cognitive risks for these infants. The aim of this review is to provide an understanding of the current evidence on the long-term developmental risks for late-preterm infants, compared with term infants (born at ≥37 weeks of gestation).

Selection of literature

Electronic databases (Cochrane library, Medline, Ovid) were searched to identify studies that explored the association between late-preterm infants and their long-term developmental outcomes. Only studies that reported predefined outcomes of interest were included. Details of the search methodology are available as Appendix 1 (online only). The study selection log is presented in Figure 1, and details of included studies in Table 4 (online only). Excluded studies and the reason for their exclusion are presented as Appendix 2 (online only). Only relevant results for the late-preterm category are presented.
 

Figure 1. Study selection log

Late-preterm birth and speech delay

A total of five studies (n = 8479 late-preterm infants and 27,410 term controls) reported this outcome (Table 1).^5–9 Stene-Larsen et al reported that the risk of speech delay was higher by 1.74 times at 18 months and 1.36 times at 36 months in late-preterm infants, compared with term controls. Rabie et al and Nepomnyaschy et al reported that late-preterm infants had significant speech delay at two years of age.⁶ By contrast, Gurkha et al observed that the need for speech therapy referrals for late-preterm infants and term infants was comparable.⁵ Studies used different methods for speech evaluation at different ages. Stene-Larsen et al evaluated speech at 18 and 36 months, while Nepomnyaschy evaluated speech at two years and four years, using parent-completed questionnaires.^6,8 Rabie et al diagnosed speech delay using ICD-9 coding,⁷ while Gurka et al identified referrals for speech therapy needs.⁵ Brown et al used the Peabody picture vocabulary test at four to five years of age.⁹
 

Late-preterm birth and cognitive delay

A total of eight studies (n = 11,511 late-preterm infants and 153,661 term controls) reported on cognitive outcomes (Table 2).^{4–6,10–14} Although cognitive delay, special education and need for early intervention may be different measures of the same outcome, only studies that reported a cognitive measure were included. Two studies, Marks et al and Ballantyne et al, used screening questionnaires and were excluded.^14,15 Studies used different methods to identify cognitive delay and performed cognitive tests at different chronological and corrected ages: Morag used the Griffith Mental Development Scales;¹⁴ Gurkha et al used the Woodcock-Johnson Psycho-Education Battery–Revised test;⁵ Baron et al used the General Conceptual Ability test;¹⁰ Talge et al used the Weschler Intelligence Scale for Children;¹² and Wolthayer et al used the Bayley Scales of Infant Development.¹³ Of the eight studies, half of them – Gurkha et al, Baron et al, Nepomnyaschy et al and Morag et al – found that the cognitive scores of late-preterm infants were comparable with those of term infants.^5,6,10,14 Morag et al reported that late-preterm infants ‘caught up’ with term controls at 12 months of corrected age and scored less than term controls at a chronological age of 12 months. Four studies – Petrini et al, Talge et al, Wolthayer et al and Morse et al – reported that late-preterm infants were more likely to show lower cognitive scores, compared with term infants.^11–13,16 Talge et al reported that, at the age of six years, late-preterm infants were twice as likely to have IQ scores of below 85, compared with term controls.
 

Late-preterm birth and motor delay

Some of the studies that reported cognitive scores also reported motor achievement scores. Nepomnyaschy et al reported that the risk of scoring >1 standard deviations (SD) below the mean for gross motor ability was comparable for late-preterm and full-term (>39 weeks of gestation) infants at two and four years of age.⁶ Morag et al evaluated motor scores using the Alberta Infant Motor scale at six months and the Griffiths test at 12 months. The authors reported that late-preterm infants showed a mean (± SD) locomotor score of 90 (± 11), significantly lower than the full-term infant score of 97 (± 12).¹⁴ However, late-preterm infants caught up with full-term infants at 12 months of corrected age (Morag et al). Wolthayer et al reported that the adjusted odds ratio (AOR) for psychomotor development index score of <70 (95% CI) was 1.56 (1.29, 1.88) at 24 months of age using Bailey’s short test version. The risk of low scores remained even after adjusting for some clinically important risk factors that included fetal growth and mode of delivery.¹³

Late-preterm birth and cerebral palsy

Two studies (n = 48,273 late-preterm infants and 1,096,283 full-term controls) reported on the link between late-preterm birth and cerebral palsy. In a Finnish study, Hirvonen et al reported that late-preterm infants had a higher risk of cerebral palsy diagnosis, with an odds ratio (OR) of 2.35 (95% confidence interval [CI]: 1.99, 2.77), compared with full-term infants by the age of seven years.¹⁷ Factors predictive of an increased risk of cerebral palsy in late-preterm groups included resuscitation at birth (OR 1.78; 95% CI: 1.09, 2.90), antibiotic treatment during the first hospitalisation (1.67; 95% CI: 1.13, 2.44), low one-minute Apgar score (1.80; 95% CI: 1.21, 2.67) and intracranial haemorrhage (12.8; 95% CI: 5.58, 29.2).¹⁷ By the age of seven years, 0.6% late-preterm infants and 0.1% full-term infants were diagnosed with cerebral palsy.¹⁷ Petrini et al evaluated the risk of cerebral palsy by five years of age.¹¹ Children born late preterm were more than three times as likely to be diagnosed with cerebral palsy when compared with full-term infants, with a hazard ratio of 3.39 (95% CI: 2.54, 4.52). However, results were not controlled for various clinical parameters, and the exact ages at diagnosis of cerebral palsy were variable.¹¹

Late-preterm birth and attention deficit hyperactivity disorder

Four studies (n = 60,328 late-preterm infants versus 101,638 full-term controls) reported on attention deficit hyperactivity disorder (ADHD; Table 3).^7,18–20 The studies used various methods for the diagnosis of ADHD. Talge et al used the Connors parent and teacher questionnaire,¹⁸ Harris et al used diagnostic and statistical manual IV criteria,¹⁹ Lindstrom used the need for stimulant medication as the basis for diagnosis,²⁰ while Rabie et al used International Classification of Diseases, ninth revision codes for the diagnosis.⁷ Rabie et al studied a subgroup of late-preterm infants who were induced because of a medical reason and termed them ‘medically indicated late-preterm infants’. They reported that medically indicated late-preterm patients had a higher risk of hyperactivity and high global index scores for ADHD. When all late-preterm infants were combined, the authors did not observe any increase in the risk of ADHD symptoms.⁷ Talge et al reported that after adjustment for parity, sociodemographics, child age, and maternal symptoms of depression and serious mental illness during pregnancy and at the child survey, only medically indicated late-preterm was associated with higher hyperactivity and global index scores (mean difference of 3.8 [95% CI: 0.5, 7.0] and 3.1 [95% CI: 0.0, 6.2]).¹⁸ The rate of ADHD also varied among the cohorts.
 

Late-preterm birth and autism

Only one study, by Guy et al, reported on the risk of autism spectrum disorder (ASD) in late-preterm infants at two years of corrected age using the modified checklist for autism questionnaire. In their population-based prospective cohort study of 548 late-preterm infants and 761 term-born infants, the authors reported that a total of 14.5% late-preterm infants versus 9.3% term controls scored above the clinical cut-off for ASD at two years of age.²¹ The risk of true positive failure rate was 2.5% versus 0.5% for late-preterm versus full-term infants on follow-up interview. However, age at follow-up testing was variable. Moreover, information on the neonatal complications were not available and were not controlled for. On further analysis, late-preterm infants with sensory issues, developmental delay, and speech and language issues were more likely to receive a high score on screening tests for ASD.

Discussion

This literature review identifies that at various stages of early childhood, some late-preterm infants are at an increased risk of developmental abnormalities, compared with term infants.

The literature reviewed suggest that most late-preterm infants had typical neurodevelopment comparable with term infants. Some late-preterm infants showed ‘catch up’ growth with gross motor,¹⁴ speech milestones,⁶ cognitive scores^12,13 by early childhood. It appears that a subset of late-preterm infants have persistent problems. Although studies have tried to identify the truly ‘at-risk’ late-preterm infant, more targeted research is needed to identify this subset of infants.

The results of this review need to be interpreted with caution. It is possible that broadening the search criteria might have resulted in the identification of a greater number of eligible studies. Studies that reported on the cognitive assessment results may have also reported on speech, motor and behavioural outcomes that were not included in the review. Therefore, selection bias could not be eliminated.

Although many studies in this review were grouped under one clinical outcome, significant clinical heterogeneity between studies was observed. Ages at speech, motor and IQ assessments, as well as diagnosis of cerebral palsy, varied between each study. Some studies performed clinical assessments at chronological age while some studies used corrected age. Only some studies provided the information on other confounding variables such as Apgar scores at birth, growth status of the infant, maternal medical conditions, and need for admission to the neonatal intensive care unit. Each of these confounders can potentially affect the neurodevelopmental outcome of the infant. Therefore, the results of the review lack generalisability. Significant variation in the selection of comparison group was also observed across studies. This varied from full term (>39 weeks of gestation) to term (>37 weeks of gestation), thereby affecting the results of assessments.^5,21 The quality of each study was not evaluated. Therefore, one needs to be careful when drawing strong conclusions from this review.

Nevertheless, the review highlights that late-preterm infants are at increased risk of various developmental abnormalities and would benefit from ongoing developmental surveillance. National Institute for Health and Care Excellence guidelines recommend that late-preterm infants with risk factors such as moderate or severe hypoxic ischaemic encephalopathy, bacterial meningitis, herpes encephalitis or brain lesions identified on magnetic resonance imaging be followed up in the first two years of life.³ For such patients, a minimum of two clinic visits in the first year between three and five months and at 12 months of corrected age and a developmental assessment at 24 months of corrected age are recommended.³ Since there is no endpoint to the long-term follow-up of late-preterm infants, follow-up should continue with varying degrees of surveillance reflecting individual patient needs beyond the first two to three years.

The economic costs associated with late-preterm birth are significant. Using data from the National Health Service, UK, Khan et al reported on the cost of late-preterm births over the first two years of life.²² The authors estimated that the mean societal cost of late-preterm infants was £5823 (AU$10,542), compared with £2056 (AU$3722) for children born at term. In view of medical complications as well as economic costs of late-preterm births, every attempt should be made to minimise delivery at late-preterm gestation. However, minimising the number of infants delivered in the late-preterm gestation may not be straightforward. Maternal conditions, such as gestational diabetes and preeclampsia, may prompt preterm induction of labour to minimise maternal and fetal complications. Fetuses with congenital anomalies or growth restriction may be spontaneously delivered early.²³ Such fetuses, irrespective of the time of gestation at birth, are more likely to have morbidities in infancy and childhood. While efforts to minimise late-preterm birth are ongoing, providing care to this large group of infants is paramount.

Conclusion

This review identifies that late-preterm infants are at an increased risk of developmental abnormalities in early childhood, compared with term infants. The review also identifies that the results are not consistently observed, and some late-preterm infants ‘catch up’ with speech, motor and cognitive outcomes. Some late-preterm infants show persistence of neurodevelopmental problems in early childhood. More research is needed in this complex area of paediatrics to understand the underlying risk factors for neurodevelopmental impairment in late-preterm infants.²⁴

GPs should be more vigilant about the increased risk of neurodevelopmental complications in late-preterm infants. While it is desirable to follow up all late-preterm infants for developmental surveillance, this is not feasible. Therefore, a targeted approach is needed. When infants undergo developmental assessment in the first 2–2.5 years, corrected age (ie age from the due date) should be used for all preterm infants born at <37 weeks of gestation.²⁵ At least two visits in the first year of life and a review at two years to discuss development are necessary for at-risk late-preterm infants.³
 

Appendix 2. Literature search methodology 

Methods

Study question: Are there differences in long-term neurodevelopmental outcomes among late-preterm infants compared with term infants (born at ≥37 weeks of gestation?

Criteria for considering studies for this review: Randomised controlled studies, observational cohort studies, cross sectional studies and case control studies that explored the association between late-preterm infants and developmental outcomes were included in this review. Studies published as abstracts, reports from national or local vital statistics not published as peer-reviewed articles were excluded.

Assessment of exposure: Studies that ascertained gestation from parent reports, dating ultrasound, hospital birth certificates, discharge summaries, vital statistics data and national databases were included.

Types of outcome measures: Studies reporting data on speech delay, gross motor delay, cerebral palsy, cognitive delay, autism spectrum disorder, attention deficit hyperactivity disorder were included.

Search strategy for identification of studies: Electronic databases (Cochrane library, Medline, Ovid) were searched on 4 February 2018 and 10 June for published studies in English language. MeSH words ‘late preterm’ OR ‘premature’ AND ‘developmental outcome’ OR ‘outcome’, ‘early intervention’, ‘cerebral palsy’, ‘gross motor delay’, ‘speech delay’, ‘autism’, ‘inattention’, ‘hyperactivity’, ‘ADHD’, alone and in different combinations were used. Alternative terminologies ‘neonate’, ‘infant’, ‘newborn’ were used to identify additional articles. The reference lists of the identified articles were searched for additional eligible studies. The articles were meticulously scanned on the basis of titles and abstracts. Selected articles were retrieved in full and were assessed for eligibility.

Only studies that reported the outcome of interest in late-preterm infants were included. If the study population included gestational age <32 weeks, or a group of 32–36 weeks, or no term comparison group, they were excluded. Given the nature of existing publications, statistical and clinical heterogeneity was expected. Hence descriptive review was planned.

Results

The results of study log is presented in Figure 1. The literature search did not identify any randomised controlled trials. Since this is not intended to be a systematic review, quality assessment or meta-analysis of the studies were not carried out. Excluded studies and the reason for their exclusion are presented as Appendix 1 and clinical characteristics of included studies are summarised in Table 1. Only relevant results for late-preterm category are presented.