MANUSCRIPT CITATION

Engan M, Engeseth MS, Fevang S, Vollsæter M, Eide GE, Røksund OD, Halvorsen T, Clemm H. Predicting physical activity in a national cohort of children born extremely preterm. Early Hum Dev. 2020; 145: 105037. PMID 32438296

REVIEWED BY

Kate L Cameron, DPT
Department of Physiotherapy, The University of Melbourne
Clinical Sciences, Murdoch Children’s Research Institute

Prof Jeanie LY Cheong, MD
Clinical Sciences, Murdoch Children’s Research Institute
Newborn Research, Royal Women’s Hospital
Obstetrics and Gynaecology, The University of Melbourne

Prof Alicia J Spittle, PhD
Department of Physiotherapy, The University of Melbourne
Clinical Sciences, Murdoch Children’s Research Institute
Newborn Research, Royal Women’s Hospital

TYPE OF INVESTIGATION

Prognosis

QUESTION

1. Is there a difference in physical activity (PA) outcomes at 11 years between children born extremely preterm and/or extremely low birthweight (EP/ELBW, ≤28 weeks’ gestation and/or <1000g) and children born at term?
2. Can motor impairment, borderline cognitive function, or behavioural problems at 5 years predict PA outcomes at 11 years for children born EP/ELBW without a neurodevelopmental disability?

METHODS

• Design: Prospective geographical cohort study (EP/ELBW children) with a term control group recruited at 11 years.
• Allocation: N/A
• Blinding: Only children born EP/ELBW participated in 5-year assessments, and assessors were not blinded to perinatal data. 11-year data was obtained using a parent-report questionnaire.
• Follow-up period: Children born EP/ELBW were assessed at 5 years. Parents of children born EP/ELBW and term completed a questionnaire when children were 11 years of age.
• Setting: Norway
• Patients: All infants born EP/ELBW in Norway during 1999-2000 were invited to participate during the neonatal period. The term group was recruited at 11 years of age as a control group for a sub-group of EP/ELBW children (those born in the Western Norway Regional Health Authority). Term-born children were recruited by finding next child born >37 weeks’ and >3000g after each child in the EP/ELBW subgroup, matching to hospital of birth and sex.
• Intervention: N/A
• Outcomes
  • Question 1: Physical activity (Children born EP/ELBW and term)
    • PA at 11 years was assessed using a newly developed, purpose built, parent-report questionnaire, with the exception of one question, which was from the World Health Organisation Health Behaviour in School Children Survey.
  • Question 2: Physical activity predictors (Children born EP/ELBW only)
    • Three possible predictors for physical activity at 11 years were obtained from the results of a neurodevelopmental assessment at 5 years including motor impairment borderline cognitive function and behavioural problems. Children who scored ≤5th percentile on the Movement Assessment Battery for Children-2nd edition (MABC-2) were classified as having motor impairment. Cognitive function was assessed using the Wechsler Preschool and Primary Scale of Intelligence-revised (WPPSI-R) full scale intelligence quotient (FSIQ). Children were considered to have borderline intellectual functioning if they scored between 70 and 84 on the FSIQ. Behaviour was assessed using the parent-report Strengths and Difficulties Questionnaire (SDQ), with a total difficulties score of ≥90th percentile indicating behavioural problems. Children in the EP/ELBW group were considered to have NDD if they had a diagnosis of cerebral palsy, a FSIQ score of <70 or severe visual or hearing impairment.
• Analysis
  • For Question 1, chi squared exact tests and the Mann-Whitney U test were used to compare PA outcomes between a) all children born EP/ELBW and term and b) children born EP/ELBW with and without NDD, adjusted for socioeconomic status.
  • For Question 2, logistic regressions were used to identify if motor impairment, borderline IQ or behavioural problems were associated with PA outcomes at 11 years for EP/ELBW without NDD only. Regressions were adjusted based on significant differences present between groups. In addition, regressions predicting sports proficiency, manual dexterity and gross motor were adjusted for motor impairment (borderlines IQ as a predictor) and borderline IQ (motor impairment and behavioural problems as predictors).  This decision was based on results from a previous study by Smits-Engelsman and Hill (2012).
  • Sample size was not calculated for this study as the number of EP/ELBW children was determined by the size of the previously recruited cohort study. They did not include an estimate of differences that could be determined by the sample size available for the study.

MAIN RESULTS

Of the 372 EP/ELBW children initially recruited, data on physical activity predictors (5 year assessment) was collected from approximately half of EP/ELBW cohort (MABC-2 n=185 (50%), SDQ n=174 (47%), WPPSI-R n=176 (47%)). At 11 years, PA questionnaires were completed for 62% (n=232/372) of children born EP/ELBW and 93% (n=57/61) of children born at term. Of the children born EP/ELBW who completed the PA questionnaire, 23 were considered to have NDD. One child was not included in the analysis as insufficient information was available to determine disability status

Differences in PA between children born EP/ELBW and term

PA outcomes at 11 years for each of the three groups, EP/ELBW with and without NDD and term-born children, are presented in Table 1. Overall, children born EP/ELBW with NDD had poorer PA outcomes than children born EP/ELBW without NDD, while children born term had more favourable outcomes than all children born EP/ELBW. When adjusted for socioeconomic factors, between groups differences between a) children born EP/ELBW with and without NDD and b) children born EP/ELBW and term, remained significant. The only exceptions were sports club activities and other organised activities, which were similar across all groups.

Table 1: Children with unfavourable physical activity results by group at 11 years

Predictors of PA in children born EP/ELBW

Predictors of PA outcomes at 11 years for children born EP/ELBW without NDD are presented in Table 2. Children with motor impairment at 5 years were more likely to play less vigorously and have poorer sports proficiency at 11 years than those without motor impairment. Borderline IQ at 5 years was associated with poor endurance, less vigorous play and clumsy manual dexterity at 11 years, while behavioural problems at 5 were associated with poor endurance, less vigorous play and poor sports proficiency at 11 years for children born EP/ELBW without NDD.

Table 2: Prediction of unfavourable PA outcomes at 11 years for children born EP/ELBW who did not have a neurodevelopmental disability using 3 predictor variables: motor impairment, borderline IQ and behavioural problems. Results are adjusted for confounding variables as outlined in methods, analysis

CONCLUSION

Children who were born EP/ELBW, both with and without NDD, had worse PA outcomes at 11 years compared with children born at term. For children born EP/ELBW who did not have NDD; motor impairment, borderline cognitive difficulty and behavioural problems at 5 years were all predictive of unfavourable PA outcomes at 11 years.

COMMENTARY

With increased survival of children born EP/ELBW it is important to  understand long-term health and development,(1) with many international cohort studies reporting poorer neurodevelopmental outcomes across multiple domains for children born EP/ELBW compared with term.(2, 3) Less is known, however, about functional outcomes, including PA participation, for children born EP/ELBW.

Using a purpose-built questionnaire at 11 years of age, Engan and colleagues found children born at term had better PA outcomes compared with children born EP/ELBW. Within the EP/ELBW group, those without NDD had better PA outcomes. Motor, cognitive and behavioural difficulties at 5 years were considered as possible predictors of PA for children born EP/ELBW, with all three predictors associated with poorer PA outcomes at 11 years.(4)

The authors specifically considered children born EP/ELBW without NDD, choosing to focus on ‘more subtle problems’, that are often overlooked but can have significant implications.

Previous cohort studies have measured PA related outcomes, including exercise capacity and PA minutes for children born preterm compared with term; with variable findings including; modest differences in PA time (5) and no difference in PA time but reduced exercise capacity.(6) As acknowledged by Engan and colleagues, differences in methodology and cohort characteristics make it difficult to make comparisons.(4)

The authors identified follow-up rates as a study limitation; the children without PA data from the EP/ELBW group were more likely to have NDD than children with data, which may have decreased the true magnitude of group differences.(4) A strength of this study was the consideration of PA as an outcome measure, focusing on what children are doing in a real-world context rather than how they perform under test conditions. Children born EP/ELBW without NDD have greater rates of developmental difficulties compared with children born at term, and this has implications for PA participation.

During childhood, PA participation promotes motor skill development and provides opportunity for social and emotional development. PA has physical (cardiovascular fitness, muscle strength etc.) and mental health benefits,(7) and may improve a range of health and developmental difficulties for which individuals born EP/ELBW are at greater risk than those born at term. (2, 8)  This research is valuable in that it looks beyond disability or impairment in children born EP/ELBW, and towards PA participation, a construct more closely aligned with meaningful quality of life. This study contributes towards emerging research considering PA outcomes for children born EP/ELBW, as well as predictors of poor PA participation, essential for providing timely intervention. This is important in the EP/ELBW population with known risk factors for poor PA participation, such as motor impairment.(2) Clinicians working with children born EP/ELBW should consider assessing PA, and work with families to promote meaningful participation in activities preferred by the child. Further research on PA, using large prospective cohort studies with EP/ELBW and term children, are needed.(1) Where possible, objective PA measures (such as accelerometery) and questionnaires with established psychometric properties should be used; though it is acknowledged that there is no consensus on the best way to measure PA participation.

REFERENCES

1. Doyle LW, Anderson PJ, Battin M, Bowen JR, Brown N, Callanan C, et al. Long term follow up of high risk children: who, why and how? BMC Pediatr. 2014;14(1):1.
2. Spittle AJ, Cameron K, Doyle LW, Cheong JL. Motor Impairment Trends in Extremely Preterm Children: 1991-2005. Pediatrics. 2018.
3. Moore T, Hennessy EM, Myles J, Johnson, SJ, Draper ES, Costeloe, KL et al. Neurological and developmental outcome in extremely preterm children born in England in 1995 and 2006: the EPICure studies. BMJ 2012; 345: e7961: e10.1136/bmj.e7961.
4. Engan M, Engeseth MS, Fevang S, Vollsæter M, Eide GE, Røksund OD, et al. Predicting physical activity in a national cohort of children born extremely preterm. Early Hum Dev. 2020;145:105037.
5. Lowe J, Watkins WJ, Kotecha SJ, Kotecha S. Physical Activity and Sedentary Behavior in Preterm-Born 7-Year Old Children. PLoS One. 2016;11(5):e0155229.
6. Welsh L, Kirkby J, Lum S, Odendaal D, Marlow N, Derrick G, et al. The EPICure study: maximal exercise and physical activity in school children born extremely preterm. Thorax. 2010;65(2):165-72.
7. Timmons BW, Naylor PJ, Pfeiffer KA. Physical activity for preschool children — how much and how? Canadian Journal of Public Health / Revue Canadienne de Sante’e Publique 2007; 98: S122.
8. Doyle LW, Irving L, Haikerwal A, Lee K, Ranganathan S, Cheong J. Airway obstruction in young adults born extremely preterm or extremely low birth weight in the postsurfactant era. Thorax. 2019;74(12):1147-53.