Article Text

Download PDFPDF

RF18 Contribution of severe neonatal morbidity to neurodevelopment at 2 years of age among very preterm infants: a mediation analysis
  1. C Bonnet1,
  2. ES Draper2,
  3. RF Maier3,
  4. M Cuttini4,
  5. L Herich4,
  6. SJ Johnson2,
  7. J Zeitlin1
  1. 1Obstetrical, Perinatal and Pediatric Epidemiology Research Team, Inserm, Paris, France
  2. 2Department of Health Sciences, Leicester University, Leicester, UK
  3. 3Children’s Hospital, Marburg Univeristy, Marburg, Germany
  4. 4Clinical Care and Management Innovation Research Area, Bambino Gesù Children’s Hospital, Rome, Italy


Background Low gestational age (GA) is the principle risk factor for poor long-term neurodevelopment among very preterm (VPT) children. Some of this effect results from a greater vulnerability to severe neonatal morbidities (SNM) which are strongly related to low GA as well as to poor neurodevelopmental outcome. We aimed to investigate how SNM mediates the relationship between GA and VPT adverse neurodevelopmental outcome at 2 years of corrected age.

Methods Data come from the area-based Effective Perinatal Intensive Care in Europe (EPICE) cohort of children born below 32 weeks’ GA in 2011/2012 in 11 European countries. Perinatal data were abstracted from medical records and follow-up was conducted using parent questionnaires at 2 years of corrected age. Children unable to walk without assistance or aids, or sit or hold their head up without support were classified with severe gross motor impairment (GMI); Non-verbal cognitive impairment (NVCI) was assessed as a score <22 on the non-verbal cognition scale of the Parent Report of Children’s Abilities-Revised (PARCA-R) and language impairment (LI) was defined as an expressive vocabulary of less than 10 words. SNM was defined as one or more of the following morbidities: intraventricular hemorrhage grades III & IV, cystic periventricular leukomalacia, retinopathy of prematurity stages III-V, surgical necrotizing enterocolitis or broncho-pulmonary dysplasia. An inverse odds weighting (IOW) mediation analysis was used to estimate the total effects of GA, and estimated indirect effects operating through SNM or directly through GA. CI were estimated by bootstrapping. Co-variables included perinatal risk factors and family socioeconomic characteristics.

Results Of 3370 children assessed in the follow-up; 5.3% had GMI, 14.9% had NVCI and 9.4% had LI. A one-week increase in GA was associated with a decreased risk of GMI (OR=0.80; 95 CI%=0.74–0.86]), of NVCI (OR=0.85; 95 CI%=0.81–0.89]) and LI (OR=0.89; 95 CI%=0.83–0.96]). 20.4% of infants had at least one severe morbidity and this was more common among children with GMI (66.0% vs. 18.4%), NVCI (38.8% vs. 17.2%) and LI (38.4% vs. 18.4%). For GMI, almost 100% of the total effect of GA on GMI was mediated by SNM. This proportion was 70% for NVCI and 40% for LI.

Conclusion Reducing SNM could lead to substantial improvements in neurodevelopmental outcomes for VPT children, especially those related to motor development.

  • preterm birth
  • neurodevelopmental consequences
  • mediation analyses

Statistics from

Request Permissions

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.