SENTENAC et al_Maindocument_JECH R1_Clean Copy_final accepted for IRIS.pdf (367.64 kB)
Download file

Maternal education and language development at 2 years corrected age in children born very preterm: Results from a European population-based cohort study

Download (367.64 kB)
journal contribution
posted on 01.04.2020, 07:47 by Marianne Sentenac, Samantha Johnson, Marie-Laure Charkaluk, Anna-Veera Seppänen, Ulrika Aden, Marina Cutinni, Rolf Maier, Mairi Männamaa, Jennifer Zeitlin
Background Socioeconomic factors influence language development in the general population, but the association remains poorly documented in children born very preterm (VPT). We assessed the impact of maternal education on language development in children born VPT and effect modification by perinatal risk.

Methods Data were from the Effective Perinatal Intensive Care in Europe (EPICE) population-based cohort of children born <32 weeks’ gestational age (GA) in 2011/2012. Regions from six countries (Estonia, France, Germany, Italy, Sweden and UK) used a validated short form MacArthur Developmental Communicative Inventories Checklist to assess language at 2 years corrected age. Perinatal variables were collected from clinical records. We assessed expressive language delay (ELD), defined as (a) not combining words; and (b) expressive vocabulary <10th percentile of norms for age and sex. Perinatal risk (low, moderate and high) was determined using GA, small for GA and neonatal morbidities. We estimated adjusted risk ratios (aRR) of ELD by maternal education with inverse weighting to account for non-response bias.

Results Of 2741 children, 24.6% were not combining words and 39.7% had a low expressive vocabulary. Low maternal education (lower secondary or less compared with a bachelor’s degree or more) increased risks of ELD: not combining words: aRR=1.52 (95% CI 1.36 to 1.69); low expressive vocabulary: aRR=1.25 (1.04 to 1.51). For children with low perinatal risk, the aRR were 1.88 (1.26 to 2.80) and 1.44 (1.06 to 1.95), respectively, compared with those with high perinatal risks: 1.36 (1.10 to 1.67) and 1.11 (0.97 to 1.27), respectively.

Conclusion Low maternal education affects ELD for children born VPT, although the association appears attenuated among those with highest perinatal risk.



J Epidemiol Community Health 2020;74:346-353.

Author affiliation

Department of Health Sciences, University of Leicester


AM (Accepted Manuscript)

Published in

Journal of Epidemiology and Community Health






BMJ Publishing Group



Acceptance date


Copyright date


Available date


Publisher version