Objective: Very preterm born children are at risk for impairments in multiple neurodevelopmental domains, but outcomes vary between individuals. The present study aimed to distinguish subgroups with distinct profiles of functioning across motor, cognitive, behavioral, and psychosocial domains. These profiles were related to neonatal and social/environmental factors.
Method: The sample included 1977 children born very preterm (<32 weeks' gestation) in 2011 from the French population-based EPIPAGE-2 cohort. Using latent profile analysis, subgroups of children were distinguished based on their functioning at 5.5 years. The relation between outcome profiles and neonatal and social/environmental factors was tested using multivariable multinomial logistic regression analysis.
Results: Four subgroups with distinct outcome profiles were distinguished: no deficit in any domain (45%); motor and cognitive deficits without behavioral/psychosocial deficits (31%); primarily behavioral and psychosocial deficits (16%); and deficits in multiple domains (8%). Male sex (odds ratio [OR] = 2.1-2.7), bronchopulmonary dysplasia (OR = 2.1-2.8), low parental education level (OR = 1.8-2.1), and parental non-European immigrant status (OR = 2.3-3.0) were independently associated with higher odds for all suboptimal outcome profiles compared to the favorable outcome profile.
Conclusion: Among 5.5-year-old very preterm born children, subgroups can be distinguished with distinct outcome profiles that vary in severity, type, and combinations of deficits. This information is important for the development of interventions that are tailored to the needs of large subgroups of children across multiple domains of functioning. General neonatal and social/environmental factors may be useful for early identification of very preterm born children at risk for general rather than domain-specific impairments.
Keywords: cognition; infant; latent class analysis; motor skills; neurodevelopmental disorders; premature.
Copyright © 2021 The Authors. Published by Elsevier Inc. All rights reserved.