PhD Candidate The University of Queensland The University of Queensland Brisbane , Queensland, Australia
Background: Children born preterm have elevated risk for socioemotional and behavioral difficulties than term-born peers. These children tend to manifest unique patterns of comorbid difficulties characteristic of autism spectrum disorder (ASD), attention-deficit/hyperactivity disorder (ADHD), and anxiety (ANX), called the Preterm Behavioral Phenotype (PBP). Current studies are restricted by small samples, limited age-ranges, and exclusive focus on multi-domain composite scores.
Objective: 1) To characterize the PBP in children born preterm at domain- and subdomain-level using latent profile analysis; 2) identify the extent to which gestational age, chronological age, and sex are associated with domain-level profile membership.
Design/Methods: Using cross-sectional study design, 2416 children born preterm aged 3–18 years (47% preschool, 31% school-age, 16% early adolescence, 6% late adolescence; 53% male; 24% extremely, 32% very, 44% late preterm) were recruited through parent organizations. Norm-referenced, parent-reported standardized measures assessed child psychopathology: Social Responsiveness Scale 2nd Edition, Strengths and Weaknesses of ADHD Symptoms and Normal Behavior Rating Scale, and Preschool Anxiety Scale/Screen for Child Anxiety Related Emotional Disorders.
Results: As seen in Figure 1, domain-level latent profile analysis revealed three distinct groups—Profile 1: PBP-low (61%) where children scored below the mean on all outcomes, Profile 2: PBP-average (29%) where children had higher than average scores on outcomes, and Profile 3: PBP-high (10%) where children scored greater than the mean on outcomes indicating more severe symptomatology. The three PBP domains were at similar severity levels within each profile, with the most marked differences for ASD. As seen in Figure 2, a comparable 3-profile categorization was found for subdomains: PBP-low (54%), PBP-average (33%), and PBP-high (13%). Relative risk (RR) of children in the PBP-high group compared to PBP-low was higher for school-aged (RR 2.10, 95% CI 1.47–3.0), males (RR 1.61, 95% CI 1.17–2.21) born extremely preterm (RR 2.35, 95% CI 1.64–3.37), with a similar risk profile evident for the PBP-average group. Conclusion(s): Findings confirm the existence of the unique PBP in the largest sample to date, with severe symptomatology restricted to a small proportion of children born preterm. Further, findings demonstrate similar profiles at the domain- and subdomain-level, supporting gestational age, chronological age, and male sex as risk factors for the PBP.
Figure 1. Latent Profile Analysis of ASD, ADHD, and ANX Domains.
Figure 2. Latent Profile Analysis of ASD, ADHD, and ANX Subdomains.
Authors/Institutions: Grace C. Fitzallen, The University of Queensland, Saint Lucia, Queensland, Australia; Alison Griffin, QIMR Berghofer Medical Research Institute, Herston, Queensland, Australia; Hudson G. Taylor, Nationwide Children's Hospital, Columbus, Ohio, United States; James N. Kirby, The University of Queensland, Saint Lucia, Queensland, Australia; Helen G. Liley, Mater Mothers' Hospital, South Brisbane, Queensland, Australia; Samudragupta Bora, Mater Research Institute, Faculty of Medicine, The University of Queensland, South Brisbane, Queensland, Australia