Is association of preterm birth with cognitive-neurophysiological impairments and ADHD symptoms consistent with a causal inference or due to familial confounds?

Preterm birth is associated with an increased risk for cognitive-neurophysiological impairments and attention-deficit/hyperactivity disorder (ADHD). Whether the associations are due to the preterm birth insult per se, or due to other risk factors that characterise families with preterm-born children, is largely unknown.We employed a within-sibling comparison design, using cognitive-performance and event-related potential (ERP) measures from 104 preterm-born adolescents and 104 of their term-born siblings. Analyses focused on ADHD symptoms and cognitive and ERP measures from a cued continuous performance test, an arrow flanker task and a reaction time task. Within-sibling analyses showed that preterm birth was significantly associated with increased ADHD symptoms (β = 0.32, p = 0.01, 95% CI 0.05 to 0.58) and specific cognitive-ERP impairments, such as IQ (β = -0.20, p = 0.02, 95% CI -0.40 to -0.01), preparation-vigilance measures and measures of error processing (ranging from β = 0.71, -0.35). There was a negligible within-sibling association between preterm birth with executive control measures of inhibition (NoGo-P3, β = -0.07, p = 0.45, 95% CI -0.33 to 0.15) or verbal working memory (digit span backward, β = -0.05, p = 0.63, 95% CI -0.30 to 0.18).Our results suggest that the relationship between preterm birth with ADHD symptoms and specific cognitive-neurophysiological impairments (IQ, preparation-vigilance and error processing) is independent of family-level risk and consistent with a causal inference. In contrast, our results suggest that previously observed associations between preterm birth with executive control processes of inhibition and working memory are instead linked to background characteristics of families with a preterm-born child rather than preterm birth insult per se. These findings suggest that interventions need to target both preterm-birth specific and family-level risk factors.