Developing a residual cognitive reserve index in a paediatric cohort with ADHD

Abstract

AbstractBackgroundCognitive reserve (CR) is thought to be protective against late‐life cognitive decline induced by neurodegenerative disease and other forms of neurological insult. Although CR is thought to be influenced by early life experiences, information about its development during childhood and adolescence is lacking. The current study operationalised CR with a residual index approach (Reed et al., 2010) optimised for a diverse paediatric cohort. Criterion and construct validity of the index was also assessed.MethodData collected across 2011‐2019 from N = 115 children (MAge= 10.48, SDAge= 0.61, n = 72 [63%] male, n = 43 [37%] with ADHD) were acquired from the Murdoch Children’s Research Institute. Elastic‐net regularised linear regression was used to generate baseline and longitudinal CR indices that represented residual variance in (change in) fluid intellect (Matrix Reasoning) not explained by demographics (i.e., sex and age) or MRI‐based brain variables (i.e., cortical surface areas and cerebellar/subcortical volumes); brain regions were selected with the aim of partialling maximal variance out of the CR residual index. Outcome variables, including word reading, math computation, and Conners 3 ADHD Index were then regressed onto baseline and longitudinal CR indices.ResultsBaseline CR predicted baseline math computation (p<.001), while change in CR predicted change in word reading ability (p<.001). Further, in individuals with low grey matter volume, higher CR was associated with higher word reading score (p = .019) and lower ADHD Index (p = .047) compared to those with lower CR, at baseline. Higher change in CR also protected word reading trajectory in individuals with higher change in WMH volumes (p = .048).ConclusionThese results suggest that individual differences in CR, as operationalised through a residual index, may be detectable in school‐age children. While findings regarding criterion and construct validity of the CR index are stronger than previous research featuring an unoptimised residual model (Irani, 2022), support for the validity of the index remains incomplete compared to established geriatric models – possibly due in part to a relatively less predictable relationship between brain structure and cognition in children versus adults. Future research would benefit from longer‐term follow‐up in paediatric samples with greater variability in neuropathological indices, cognitive functioning, academic achievement, and functional outcomes.