Remapping the cognitive and neural profiles of children who struggle at school

Abstract

Our understanding of learning difficulties largely comes from children with specific diagnoses or individuals selected from community/clinical samples according to strict inclusion criteria. Applying strict exclusionary criteria overemphasizes within group homogeneity and between group differences, and fails to capture comorbidity. Here, we identify cognitive profiles in a large heterogeneous sample of struggling learners, using unsupervised machine learning in the form of an artificial neural network. Children were referred to the Centre for Attention Learning and Memory (CALM) by health and education professionals, irrespective of diagnosis or comorbidity, for problems in attention, memory, language, or poor school progress (n = 530). Children completed a battery of cognitive and learning assessments, underwent a structural MRI scan, and their parents completed behavior questionnaires. Within the network we could identify four groups of children: (a) children with broad cognitive difficulties, and severe reading, spelling and maths problems; (b) children with age-typical cognitive abilities and learning profiles; (c) children with working memory problems; and (d) children with phonological difficulties. Despite their contrasting cognitive profiles, the learning profiles for the latter two groups did not differ: both were around 1 SD below age-expected levels on all learning measures. Importantly a child's cognitive profile was not predicted by diagnosis or referral reason. We also constructed whole-brain structural connectomes for children from these four groupings (n = 184), alongside an additional group of typically developing children (n = 36), and identified distinct patterns of brain organization for each group. This study represents a novel move toward identifying data-driven neurocognitive dimensions underlying learning-related difficulties in a representative sample of poor learners.

Beyond the Core-Deficit Hypothesis in Developmental Disorders

“…Emerging first in adult psychiatry (Cuthbert & Insel, 2013; Morris & Cuthbert, 2012), transdiagnostic approaches focus on identifying underlying symptom dimensions that likely span multiple supposed categories (e.g., Astle, Bathelt, The CALM Team, & Holmes, 2019; Bathelt, Holmes, et al, 2018; Bryant, Guy, The CALM Team, & Holmes, 2020; Hawkins, Gathercole, Astle, The CALM Team, & Holmes, 2016; Holmes et al, 2019; Siugzdaite et al, 2020). Within developmental science there are good examples of researchers cutting across hitherto unquestioned diagnostic boundaries in order to identify cognitive symptoms that underpin learning, but they remain relatively rare (e.g., Astle et al, 2019; Casey, Oliveri, & Insel, 2014; Hulme & Snowling, 2009; Peng & Fuchs, 2016; Sonuga-Barke & Coghill, 2014; Zhao & Castellanos, 2016). A review of a transdiagnostic approach is well beyond the scope of the current article, but suffice it to say, contemporary developmental science needs larger and more diverse samples.…”

“…But relative to SEM-based approaches and network analysis, machine-learning applications remain underdeveloped. Although popular in other areas of science with similar challenges, these methods have yet to gain much traction within the study of developmental differences (but see Astle et al, 2019). These algorithms are highly flexible, and the resulting models can easily accommodate nonlinear relationships, make predictions about unseen data, be combined with simulations, incorporate different data types, and open the way to tools for testing generalization, such as cross-validation.…”

“…The left-most panel shows the best-matching unit for all children, and the subsequent panels show those for children with different diagnoses. The training data set included measures of fluid reasoning, vocabulary, verbal and spatial short-term and working memory, and phonological awareness (see also Astle, Bathelt, The CALM Team, & Holmes, 2019). ADHD = attention-deficit/hyperactivity disorder; ASD = autism spectrum disorder.…”

Toward a Science of Effective Cognitive Training

“…And yet these individual differences are often overlooked, and current approaches broadly take univariate statistical approaches, which are unable to identify individual cognitive profiles of performance on the basis of rich multivariate data. Contemporary multivariate analysis methods offer a radical rethink of training and associated transfer (Astle, Bathelt, The CALM Team, & Holmes, 2019; Bathelt, Holmes, & Astle, on behalf of the CALM Team, 2018) by focusing on training-related changes in task relationships.…”

The dyslexia debate: life without the label

“…The search for a unique causal factor in reading disability/dyslexia has largely been replaced by recognition that the problem is multifactorial, involving a range of risk and protective factors at genetic, neural, cognitive and environmental levels, with the weighting of each of these varying across individuals (Pennington, 2006;Pennington et al, 2012;Vandermosten et al, 2016). Continuing the current practice of diagnosing dyslexia on the basis of the presence of one or more of a long list of cognitive and behavioural symptoms, common to most poor readers, is no longer appropriate and may explain why Astle et al (2019) found little relationship between diagnoses of 530 children struggling at school and their cognitive profiles.…”