Back

The neurodevelopmental spectrum: phenotypic architecture, etiology, and predictive utility across development

Michelini, G.; Liao, W.; Lu, S.; Caserini, C.; Eley, T. C.; Ronald, A.; Wilson, S.; Malanchini, M.; Rimfeld, K.

2025-07-31 psychiatry and clinical psychology
10.1101/2025.07.31.25332492 medRxiv
Show abstract

Neurodevelopmental conditions are highly heritable, heterogeneous, and frequently co-occur with one another. Transdiagnostic dimensional approaches have advanced our understanding of psychiatric disorders and informed their classification, but have largely omitted neurodevelopmental conditions. We investigated the structure of a transdiagnostic "neurodevelopmental spectrum", its etiology, and its ability to predict functional outcomes across development, using longitudinal data for >10,000 twins from the Twins Early Development Study (TEDS). Hierarchical exploratory factor modeling of traits/symptoms from a broad questionnaire battery delineated a phenotypic neurodevelopmental spectrum alongside internalizing and externalizing dimensions at ages 7, 12, and 16. Twin, polygenic score, and longitudinal analyses showed that this neurodevelopmental spectrum was highly heritable across development (h2=0.60-0.82) and predicted by polygenic scores (PGS) for neurodevelopmental, cognitive and educational phenotypes (R2 up to 2.30% in single-PGS analyses, 3.36% in multi-PGS analyses) as well as by perinatal environmental and early developmental factors (e.g., low birth weight, language delays) (R2 up to 8.65%). Differences between children in this neurodevelopmental spectrum predicted cognitive and educational outcomes both concurrently and longitudinally (R2 up to 21%), largely due to overlapping genetic effects. Most results were unchanged when controlling for other transdiagnostic dimensions, indicating specificity of associations with the neurodevelopmental spectrum. Our new data on the phenotypic architecture, etiology, and predictive utility of the neurodevelopmental spectrum across development supports the integration of this spectrum into transdiagnostic frameworks, with important implications for advancing future research, psychiatric classification, and clinical care.

Published in Molecular Psychiatry (predicted rank #1) · training set

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.