Neuroimaging Summary Scores Predict Trajectories of Psychotic-Like Experiences in Youth
- 1Boston Children's Hospital
Abstract
Objective. Persistent, distressing psychotic-like experiences (PLEs) are associated with neurobiological alterations and increased psychosis risk. We combined individual-level neuroimaging measures with effect sizes from large neuroimaging studies to create a summary score ('Psychosis Neuroscore') reflecting neuroanatomic liability for psychosis, and examined its ability to predict PLE trajectories in young adolescents. Method. Using latent growth mixture models, we estimated PLE trajectories from four annual visits of the Adolescent Brain Cognitive Development Study (N=9584, ages 9-10 at baseline). Using baseline T1-weighted and diffusion-weighted imaging data, we calculated Psychosis Neuroscores, as well as Neuroscores for two psychiatric disorders with late adolescent/adult onset (Major Depressive Disorder, Bipolar Disorder). We compared Psychosis Neuroscores to i) other psychiatric Neuroscores, ii) modifiable risk factors, and iii) established risk factors in predicting trajectory membership. Results. We identified four trajectories of distressing PLEs: Persistent Elevated (N=1,968, 21%), Gradual Decreasing (N=3,424, 36%), Rapid Decreasing (N=1,593, 17%) and Low/No Distress (N=2,599, 27%). Adolescents with Persistent Elevated PLEs had significantly higher Multimodal (combined T1 and diffusion-weighted) and T1-weighted Psychosis Neuroscores than all other trajectories (Odds Ratios [ORs] 1.27-1.34,pFDR<.01). Bipolar Disorder Neuroscores showed a similar pattern (ORs 1.16-1.23,pFDR<.01). Psychosis Neuroscores showed comparable associations with established risk factors in predicting trajectory membership, but smaller associations than modifiable risk factors, including screen time, physical activity, and sleep disturbances. Conclusion. Psychosis Neuroscores differentiate youth with persistent PLEs from those with decreasing, remitting or low PLEs, demonstrating their potential utility for early risk stratification. Integration with established risk factors may enhance psychosis risk prediction in youth.
Related
- Multimodal Prediction of Distressing and Persistent Psychotic-Like Experiences in Adolescence across Environment, Genetics, and Neuroimaging
- A twin study of genetic and environmental aetiologies of associations between sensory processing sensitivity and psychotic-like experiences
- Environmental etiology of psychotic experiences in Ethiopia: Effects of khat use, trauma load and seasonality
- Longitudinal patterns of sensory and paranoid psychotic-like experiences in adolescents: Associations with distress, psychopathology, and substance use
- Multimodal neuroimaging signature of sleep problems captures preadolescent mental health trajectories
BibTeX
@misc{cooper2026neuroimaging,
title = {{Neuroimaging Summary Scores Predict Trajectories of Psychotic-Like Experiences in Youth}},
author = {R. E. Cooper and R. Sahasrabudhe and D. C. Glahn and M. Jalbrzikowski},
howpublished = {medRxiv},
year = {2026},
month = jun,
doi = {10.64898/2026.06.03.26354754},
url = {https://www.medrxiv.org/content/10.64898/2026.06.03.26354754v2},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - Neuroimaging Summary Scores Predict Trajectories of Psychotic-Like Experiences in Youth
AU - R. E. Cooper
AU - R. Sahasrabudhe
AU - D. C. Glahn
AU - M. Jalbrzikowski
PY - 2026
DA - 2026/06/03
PB - medRxiv
DO - 10.64898/2026.06.03.26354754
UR - https://www.medrxiv.org/content/10.64898/2026.06.03.26354754v2
AB - Objective. Persistent, distressing psychotic-like experiences (PLEs) are associated with neurobiological alterations and increased psychosis risk. We combined individual-level neuroimaging measures with effect sizes from large neuroimaging studies to create a summary score ('Psychosis Neuroscore') reflecting neuroanatomic liability for psychosis, and examined its ability to predict PLE trajectories in young adolescents. Method. Using latent growth mixture models, we estimated PLE trajectories from four annual visits of the Adolescent Brain Cognitive Development Study (N=9584, ages 9-10 at baseline). Using baseline T1-weighted and diffusion-weighted imaging data, we calculated Psychosis Neuroscores, as well as Neuroscores for two psychiatric disorders with late adolescent/adult onset (Major Depressive Disorder, Bipolar Disorder). We compared Psychosis Neuroscores to i) other psychiatric Neuroscores, ii) modifiable risk factors, and iii) established risk factors in predicting trajectory membership. Results. We identified four trajectories of distressing PLEs: Persistent Elevated (N=1,968, 21%), Gradual Decreasing (N=3,424, 36%), Rapid Decreasing (N=1,593, 17%) and Low/No Distress (N=2,599, 27%). Adolescents with Persistent Elevated PLEs had significantly higher Multimodal (combined T1 and diffusion-weighted) and T1-weighted Psychosis Neuroscores than all other trajectories (Odds Ratios [ORs] 1.27-1.34,pFDR<.01). Bipolar Disorder Neuroscores showed a similar pattern (ORs 1.16-1.23,pFDR<.01). Psychosis Neuroscores showed comparable associations with established risk factors in predicting trajectory membership, but smaller associations than modifiable risk factors, including screen time, physical activity, and sleep disturbances. Conclusion. Psychosis Neuroscores differentiate youth with persistent PLEs from those with decreasing, remitting or low PLEs, demonstrating their potential utility for early risk stratification. Integration with established risk factors may enhance psychosis risk prediction in youth.
ER -