Sex- and Age-Specific Deviations in Cerebellar Structure and Their Link With Symptom Dimensions and Clinical Outcome in Individuals at Clinical High Risk for Psychosis

Abstract

Background: The clinical high-risk (CHR) period offers a temporal window into neurobiological deviations preceding psychosis onset, but little attention has been given to regions outside the cerebrum in large-scale studies of CHR. Recently, the North American Prodrome Longitudinal Study (NAPLS)-2 revealed altered functional connectivity of the cerebello-thalamo-cortical circuitry among individuals at CHR; however, cerebellar morphology remains underinvestigated in this at-risk population, despite growing evidence of its involvement in psychosis.

Study design: In this multisite study, we analyzed T1-weighted magnetic resonance imaging scans obtained from N = 469 CHR individuals (61% male, ages = 12-36 years) and N = 212 healthy controls (52% male, ages = 12-34 years) from NAPLS-2, with a focus on cerebellar cortex and white matter volumes separately. Symptoms were rated by the Structured Interview for Psychosis-Risk Syndromes (SIPS). The outcome by two-year follow-up was categorized as in-remission, symptomatic, prodromal-progression, or psychotic. General linear models were used for case-control comparisons and tests for volumetric associations with baseline SIPS ratings and clinical outcomes.

Study results: Cerebellar cortex and white matter volumes differed between the CHR and healthy control groups at baseline, with sex moderating the difference in cortical volumes, and both sex and age moderating the difference in white matter volumes. Baseline ratings for major psychosis-risk dimensions as well as a clinical outcome at follow-up had tissue-specific associations with cerebellar volumes.

Conclusions: These findings point to clinically relevant deviations in cerebellar cortex and white matter structures among CHR individuals and highlight the importance of considering the complex interplay between sex and age when studying the neuromaturational substrates of psychosis risk.

Keywords: CHR; North American Prodrome Longitudinal Study; prodrome; schizophrenia; structural magnetic resonance imaging; ultra-high risk.

Fig. 1.

(A) Histogram showing age distributions of study participants. (B) Example cerebellar segmentation mask. (C) and (D) Cerebellar cortex and WM volume distributions across age. Solid lines: line of linear fit, error bands: 95% confidence interval. CHR, clinical high-risk; HC, healthy control; WM, white matter.

Fig. 2.

(A)–(C) Predictor effect plots illustrating sex- and age-specific differences in baseline cerebellar cortex and WM volumes between CHR and HC participants. Error bands: 95% confidence interval, dashed circles: crossover point of interaction. CHR, clinical high-risk; HC, healthy control; WM, white matter.

Fig. 3.

Predictor effect plots illustrating sex-specific associations between baseline cerebellar cortex volumes and symptom ratings among CHR participants. Error bands: 95% confidence interval. CHR, clinical high-risk.

Fig. 4.

Predictor effect plots illustrating sex- and age-specific associations between baseline cerebellar WM volumes and future clinical outcomes among CHR participants. Error bands: 95% confidence interval. CHR, clinical high-risk; WM, white matter.