While adolescent-onset schizophrenia (ADO-SCZ) and adolescent-onset bipolar disorder with psychosis (psychotic ADO-BPD) present a more severe clinical course than their adult forms, their pathophysiology is poorly understood. Here, we study potentially state- and trait-related white matter diffusion-weighted magnetic resonance imaging (dMRI) abnormalities along the adolescent-onset psychosis continuum to address this need.

Forty-eight individuals with ADO-SCZ (20 female/28 male), 15 individuals with psychotic ADO-BPD (7 female/8 male), and 35 healthy controls (HCs, 18 female/17 male) underwent dMRI and clinical assessments. Maps of extracellular free-water (FW) and fractional anisotropy of cellular tissue (FAT) were compared between individuals with psychosis and HCs using tract-based spatial statistics and FSL's Randomise. FAT and FW values were extracted, averaged across all voxels that demonstrated group differences, and then utilized to test for the influence of age, medication, age of onset, duration of illness, symptom severity, and intelligence.

Individuals with adolescent-onset psychosis exhibited pronounced FW and FAT abnormalities compared to HCs. FAT reductions were spatially more widespread in ADO-SCZ. FW increases, however, were only present in psychotic ADO-BPD. In HCs, but not in individuals with adolescent-onset psychosis, FAT was positively related to age.

We observe evidence for cellular (FAT) and extracellular (FW) white matter abnormalities in adolescent-onset psychosis. Although cellular white matter abnormalities were more prominent in ADO-SCZ, such alterations may reflect a shared trait, i.e. neurodevelopmental pathology, present across the psychosis spectrum. Extracellular abnormalities were evident in psychotic ADO-BPD, potentially indicating a more dynamic, state-dependent brain reaction to psychosis.

There is considerable evidence that white matter abnormalities play a key role in the pathogenesis of a number of major psychiatric disorders, including schizophrenia, bipolar affective disorder, and obsessive-compulsive disorder. Few studies, however, have compared white matter abnormalities early in the course of the illness.

A total of 102 children and adolescents participated in the study, including 43 with early-onset schizophrenia, 13 with early-onset bipolar affective disorder, 17 with obsessive-compulsive disorder, and 29 healthy controls. Diffusion tensor imaging scans were obtained on all children and the images were assessed for the presence of non-spatially overlapping regions of white matter differences, a novel algorithm known as the pothole approach.

Patients with early-onset schizophrenia and early-onset bipolar affective disorder had a significantly greater number of white matter potholes compared to controls, but the total number of potholes did not differ between the two groups. The volumes of the potholes were significantly larger in patients with early-onset bipolar affective disorder compared to the early-onset schizophrenia group. Children and adolescents with obsessive-compulsive disorder showed no differences in the total number of white matter potholes compared to controls.

White matter abnormalities in early-onset schizophrenia and bipolar affective disorder are more global in nature, whereas children and adolescents with obsessive-compulsive disorder do not show widespread differences in FA.

Adolescent-onset schizophrenia (AOS) is associated with cognitive impairment and poor clinical outcome. Cognitive dysfunction is thought to reflect functional dysconnectivity between the frontal cortex and the striatum, Previous work [1] has shown frontostriatal dysconnectivity in large WM tracts explain core cognitive deficits, with processing speed, which is affected by alterations in WM connectivity, being an intermediary variable.

To undertake a follow-up MRI study using whole-brain structural connectomics to track topological changes in the follow-up (1st episode versus follow-up), in order to characterize the early stages (evolution of the first two years) of the disorder.

A follow-up study of 25 AOS subjects and 25 age and sex-matched healthy subjects.

Network theory will be applied to identify topological alterations in structural networks, including frontostriatal white matter (WM) tracts in relation to cognition and outcome measures.

The authors have not supplied their declaration of competing interest.

Adolescent-onset schizophrenia (AOS) is associated with cognitive impairment and poor clinical outcome. Cognitive dysfunction is hypothesised, in part, to reflect functional dysconnectivity between the frontal cortex and the striatum, although structural abnormalities consistent with this hypothesis have not yet been demonstrated in adolescence.

To characterise frontostriatal white matter (WM) tracts in relation to cognition in AOS.

A MRI volumetric and diffusion tensor imaging study.

Thirty-seven AOS subjects and 24 age and sex-matched healthy subjects.

Using probabilistic tractography, cortical regions with the highest connection probability for each striatal voxel were determined, and correlated with IQ and specific cognitive functions after co-varying for age and sex. Fractional anisotropy (FA) from individual tracts was a secondary measure.

Bayesian Structural Equation modeling of FA from 12 frontostriatal tracts showed processing speed to be an intermediary variable for cognition. AOS patients demonstrated generalised cognitive impairment with specific deficits in verbal learning and memory and in processing speed after correction for IQ. Dorsolateral prefrontal cortex connectivity with the striatum correlated positively with these measures and with IQ. DTI voxel-wise comparisons showed lower connectivity between striatum and the motor and lateral orbitofrontal cortices bilaterally, the left amygdalohippocampal complex, right anterior cingulate cortex, left medial orbitofrontal cortex and right dorsolateral prefrontal cortex.

Frontostriatal dysconnectivity in large WM tracts that can explain core cognitive deficits are evident during adolescence. Processing speed, which is affected by alterations in WM connectivity, appears an intermediary variable in the cognitive deficits seen in schizophrenia.