Magnetic resonance imaging (MRI) studies have documented group-level brain structural and microstructural differences between individuals with psychotic disorders and unaffected individuals (Birur et al., 2017; Bose et al. 2009; Ching et al. 2022; Falkai, Schmitt, and Andreasen 2018; Schmidt et al. 2016; Andreassen et al. 2023; Haukvik et al. 2018; Hibar et al. 2018). However, whether these differences represent risk factors or develop together with the disorder or in relation to medication is not well understood. Studies of individuals at risk, whether it be a genetic risk or individuals who endorse subclinical psychosis symptoms, can provide new insights into developmental processes in psychotic disorders. In the present study, we utilized a state-of-the-art MRI-based signal intensity measure, gray/white matter contrast (GWC), and explored brain microstructure in young adults with genetic high risk or those endorsing psychotic experiences in comparison to unaffected individuals. Additionally, we tested whether individual differences in GWC in these risk groups were related to the number and experiences of psychosis-like symptoms.

Patients with schizophrenia and bipolar disorder have been found to have widespread structural brain differences compared to unaffected individuals. Both groups, on average, show lower cortical volume, with a greater reduction in volume in schizophrenia affecting frontal, temporal and occipital regions (Madre et al. 2020; Rimol et al. 2012). Patients with schizophrenia also show widespread thinner cortices and smaller cortical surface area, with the largest effects observed in the frontal and temporal regions (van Erp et al. 2016). In patients with bipolar disorder, studies predominantly result thinner cortices, but not smaller surface area, in comparison to unaffected individuals (Hibar et al. 2018; Madre et al. 2020). Patient groups also show, on average, larger ventricles, and smaller bilateral hippocampi and left thalamus (Haukvik et al. 2018; Hibar et al. 2018; van Erp et al. 2016), though patients with schizophrenia appear to have smaller hippocampal and amygdala volumes and larger right putamen volume in comparison to patients with bipolar disorder (Haukvik et al. 2018; Ohi et al. 2022; van Erp et al. 2016).

One approach to determine whether the observed differences in brain structure in patients with psychotic disorders precede illness onset is to study individuals at elevated risk for psychosis (Paolo Fusar-Poli 2017). This is particularly important for understanding the pathological mechanisms driving clinical and functional deterioration commonly seen in psychosis, which could inform the development of interventions aimed at preventing progression from a high-risk state to a clinical disorder (Cannon 2008). Individuals can be considered at high genetic risk for developing a psychotic disorder based on familiar risk (Smieskova et al. 2010a) or the presence of common or rare genetic variants associated with psychotic disorders (Karayiorgou, Simon, and Gogos 2010). Individuals from the general population who endorse psychotic experiences (PE), as well as help-seeking individuals who do not meet diagnostic criteria for a psychotic disorder, are also considered at increased risk for developing psychosis (Paolo Fusar-Poli 2017; Modenato et al. 2021; Yung et al. 1996; Jalbrzikowski et al. 2021; Drakesmith et al. 2015).

Monozygotic twins of individuals with a psychotic disorder are estimated to have a 40–50 % lifetime risk of developing a psychotic disorder, while first-degree relatives of patients with schizophrenia have an approximately 10-fold increased risk for later illness (Smieskova et al., 2010a). Familial risk for schizophrenia has been associated with smaller brain volumes and lower cortical thickness relative to subjects from low-risk families (R. C. K. Chan et al. 2011; Ivleva et al. 2013; Zwarte et al. 2019). Genetic risk can also be studied using polygenic risk scores (PRS). The PRS is a result of a large number of common genetic variants that are likely contributing to a disorder (Ripke et al. 2014). Each variant has a small effect, but by combining them one can derive a score with larger predictive power (Choi, Mak, and O'Reilly 2018). These PRSs can be calculated in unaffected individuals to study the generic risk of various diseases (Daetwyler, Villanueva, and Woolliams 2008). High PRS for schizophrenia has – in unaffected adults – been associated with lower cortical thickness in lateral orbitofrontal, inferior frontal, and posterior cingulate regions (Zhu et al. 2021). Contrary to these findings, however, a meta-analysis (van der Merwe et al. 2019) did not find any significant associations between PRS for schizophrenia and grey matter volume, white matter volume, globus pallidus volume and total brain volume.

The risk of developing a psychotic disorder ranges from 41 % to 54 % within one year and 18 % to 20 % within two years among individuals with clinical high risk for psychosis (CHR) (Paolo Fusar-Poli 2017; Paolo Fusar-Poli et al. 2015; 2020; Jalbrzikowski et al. 2021). A large body of work has used MRI to investigate brain structural differences in individuals at CHR for psychosis (Chung et al. 2019; Del Re et al. 2021; Fornito et al. 2008; P. Fusar-Poli et al. 2011; Iwashiro et al. 2012; Klauser et al. 2015; Koutsouleris et al. 2009; Kwak et al. 2019; Mechelli et al. 2011; Sun et al. 2009; Takayanagi et al. 2017; Tomyshev et al. 2019; Velakoulis et al. 2006; Ziermans et al. 2009; Zikidi et al. 2020). The largest to-date study of CHR individuals found that in comparison to unaffected individuals, individuals with CHR showed widespread lower cortical thickness, but no differences in surface area or subcortical volumes (Jalbrzikowski et al. 2021). Moreover, CHR individuals who later developed a psychotic disorder had lower cortical thickness in paracentral, superior temporal and fusiform regions in comparison to both CHR individuals who did not develop psychosis and unaffected individuals (Jalbrzikowski et al. 2021). Similar structural differences have been found among young adults who endorse psychotic experiences compared to unaffected individuals. For example, studies have found group-level differences in brain microstructures such as fractional anisotropy (Peters and Karlsgodt 2015; Smigielski et al. 2022; León-Ortiz et al. 2022; Drakesmith et al. 2016), differences in grey matter volume in the left supramarginal gyrus (Drakesmith et al. 2015), and greater brain volume in the middle frontal gyrus into the superior frontal gyrus (Fonville et al. 2019). However, another study found no differences in gray matter volume in a whole-brain analysis (Fonville et al. 2015).

Although the studies discussed above indicate that differences in brain structure exist prior to the onset of psychotic disorders, newer and less explored MRI signal intensity measures can provide additional information about brain microstructure (Norbom et al. 2019). These might offer increased biological specificity, as GWC is presumed to be a proxy measure for differences in intracortical and subjacent white matter myelin content (Eickhoff et al. 2005; Salat et al. 2009; Stüber et al. 2014). One such approach is to compute GWC from intensities sampled across the cortical mantle and within closely subjacent white matter (Salat et al. 2009), where higher GWC indicates a greater discrepancy between grey and white matter. Examining GWC among young adolescents could improve the sensitivity to detect differences between individuals at risk for psychosis and unaffected individuals. This period aligns with the protracted maturation of subjacent myelination, estimated to continue until the late 20 s or early 30 s (Drakulich et al. 2021; Norbom et al. 2020; Westlye et al. 2010). This period of myelin maturation also aligns with the vulnerability period for psychotic disorders, and dysmyelination has been proposed as one possible mechanism for these disorders (Bartzokis 2003; Whitford et al. 2012). Abnormalities in myelination have been associated with dysfunctional connectivity between the frontal lobes and extra frontal and subcortical structures in schizophrenia (Mighdoll et al. 2015).

During childhood and adolescence, GWC shows an age-related decrease, thought to partly reflect protracted intracortical myelination (Norbom et al. 2019). Only a few studies to date have used GWC to examine brain microstructure in patients with psychotic disorders (Jørgensen et al. 2016; Kong et al. 2012; 2015; Chwa et al. 2020), while no studies have examined GWC in psychosis risk groups. A study by Jørgensen et al. (2016) including adults with schizophrenia, bipolar disorder and unaffected individuals found higher GWC in pre- and postcentral gyri, the transverse temporal gyri, posterior insula, and parieto-occipital regions in patients with schizophrenia. The study also found higher contrast primarily in the left precentral gyrus in patients with bipolar disorder, but no significant differences between the two patient groups. Moreover, Jørgensen et al. (2016) found that increased GWC was associated with increased severity of hallucinations in patients, but no associations for delusions or medication. A study by Chwa et al. (2020) similarly showed, that in comparison to unaffected individuals, individuals with schizophrenia had higher GWC in the right superior frontal lobe encompassing the sensorimotor region. Additionally, they observed an association between lower GWC and increased exposure to second-generation antipsychotics within the superior frontal lobes. A large population-based study of youth by Norbom et al. (2019) also found that increased GWC was associated with more psychosis-spectrum symptoms. Contrary to the results from these three studies, Kong et al. (2012) compared GWC in a small sample of adult patients with schizophrenia and unaffected individuals and observed lower GWC in the patient group in large portions of the cortex, including frontal, temporo-parietal and lateral occipital regions. In a follow-up study, Kong et al. (2015) also found lower regional GWC in the patient group. In sum, the three largest existing studies indicate increased GWC primarily in highly myelinated sensory and motor regions in patients with schizophrenia and increased GWC to be associated with psychosis symptoms, although discrepant findings have also been reported. Our understanding of the progressive effect these disorders have on GWC is however limited as no studies to date have examined GWC in individuals with genetic risk for psychosis or those who endorse psychotic experiences.

The present study aimed to explore differences in brain microstructure before the onset of psychotic disorders by comparing GWC between young adults with genetic risk, those endorsing psychotic experiences, and unaffected individuals. We hypothesized that we would observe regionally higher GWC in the psychotic experience group compared to unaffected individuals, based on previous findings (Chwa et al. 2020; Jørgensen et al. 2016; Norbom et al. 2019). To our knowledge, there is a lack of research on GWC in groups with genetic risk for psychosis, with previous studies reporting limited evidence for associations between PGS for schizophrenia and brain macrostructure (Lancaster et al. 2019; Papiol et al. 2014; van der Merwe et al. 2019). Furthermore, no association between risk for schizophrenia and genes associated with myelination have been found (Goudriaan et al. 2014; Stokowy et al. 2018). Thus, we hypothesized that there would be no significant difference in GWC between the high genetic risk group and the low genetic risk group. We also hypothesized regionally higher GWC in the group of individuals who endorsed psychotic experiences relative to the genetic risk group. Finally, we hypothesized that individual differences in the number and experiences of psychosis-like symptoms would be positively associated with the GWC in the psychotic experiences group, based on findings from a previous study observing a positive association between hallucination symptoms and GWC in patients with psychotic disorders (Jørgensen et al. 2016). We did not expect to observe a similar association in the genetic risk group since previous studies have reported no association between PRS for schizophrenia and psychotic symptoms (Jones et al. 2016; Lancaster et al. 2019). Overall, the current study aimed to provide new knowledge of the neural mechanisms underlying risk for psychotic disorders and investigate whether previously reported differences in GWC in patients represent pre-existing risk factors or emerge later in the course of the disorders.