Cognition-mediated genetic influences on psychotic symptoms in adolescence

532 

Abstract Submission 

Sarah Chang¹, Dylan Hughes¹, Sullivan Salone¹, Jinhan Zhu², Mahnoor Hyat², Jennifer Forsyth³, Carrie Bearden⁴

¹University of California, Los Angeles, Los Angeles, CA, ²University of Washington, Seattle, WA, ³University of Washington, Seattle, CA, ⁴University of California at Los Angeles, Los Angeles, CA

Sarah Chang  
University of California, Los Angeles
Los Angeles, CA

Dylan Hughes  
University of California, Los Angeles
Los Angeles, CA

Sullivan Salone  
University of California, Los Angeles
Los Angeles, CA

Jinhan Zhu  
University of Washington
Seattle, WA

Mahnoor Hyat  
University of Washington
Seattle, WA

Jennifer Forsyth  
University of Washington
Seattle, CA

Carrie Bearden, PhD  
University of California at Los Angeles
Los Angeles, CA

Across domains of cognitive deficits in schizophrenia (SCZ), attention is among the most significantly impaired. Deficits are present long before onset of overt illness (Reichenberg et al. 2010), and are also observed in unaffected first-degree relatives (Egan et al. 2000). Collectively, findings suggest that attention impairments and accompanying alterations in neurodevelopment may reflect genetic liability to SCZ. While hundreds of genome-wide significant variants associated with SCZ diagnosis have been identified (Trubetskoy et al. 2022), less is known about the genetic etiology of psychotic symptoms across development, and variants implicated in SCZ may be developmentally dynamic. For example, previous work in adolescence shows that polygenic scores (PGS) for ADHD, but not SCZ, are associated with subthreshold psychotic symptoms (Hughes et al. 2023). This study seeks to extend these findings by evaluating whether attention and/or attention-related functional connectivity (FC) may partially explain the effect of genetic risk for ADHD on psychotic-like experiences (PLEs) in adolescence.

For a visual representation of these analyses, please see Figure 1. This study leveraged data from 5,385 adolescents of European descent from the Adolescent Brain and Cognitive Development (ABCD) study, spanning 5 yearly study visits (age at baseline study visit: 9-11, female: 49.75%) (Casey et al. 2018). Using linear mixed models, we evaluated associations between ADHD genetic risk, attention, attention-related FC, and psychotic-like experiences. Secondly, we hypothesized that attention or attention-related FC would mediate the relationship between genetic risk and PLEs.

Attention was quantified using variability in reaction time (intra-individual variability; IIV) from the NIH Toolbox cognitive battery (Chang et al. 2022). For the ABCD study, MRI acquisition and processing has been described previously (Hagler et al. 2018). Using resting state scans, we focused on within-network FC of three attention-related networks (dorsal attention network, cingulo-opercular network, and default mode network) and between-network FC of the default mode-dorsal attention networks and default mode-cingulo-opercular networks. To address motion and physiological artifacts, we removed resting state scans with framewise displacement (FD) larger than 5mm or had more than half of their volumes censored (due to volumes with framewise displacement  > 0.3 mm or derivative of root mean square variance over voxels  > 50), consistent with Chen et al. 2023. Genetic risk was estimated using PRS-continuous shrinkage (PRS-CS) and ancestral population was estimated via principal components analysis and random forest classification (Ge et al. 2019). PLEs were measured via the Prodromal Questionnaire - Brief Child Version (Karcher et al. 2018).

We found that greater ADHD PGS was associated with worse attention (ie. greater IIV; β=0.07 [CI: 0.05,0.1]; q=3.81e-10) and more severe psychotic-like experiences (β=0.1 [0.08,0.12], q<2e-16). Worse attention (ie., greater IIV) was also associated with more severe psychotic-like experiences (β=0.1 [0.08,0.11]; q<2e-16). Moreover, we found that worse attention partially mediated the relationship between ADHD PGS and psychotic-like experiences, explaining 7% of the association, according to 10,000 simulation bootstrapping (indirect effect = 0.64 [0.43,0.85], q<2e-16). We also found that reduced within-network dorsal attention network FC (β=0.08 [0.04,0.11]; q=4.18e-5) and weaker between-network default mode-dorsal attention FC (β = - 0.04 [-0.08, -0.01]; q=0.033) was associated with more severe psychotic-like experiences, but had no relationship with ADHD PGS.

Cognitive deficits are non-responsive to current treatments for schizophrenia, motivating this research. Understanding the pathophysiology of cognitive deficits during the early stages of psychosis may be critical for early intervention.

Psychiatric (eg. Depression, Anxiety, Schizophrenia) ¹

Early life, Adolescence, Aging ²

Cognition

Schizophrenia

Other - Genetic Risk