Contributions of Polygenic Risk Score for Genomic Common Executive Function and P- Factors to Longitudinal Variation in Cognitive Performance.

The PsyCourse Study, conducted in Germany and Austria, researches the longitudinal course of severe mental disorders across the 'psychotic-to-affective continuum', including neurotypic individuals. Over 18 months and four measurement points, participants underwent comprehensive assessments, including collection of biomaterial, at approximately 6-month intervals. Metadata on the assessments are available online. This study focuses on polygenic scores (PRS) for a common executive function (cEF) factor and their role in longitudinal cognitive variations. EFs, crucial for regulating thoughts and social interactions, are impaired in psychiatric disorders such as major depressive disorder, bipolar disorder, and schizophrenia, and are mediated via the dorsolateral prefrontal cortex. The present thesis examines whether cEF PRS, which have been shown to play a significant role in cross-sectional cognitive performance, also play a role in cognitive variation over time. Specifically, cognitive performance improves over repeated assessments (i.e. study visits), interpreted as effects of increasing familiarity or strategy. We examined cognitive test performance over time in three repeatedly assessed cognitive tests: Trail Making Test Part B (TMT-B), Verbal Digit-Span Backward (VDS-B), and Verbal Digit-Span Forward (VDS-F). These tests tap the executive domain to varying degrees. Statistical analysis with Linear Mixed Models indicates, as expected, significant effects of time reflecting improved test scores through repeated test administrations. An overall effect of cEF PRS was observed for TMT-B and VDS-B. There was a nominally significant Visit x cEF PRS interaction for the TMT-B, which did not survive adjustment for multiple testing. Thus, as hypothesized, the non-significant Visit × cEF PRS interactions suggests that enhancements in cognitive scores are more linked to familiarity or strategy than genetic factors. Future studies should examine latent phenotypic EF scores, to eliminate measurement error.