T104. IS GENETIC BURDEN IN BIPOLAR DISORDER ASSOCIATED WITH COGNITIVE PERFORMANCE? A POLYGENIC RISK SCORE ANALYSIS

Individuals with bipolar disorder (BD) experience cognitive deficits across various domains, including attention, memory, and executive functions. The use of polygenic risk scores (PRS), considering the cumulative effects of multiple genetic variants, might provide a more comprehensive understanding of the genetic architecture underlying cognitive impairment in BD. We explored the associations between several PRS and neurocognitive performance in 78 individuals with BD and 69 healthy controls. PRSs for schizophrenia (SZ-PRS), BD (BD-PRS), and major depression (MDD-PRS) were computed using Bayesian regression and continuous shrinkage priors (PRS-CS) based on the latest genome-wide association studies for these disorders. Neurocognitive performance was assessed through a comprehensive neuropsychological battery including measures of attention (CPT-II), working memory (arithmetic, letter number and digits subtest of WAIS III), processing speed (symbol search and key number of WAIS-III and TMT part A), verbal memory (CVLT and logical memory of WMS-III), visual memory (ROCF immediate recall), and executive functions (WCST, SCWT, TMT partB, ROCF copy and verbal fluency from COWAT test). In addition, measures of social cognition were also included (MSCEIT, AIHQ, HT and RMET). Considering only the BD sample, linear regression models were used to test their association with the calculated PRSs. Analyses were corrected for ancestry components and other relevant covariates. The Bonferroni correction was used to account for multiple comparisons. BD-PRS significantly differentiated HC from patients in the whole sample (B=2.59, SE=0.74, p=0.007). In the BD sample, regarding neuropsychological results, BD-PRS was associated with lower performance in processing speed in TMT-A (B=-0.79, SE=29.2, p=0.01) and Digit Symbol subset of WAIS-III (B=85.69, SE=41.68, p=0.04.), and also with visual memory assessed with ROCF - immediate recall (B=40.6, SE=13.9, p=0.006). Additionally, an association between BD-PRS and social cognition was also found in the AIHQ test (B=-18.2, SE=4.98, p=0.01). In patients with BD, SZ-PRS and MDD-PRS were not associated with cognitive impairment. Our results showed an association of BD-PRS with worse neurocognitive performance in processing speed and visual memory domains in the BD sample. We also found an association of BD-PRS with lower scores in the attributional bias domain of social cognition. Although some studies have found no associations between BD genetic burden and cognitive performance, our findings support the idea that cognitive impairments in BD are partially mediated through a higher genetic burden for BD. This could be indicative that cognitive performance arises, at least in part, due to genetic factors, thus, suggesting independence of the disease course. In turn, no association with other psychiatric genes, such as schizophrenia and depressive disorder, was found, suggesting that cognitive impairment in processing speed and visual memory are BD specific impairments. This finding could further understanding of the BD phenotype in terms of specific neuropsychological profile. To conclude, our study may have clinical implications in predicting the disease course in those patients who show poorer outcomes in these cognitive domains, facilitating, thus, early interventions and personalized treatment approaches. Further studies based on cohorts with a larger sample size are needed to replicate these findings.