Effective psychological treatments exist for common psychiatric disorders, in form of cognitive behavior therapy (CBT), but a significant portion (30-50%) of patients do not respond sufficiently to CBT. Not enough is known about which patients will benefit from CBT, with reliable predictors of individual differences in symptom change currently lacking. Genetic differences have been suggested to play a role, but existing evidence is limited and inconclusive.Study I described the set-up of the unique MULTI-PSYCH cohort, including 2,668 participants with panic disorder, social anxiety disorder and major depressive disorder that received internet-delivered CBT (ICBT). MULTI-PSYCH also contains genetic information and data from several Swedish register with variables spanning from birth to up to 10 years post ICBT.Study II examined associations between polygenic scores (PGS) and symptom change following ICBT in 2,668 participants from the MULTI-PSYCH cohort. Symptom change was assessed from pre- to post-treatment. Linear mixed- effects models with repeated measures revealed a significant negative association between the PGS for educational attainment (PGS-EDU) and symptom severity in the adjusted analysis (B =-. 69, p =. 034). This suggests that individuals with a higher genetic predisposition for educational attainment improved more during CBT. Additionally, a significant interaction between PGS- EDU and time was observed, indicating that PGS-EDU also was associated to symptom change rate over the course of treatment.Study III included 1,598 adults and children diagnosed with obsessive-compulsive disorder (OCD) who underwent CBT, drawn from the NORDIC cohort. Linear mixed models, adjusted for age, sex, genotyping batch, and the first five ancestry principal components, were used to examine the association between PGS and change in symptom severity from pre- to post-treatment. The PGS for schizophrenia was modestly but significantly associated with symptom change (B =. 013, p =. 04), suggesting that individuals with a higher genetic risk for schizophrenia experienced a smaller reduction in symptoms following CBT. No other PGS showed significant associations with symptom change.Study IV was a genome-wide association study (GWAS) of symptom change following CBT using data from the MULTI-PSYCH and NORDIC cohorts. The sample included 3,113 adults and children who received CBT for anxiety, depression, or OCD. The primary aim was to estimate how much of the variation in symptom change could be explained by common genetic variants (SNPs) and to identify SNPs associated with treatment outcome. Secondary analyses explored associations with additional clinical outcomes, such as pre- and post-treatment symptom severity and remission status. While no individual SNPs reached genome-wide significance for any outcome, including symptom change and remission, we observed a moderate and statistically significant SNP-based heritability for symptom improvement (h2snp =. 221, SE =. 123, p =. 036). These findings suggest that common genetic variation plays a modest role in treatment response. However, identifying specific genetic markers will likely require much larger sample sizes.Study V investigated the role of rare genetic variants in individuals with OCD, utilizing data from the NORDIC cohort. We identified copy number variants (CNVs) using genotype array data from 2,248 individuals with OCD and 3,608 unaffected controls from Sweden and Norway. The study found that rare, large (>30 kb in size). CNVs were more prevalent in OCD cases compared to controls, dominated by CNVs overlapping protein coding regions. Secondary analyses revealed that individuals carrying neurodevelopmental duplications were significantly more likely to present with comorbid autism, and those with deletions overlapping neurodevelopmental genes exhibited a poorer response to treatment. These findings support the involvement of rare CNVs in OCD risk.Together, these studies contribute to the growing understanding of how genetic variation influences psychological CBT outcomes. This thesis adopted a comprehensive strategy by including multiple psychiatric conditions, age groups, and genetic methodologies. The findings highlight the complexity of predicting treatment response and indicate that, while genetic factors are not yet clinically actionable, they may eventually form part of a broader, multimodal strategy for personalized psychiatry. They also emphasize the value of large, high-quality datasets in psychiatric genetics. To enhance predictive accuracy and clinical relevance, future research should integrate genetic data with environmental, clinical, and neurobiological measures. As sample sizes grow and analytic methods improve, the field may progress toward more precise, individualized treatment recommendations.List of scientific papersThe following publications and manuscripts were included in the thesis:I. Boberg J, Kaldo V, Mataix-Cols D, Crowley JJ, Roelstraete B, Halvorsen M, Forsell E, Isacsson NH, Sullivan PF, Svanborg C, Andersson EH, Lindefors N, Kravchenko O, Mattheisen M, :unselected: Danielsdottir HB, Ivanova E, Boman M, Fernández de la Cruz L, Wallert J 13(10):e069427. https://doi.org/10.1136/bmjopen-2022-069427II. Bäckman J, Wallert J, Halvorsen M, Crowley JJ, Mataix-Cols D Mataix-Cols D, Rück C 30(4):1510- 1517. https://doi.org/10.1038/s41380-024-02763-7
Julia Boberg (Fri,) studied this question.