Contributions of common and rare genetic variation to different measures of mood and anxiety disorder in UK Biobank

Katzourou, Ioanna, Barroso, Ines, Benger, Lauren, Ingason, Andres, Stow, Daniel, Tsang, Ruby, Wood, Megan, Kirov, George ORCID: https://orcid.org/0000-0002-3427-3950, Walters, James ORCID: https://orcid.org/0000-0002-6980-4053, Owen, Michael J. ORCID: https://orcid.org/0000-0003-4798-0862, Holmans, Peter ORCID: https://orcid.org/0000-0003-0870-9412, Van den Bree, Marianne B.M ORCID: https://orcid.org/0000-0002-4426-3254 and Consortium, LINC 2025. Contributions of common and rare genetic variation to different measures of mood and anxiety disorder in UK Biobank. BJPsych Open 11 (3) , e97. 10.1192/bjo.2025.43

Preview

PDF - Published Version Available under License Creative Commons Attribution. Download (884kB) | Preview

Abstract

Mood and anxiety disorders co-occur and share symptoms, treatments and genetic risk, but it is unclear whether combining them into a single phenotype would better capture genetic variation. The contribution of common genetic variation to these disorders has been investigated using a range of measures, however the differences in their ability to capture variation remain unclear, while the impact of rare variation is mostly unexplored. Aims: We aimed to explore the contributions of common genetic variation and Copy Number Variations associated with risk of psychiatric morbidity (P-CNVs) to different measures of internalising disorders. Method: We investigated eight definitions of mood and anxiety disorder, and a combined internalising disorder, derived from self-report questionnaires, diagnostic assessments, and electronic healthcare data (EHR). Association of these definitions with polygenic risk scores (PRSs) of major depressive disorder and anxiety disorder, as well as presence of a P-CNV, was assessed. Results: The effect sizes of both PRSs and P-CNVs were similar for mood and anxiety disorder. Compared to mood and anxiety disorder, internalising disorder resulted in higher prediction accuracy for PRSs, and increased significance of associations with P-CNVs for most definitions. Comparison across the eight definitions showed that PRSs had higher prediction accuracy and effect sizes for stricter definitions, whereas P-CNVs were more strongly associated with EHR- and self-report- based definitions. Conclusions: Future studies may benefit from using a combined internalising disorder phenotype, and may need to consider that different phenotype definitions may be more informative depending on whether common or rare variation is studied.

Item Type:	Article
Date Type:	Publication
Status:	Published
Schools:	Schools > Medicine
Publisher:	Cambridge University Press
ISSN:	2056-4724
Date of First Compliant Deposit:	30 April 2025
Date of Acceptance:	24 February 2025
Last Modified:	20 May 2025 11:45
URI:	https://orca.cardiff.ac.uk/id/eprint/177881

Actions (repository staff only)

Edit Item