Multi-omics investigation of major depressive disorder and antidepressant exposure

Abstract

Major Depressive Disorder (MDD) is a complex psychiatric condition affecting more than 280 million people around the world and is estimated to be a leading cause of disability by 2030. The exact biological mechanisms of MDD are unknown and current pharmacological treatments (antidepressants) for MDD are limited. Most antidepressants take around 4-8 weeks for a clinical improvement in symptoms, and an estimated 40% of people are resistant to treatment. Furthermore, antidepressants are associated with a range of side effects, including gastrointestinal disturbances, nausea and sexual dysfunction. In sum, individuals suffering from MDD will undergo various antidepressant regimens over a period of months with a host of side effects for often sub-optimal changes to their symptoms. Therefore, there is a clinical need to identify more efficacious antidepressants with a lessened side effect profile.

Furthermore, an increased understanding of how and when antidepressants exert their effects may also enable stratification of current treatments to those more likely to respond. Both goals require an increased understanding of the biological basis of MDD and the biological impact of currently prescribed antidepressants. This thesis aims to investigate both aspects using various multi-omics data in large population cohorts, primarily the UK Biobank and Generation Scotland.

Chapter 1 consists of a general introduction to MDD and antidepressants, followed by Chapter 2 which describes the datasets used in this thesis. The next three chapters constitute the main analyses conducted and are each accompanied by in-depth methodologies. Chapter 3 examines the associational and causal relationship between circulating metabolite measures and lifetime MDD in the UK Biobank. Our findings indicate a potential causal role of omega-3 and omega-6 fatty acids measures in MDD, with a colocalised region in the FADS cluster on chromosome 11. Due to high linkage disequilibrium (LD) in this region, it is challenging to identify the potential causal gene underlying these observations. Therefore, Chapter 4 aims to investigate this further using rare variants, which typically exhibit less LD and may provide further evidence of causality.

This found evidence of disparate signals in the FADS region on omega-3 fatty acids and MDD. Specifically, our analysis indicated a nominally significant burden of rare pathogenic variants in FADS1 in MDD, which was primarily driven by a singular missense variant in Exon 1. Next, Chapter 5 examines the relationship between DNA methylation and both self-reported and prescription-derived antidepressant exposure in Generation Scotland.

Our findings indicated robust associations between DNA methylation and antidepressant exposure, specifically in DGUOK-AS1 and KANK1. Finally, Chapter 6 discusses the findings of this work, its limitations and potential future directions for further elucidating on the biological basis of MDD and the effects of antidepressant exposure.

Overall, this thesis presents a series of analyses utilising multi-omics data to learn more about the biological differences in those with MDD and the potential impact of antidepressant medications. These findings contribute to a growing body of work to increase understanding of MDD and antidepressants, with the goal of developing new effective treatments.