Investigating Shared Genetic and Environmental Aetiology between Psychiatric Disorders and Rheumatoid Arthritis

Abstract

Complex diseases are defined by having a multifactorial aetiology, consisting of multiple genetic and environmental risk factors. Complex diseases are often associated with unusual patterns of comorbidity. They are also typified by suboptimal nosology, being classified according to historical diagnostic boundaries that may not be strongly justified given emerging evidence on pathophysiology. Epidemiological studies have shown an unusual pattern of comorbidity between the psychiatric and autoimmune disorders – two broad categories of complex disease, however the aetiology underlying this overlap is yet to be established.

We present three investigations into the overlap between the psychiatric and autoimmune disorders. First, we review the epidemiological literature of the phenotypic relationship between schizophrenia and rheumatoid arthritis and perform a meta-analysis of studies meeting inclusion criteria. Next we investigate evidence for an enrichment of schizophrenia genetic risk amongst controls for rheumatoid arthritis using a number of existing statistical genetic techniques. We find no evidence that common genetic variation influences the low prevalence of rheumatoid arthritis in schizophrenia cases.

In a longitudinal population cohort we model depression genetic risk and its influence on the onset of depression, autoimmune disorders, and the comorbidity between the two. We find evidence that autoimmune disorder onset increases the risk of subsequent depression onset, independent of depression genetic risk.

In a cohort of rheumatoid arthritis patients, we investigate the role of depression genetic risk and rheumatoid arthritis severity on disease progression. We find that low mood is a significant predictor of worse treatment outcomes, including inflammatory components of rheumatoid arthritis disease severity.

To interrogate the genetic aetiology underlying comorbidity, we extend the polygenic risk score (PRS) approach in two ways. First, we develop software, PRSice, to perform PRS analyses. Secondly, we develop a novel PRS method to calculate PRS in cross-disorder scenarios.

Date of Award	2016
Original language	English
Awarding Institution	King's College London
Supervisor	Cathryn Lewis (Supervisor) & Paul O'Reilly (Supervisor)