Exploring the roles of sociodemographic, clinical, and genetic factors on the prognosis of anxiety and depressive disorders

Student thesis: Doctoral ThesisDoctor of Philosophy


Anxiety and depressive disorders are common mental health problems and are among the costliest to society. A longer course of illness, which can be exacerbated by delays in treatment-seeking and receipt, as well as poor treatment selection, adherence and response, is associated with poor later life outcomes such as disability, unemployment, and suicide. As such, early risk identification, targeted interventions and development of tools which enable timely and optimal treatment selection need to be research priorities. Work is urgently needed to identify factors associated with treatment-related outcomes, to help improve treatment selection models. Following a general introduction to anxiety, depression and some of the opportunities for improving prognostic outcomes (Chapter 1), this thesis presents four research studies with the overarching aim of identifying novel associations with treatment-related outcomes. The first two empirical studies build on prior research showing that only a fraction of individuals with anxiety and depression seek and receive treatment for their symptoms, highlighting treatment-seeking and receipt as key intermediate outcomes. These chapters include analyses performed on data from the UK Biobank, which aimed to identify sociodemographic (Chapter 2) and genetic factors (Chapter 3) associated with treatment-seeking and receipt. Chapter 4 presents the largest genome-wide association meta-analysis of prognostic outcomes following psychological therapy, using data from several clinical cohorts to test for the effects of genetic and clinical factors. In the final study (Chapter 5), a novel, minimal phenotyping approach was used to test for associations with psychological therapy outcomes, in the Genetic Links to Anxiety and Depression (GLAD) Study. All analyses in this thesis aimed to detect associations between treatment-related outcomes and sociodemographic, clinical or genetic factors, to inform the use of such variables in future prediction models of prognostic outcomes and the development of treatment screening tools. The final chapter presents a general discussion of the findings from these analyses, including the limitations and potential future directions of this research.
Date of Award1 Jul 2021
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorThalia Eley (Supervisor) & Jonathan Coleman (Supervisor)

Cite this