Investigating age at onset in bipolar disorder

Thesis

Bipolar disorder (BD) is a complex mental health disorder with a heterogenous clinical course, diagnostic delays, high relapse rates, and sub-optimal treatment outcomes. Age at onset (AAO) has been proposed as a useful specifier for defining more homogeneous BD subgroups that can inform clinical course and symptom profiles. This thesis aimed to investigate the utility of AAO as a clinical specifier by defining and validating AAO subgroups; identifying predictive factors for BD AAO; investigating the relationship between premorbid factors, AAO, and functional outcomes; and examining the association between AAO and mood instability.

Using mixed methods across four experimental chapters, this thesis provides novel insights into the role of AAO in BD. Chapter 2 presents a systematic review of AAO distributions in BD and provides a recommended AAO definition. Chapter 3 uses machine learning approaches to explore predictive factors for BD AAO. Chapter 4 investigates the potential pathways between premorbid factors, BD AAO, and functional outcomes using prospective data. Chapter 5 examines the association between AAO and mood instability using longitudinal mood monitoring data. Findings reveal a trimodal distribution of AAO in BD, with distinct early-life risk factors, which may represent potential causal pathways to clinical outcomes. Additionally, mood instability is identified as a promising target for intervention in the clinical trajectory of BD. These results have important theoretical and practical implications, informing early intervention strategies and providing further evidence for the distinctiveness of AAO subgroups.

The thesis concludes with a general discussion of the findings and implications for future research. Overall, this thesis provides valuable insights into the role of AAO in BD, highlights the importance of identifying more homogeneous subgroups to improve diagnosis and treatment outcomes, and underscores the need for continued research in this area.

Authors: Bolton, S

• DOI: 10.5287/ora-xq5mvegry
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: bipolar disorder; machine learning; structural equation modelling; age at onset; BD; AAO; adolescent mental health
• Subjects: Psychology; Clinical psychology; Bipolar disorder; Psychiatry