Re-examining adolescent bipolar disorder and related psychopathology using meta-analysis and item response theory

Abstract

The aims of this thesis were to summarise and synthesise the current research into BD; critically evaluate existing literature to assess whether age of onset is associated with poorer outcomes in BD; and examine whether individual symptoms may be clinically useful as risk markers in childhood and adolescence. Study one was a meta-analysis of existing research investigating outcomes associated with an early onset of BD. Data was analysed from fifteen papers that compared clinical presentation and outcomes in BD grouped according to age of onset (Total n = 7370). Clinical features found to have the strongest relationship with an earlier age of onset were those amenable to intervention such as comorbid anxiety, substance use, and treatment delay. Study two used a novel analytical approach to evaluate whether individual clinical symptoms differed in their capacity to discriminate between those scoring high and low on underlying traits of depression and mania; or in the information they provided in relation to severity. The sample consisted of n=186 participants aged 12–21yrs including n=105 with a first degree relative diagnosed with BD (At Risk); n=63 control participants; and n=18 with a confirmed diagnosis of BD. Results support hypotheses from previous research that specific mood symptoms are more informative of risk in BD than general symptoms; and are in line with previous findings that indicate increased energy is a core feature of mania. These findings are important in relation to ongoing controversy around diagnoses of paediatric BD, and the broadening of diagnostic criteria. Overall, the studies in this thesis provide information useful to clinicians in identifying at risk populations that may benefit from early support, monitoring and intervention; and identify key risk areas in adolescent populations informing important areas of future research.