An Epidemiological Investigation of the Relationship Between Hallucinations and Suicidal Behaviour Across the Adult Lifespan and in High-Risk Groups.

Background: Psychotic experiences (PE) are common in the general population, and hallucinations are most commonly reported. Research over the past ten years has, in particular, highlighted a relationship between PE and suicidal behaviour.

Aims: The overall aim of this thesis was to investigate hallucinations as a risk marker for suicidal behaviour in (1) the general population, including looking at variation with regard to age, and (2) in specific groups already known to have elevated risk of suicidal behaviour, including (2a) individuals with a chronic neurological disorder (seizure disorders) and (2b) individuals with specific psychosocial adversity (victims of sexual assault). A further aim was to explore developmental sensitivity in the relationship between psychosocial adversity (sexual victimization) and psychotic experiences. In order to look at the overall general population, a quantitative synthesis of the literature was conducted examining the longitudinal relationship between psychotic experiences and subsequent suicidal ideation, attempts and deaths in the general population (Study 1). Second, the prevalence and clinical significance of hallucinations across the adult lifespan was assessed (Study 2). Third, it was examined whether hallucinations are a risk marker for psychopathology, suicidal ideation and suicide attempt in individuals with a seizure history (Study 3). Then, the prevalence of hallucinations, delusional beliefs and psychotic disorder in the context of sexual assault was investigated (Study 4). Finally, it was examined whether hallucinations were a risk marker for suicidal behaviour specifically in the context of sexual assault (Study 5).

Method: A systematic review and meta-analysis was conducted for study 1. Analyses for the remaining studies were conducted using the Adult Psychiatric Morbidity Surveys (APMS). The APMS studies are cross-sectional, national household surveys representative of the general UK population. They have been conducted every seven years since 1993 (2000, 2007, 2014) and represent the largest epidemiological survey worldwide that uses consistent methods to assess the prevalence of treated and untreated mental health conditions. The consistent methods allow different study waves to be combined to allow robust testing of hypotheses using large samples.

Results: PE were associated with increased risk for subsequent suicidal ideation, suicide attempt and suicide death in the general population (Study 1). The prevalence of hallucinations decreased significantly across the adult lifespan from age 16-95 years. Whilst hallucinations were associated with increased risk for psychopathology and suicidal behaviour in all age groups when compared to individuals without hallucinations, hallucinations in older adults were less likely to be associated with a co-occurring mental disorder, suicidal ideation or suicide attempt compared to early adulthood and middle age (Study 2). There was an increased prevalence of hallucinations in individuals with a seizure history. Hallucinations were a risk marker for psychopathology, suicidal ideation, and suicide attempt in individuals with a seizure history, with 65% meeting criteria for one or more mental health disorder and 53% disclosing at least one suicide attempt (Study 3). Individuals with a history of sexual assault also had a higher prevalence of hallucinations, delusional beliefs and psychotic disorder compared to individuals without a sexual assault history (Study 4) and hallucinations were a strong risk marker for suicidal behaviour in the context of sexual assault (Study 5). The relationship between hallucinations and suicidal ideation and behaviour in all studies was not explained by other, non-psychotic measures of psychopathology.

Conclusion: Hallucinations are common in the general population and occur at an elevated prevalence in groups known to be at high-risk for suicide. What is more, even within groups at already elevated risk, hallucinations are strong markers of increased risk for suicidal behaviour. There is age-related variation in the clinical significance of hallucinations in that they are less likely to be associated with mental ill health, suicidal ideation and suicide attempt in older age compared to hallucinations in younger and middle-age. Hallucinations are a valuable risk marker for identifying individuals with a particularly increased odds of suicidal ideation and behaviour in high-risk groups.