Over the past several years, there has been a lively academic debate about what it means to have psychotic symptoms. Although these symptoms (most commonly auditory hallucinations) can be part of the expression of a psychotic illness, on their own they are neither necessary nor sufficient for a diagnosis.

In fact, researchers from many countries have now shown that voice hearing in the general population is surprisingly common – around 10% of those questioned report having heard a voice at least once in their lives, although these experiences are usually not at the level of those with a frank psychotic disorder.

The question remains, though, whether those who have these experiences are at increased risk for mental health problems – and specifically whether these symptoms are an under recognized marker of suicidal behaviour.

Improving suicide risk assessment would obviously be hugely important to public health; general psychopathology (mainly depression) is too common to be a useful indicator of an ‘at risk’ group. Such a marker would be especially useful in young adolescents, who have the highest risk for suicidal behaviour.

Methods

Kelleher and colleagues set out to investigate whether psychotic symptoms, reported at baseline in a general community sample of young people, would predict suicide attempts in the months afterwards. The study took place in 17 schools from south-west Ireland, recruiting 1,112 students aged between 13 and 16, who were assessed at baseline and then followed-up after 3 and 12 months. All were assessed using the Strengths & Difficulties Questionnaire (SDQ) to determine the presence or absence of various disorders. Children in the top quintile on the total score were regarded as the ‘psychopathology’ group.

Presence or absence of psychotic symptoms was determined using the Adolescent Psychotic Symptoms Screener, specifically the item ‘Have you ever heard voices or sounds that no one else can hear?’ Although this item has extremely good positive and negative predictive value for the presence of psychotic symptoms, these need not be full threshold.

The outcome of interest was the score on the Paykel Suicide Scale, and particularly the question ‘Have you ever made an attempt to take your own life?’

The researchers conducted multivariate analyses that controlled for age, household type, and country of birth of participant, mother, and father, as well as previous use of cannabis.

Results

Of the 1,112 students assessed at baseline, 90% completed 3-month and 88% 12-month follow-up. 77 children (7%) reported psychotic symptoms at baseline, and over the 3 assessment points of the study there were 96 reported suicide attempts (involving 45 participants).

Baseline psychotic symptoms proved to be a highly significant predictor of suicide attempts at both 3 and 12 months:

• 3 months (7% compared to 1%, OR 10.01 (95% CI 2.24-45.49))
• 12 months (20% compared to 2.5%, OR 11.27 (95% CI 4.44-28.62))

This alone is important enough, but the researchers also examined the impact of reported psychopathology on subsequent suicide attempts, and the interaction with psychotic symptoms.

193 of the sample (17%) were in the psychopathology group, defined as scoring in the top quintile on the SDQ. Unsurprisingly, this group were more likely to report psychotic symptoms (23% vs 4% in the non-psychopathology sample). Even in this higher risk group, the presence of psychotic symptoms at baseline was very strongly related to suicide attempts in the subsequent period – an odds ratio of nearly 18 at 3 months (95% CI 3.6-88.8) and 32.7 at 12 months (95% CI 10.4-102.4). When this risk was stratified by symptom comorbidity (i.e. whether the individual met criteria for disorder in 0, 1, 2, or 3 domains), the odds ratio for both psychotic symptom prevalence and prevalence of suicide attempt increased dramatically.

Of course, it is possible that those individuals without psychotic symptoms at baseline who subsequently attempted suicide may have been experiencing such symptoms at or around the time of the suicidal behaviour. In order to explore this possibility, the researchers looked at the co-occurrence of psychotic symptoms and a suicide attempt in the 14 days before an assessment point. Those with psychopathology but without psychotic symptoms in the past 2 weeks showed no increased risk of a recent suicide attempt (OR 1.1, 95% CI 0.15-8.06), whereas those with psychotic symptoms had nearly 70-fold increased odds of suicide attempt (OR 67.5, 95% CI 11.4-399.2).

Summary

• Young adolescents who report hearing voices are at a 10-fold increased risk for a suicide attempt in the following 12 months
• This risk increases in the presence of non-psychotic psychopathology, and the greater the symptom burden the more likely a suicide attempt is. However, this may be a result of greater symptom burden being associated with increased rates of psychotic symptoms
• Very recent suicide attempts (in the past two weeks) were extremely strongly related to psychotic symptoms

Limitations

• The presence of psychotic symptoms was assumed on the basis of a single question, relating to voice hearing
• The absolute number of suicide attempts (and particularly attempts in the 2 weeks prior to assessment) was fairly low – 96 in total and only 14 in the previous 2 weeks – which has resulted in some very wide confidence intervals

Discussion

The study very clearly shows that psychotic symptoms in the general adolescent population are a clinical marker of high risk for a suicide attempt. In particular, the co-occurrence of psychopathology and psychotic symptoms presents a particularly high risk for suicide, and suggests that young people presenting with general psychological distress should be carefully assessed for possible psychotic symptomatology – even where those psychotic symptoms are attenuated (that is, with intact reality testing).

It is not clear why psychotic symptoms should be associated with such an increased rate of suicide attempts. It may be that psychotic symptoms index a general level of severity, or that those with such symptoms are more susceptible to stress or have poorer coping skills. Specific interventions to improve coping in these individuals might be warranted.

If you need help

If you need help and support now and you live in the UK or the Republic of Ireland, please call the Samaritans on 116 123.

If you live elsewhere, we recommend finding a local Crisis Centre on the IASP website.

We also highly recommend that you visit the Connecting with People: Staying Safe resource.

Link

Kelleher I. et al Psychotic Symptoms and Population Risk for Suicide Attempt: A Prospective Cohort Study. JAMA Psychiatry. 2013 Jul 17. doi: 10.1001/jamapsychiatry.2013.140. [Epub ahead of print] [PubMed abstract]