Many have grappled with the question of where ‘normal’ ends and ‘madness’ begins. After all, one is surely either ‘psychotic’ or ‘not psychotic’? Yet, a growing body of evidence would suggest psychosis or ‘psychotic-like symptoms’ exist in the general population on a sort of continuum.

For example, we know from a 2013 meta-analysis (Linscott et al, 2013), the estimated prevalence of so-called psychotic-like experiences was 7.2%, clearly much higher than the lifetime morbid risk for actual psychotic disorder. Thus, it appears the reality of psychosis (and indeed ‘normality’) is far from all or nothing.

In a recent JAMA Psychiatry paper (McGrath et al, 2015) the authors report on psychotic experiences (PEs) in the general population, drawing on data from an impressively large international sample of 31,261 adult respondents from 18 countries.

In their own words, this is ‘the most comprehensive description of the epidemiological landscape of PEs published to date’.

Methods

This study aimed to capture detailed information on particular psychotic-like experiences (prevalence, frequency and correlates) including differences by age, gender, country of origin and level of income.

How did they go about it?

The authors collated data from the World Health Organisation (WHO) Mental Health Surveys, conducted from 2001-2009. These are described as rigorously implemented general population surveys aimed at addressing the global burden of mental disorders. Further details on the global plan can be found at http://www.hcp.med.harvard.edu/wmh/.

Which countries were included?

18 of the 26 countries surveyed included the Composite International Diagnostic Interview (CIDI) psychosis module. They are: low and middle income countries (Colombia, Iraq, Nigeria, Peoples Republic of China and Peru); upper middle income countries (Brazil, Lebanon, Mexico and Romania); and high income countries (Belgium, France, Germany, Italy, Netherlands, New Zealand, Portugal, Spain and the United States).

Previous studies using general population surveys have tended to focus on higher income countries, with some exceptions (Nuevo et al, 2012).

The response rate (weighted mean) was fairly high overall at 72.1%, with the lowest response rates in France (45.9%) and Belgium (50.6%).

The sample

All respondents were adults (age range 18-100) and most countries used nationally representative samples.

People who reported having an existing mental disorder (asked as part of the interview) were excluded.

What was asked?

This is a key area where the authors argue that their data is superior to similar surveys conducted in the past. Six questions covering different PEs were asked. This was considered to be an improvement on previous work because:

• Importantly, they distinguished between HEs (visual and auditory hallucinations), DEs (bizarre delusional items), and paranoid delusional items, moving beyond crude dichotomous ‘PE Vs non-PE’ categorisations
• The survey explored not only lifetime prevalence but also frequency of the above experiences (once, 2-5x, 6-10x, 11-100x and >100)
• They asked separately about experiences in the LAST 12 MONTHS
• There were questions to ‘rule out’ PEs secondary to drug or alcohol use (though organic disorders/physical illness was not considered, potentially a cause of visual hallucinations or other phenomena)

Data Analysis

They calculated weighted prevalence estimates for the various PE types, and provided odds ratios and 95% confidence intervals. The multivariate regression model was adjusted for country, age at interview, sex, employment status, marital status, education level, household income and nativity.

Results

• The mean lifetime prevalence (SE) of having a PE was 5.8% (0.2%)
• Hallucinatory experiences were more common than other types (HEs were at 5.2% [0.2%] and DEs at only 1.3% [0.1%])
• The vast majority of people with PEs had only one such experience (72%)
• The more different types of PEs you experienced, the more episodes you are likely to have
• Women experience significantly more PEs than men (6.6% [0.2%] vs 5.0% [0.3%]) – this is true for hallucinations but interestingly not for DEs
• The lifetime prevalence estimates were significantly higher in middle/high income countries (7.2% [0.4%], 6.8% [0.3%]) than low income ones (5.0% [0.3%])
• Being non-married, unemployed and on low income also increased your rates of PE

Strengths

• Sample: this is a very large, international sample, covering high and low income countries (notably more than half European, but the results are weighted)
• Statistical analysis appears to be solid (from my understanding of it!)
• Survey: more detailed questions than many other studies, steps to exclude drugs/alcohol, yielding much more meaningful results

Limitations

• Of course, with cross-sectional data no inferences can be made on causality (i.e. are unemployed people more at risk of having PEs or are people with PEs more at risk of being unemployed?)
• They used lay interviewers with basic training, which could affect the quality of the interviews
• While they aimed to exclude people who reported a diagnosis of mental disorder, cases could easily have been missed, and therefore not excluded
• These are not the easiest symptoms to report to a stranger face-to-face; there is likely to be underreporting, making estimates conservative
• The questions focus on lifetime prevalence, so there is a risk of recall bias
• There is no mention of missing data levels, and accordingly no explanation for how this was handled
• There is evidence (for example from the ALSPAC cohort) that young people present with PEs before their teens, and a large group of interest will not have been studied by focusing solely on adults – but this may well be something that can be looked at in future
• Studies of ‘ultra high risk’ of psychosis populations have found those that do convert to psychosis tended to have negative symptoms, poor functioning and disorders of thought content (Nelson et al, 2013), none of which are captured in this survey

Conclusions and further thoughts

This is an important paper, which highlights the need to think more carefully about what we mean by psychotic experiences and indeed psychosis.

The authors have demonstrated different types of PE behave differently, and there is variation by gender and country of origin. Quite how these experiences fit in with psychotic illness itself is unclear at present, and work with ultra-high risk for psychosis groups suggest most people don’t actually go on to develop psychotic illness over time (Nelson et al, 2013).

Crucially, this adds to the debate over what should be done about such experiences, and future work looking at how having these experiences influence health/functioning/quality of life outcomes will be of interest. As a psychiatrist, I am cautious about overmedicalising experiences, and hope any calls for treating PEs in otherwise healthy, functioning people will be met with sensible scepticism, while we gather better evidence.

Links

Primary paper

McGrath JJ, Saha S, Al-Hamzawi A, et al. Psychotic Experiences in the General Population: A Cross-National Analysis Based on 31 261 Respondents From 18 Countries. JAMA Psychiatry. Published online May 27, 2015. doi:10.1001/jamapsychiatry.2015.0575. [PubMed abstract]

Other references

Linscott RJ, van Os J. An updated and conservative systematic review and meta-analysis of epidemiological evidence on psychotic experiences in children and adults: on the pathway from proneness to persistence to dimensional expression across mental disorders. Psychological Medicine. 2013;43(6):1133-49. Epub 2012/08/02.

Nuevo R, Chatterji S, Verdes E, Naidoo N, Arango C, Ayuso-Mateos JL. The continuum of psychotic symptoms in the general population: a cross-national study. Schizophrenia Bulletin. 2012;38(3):475-85. Epub 2010/09/16.

Nelson B, Yuen HP, Wood SJ, Lin A, Spiliotacopoulos D, Bruxner A, et al. Long-term follow-up of a group at ultra high risk (“prodromal”) for psychosis: the PACE 400 study. JAMA Psychiatry. 2013;70(8):793-802. Epub 2013/06/07.