Trauma and psychotic experiences: results from a transnational survey


People with a history of trauma are often seen in clinical practice, as are those who describe hearing voices or seeing things. However, the relationship between these problems is poorly understood. Associations between PTSD and first episode psychosis have previously been discussed on The Mental Elf. Psychosis and trauma also often co-exist (Schafer and Fisher 2011). Psychodynamic interpretations of psychosis as a defense against adverse life events (e.g. Martindale B. and Summers A., 2013) support the idea that trauma leads to psychosis.

However, it’s important to consider the possible explanations here:

  • Does the trauma cause the psychosis directly?
  • Does the trauma cause mental health problems such as PTSD, which then causes psychosis?
  • Might pre-existing mental health problems increase the risk of trauma and psychosis?

A recent study carried out by the WHO World Mental Health Survey Collaborators led by John McGrath (2017) investigates the association between traumatic events and subsequent psychotic experiences, for different types of trauma and the number of traumatic events.

Does trauma cause psychosis or is it more complicated than that?

Does trauma cause psychosis or is it more complicated than that?


  • A survey with 24,464 respondents from the general population, gathered across 16 countries (The UK was not among them, but 8 European countries were included, alongside countries from North and South America, Africa, the Middle East and the South Pacific)
  • The survey utilised the CIDI (Composite International Diagnostic Interview): a standardised semi-structured interview tool based on DSM-IV criteria, administered by trained interviewers, and translated into the language appropriate for the country in which it was being used
  • Participants were assessed for psychosis using standard CIDI procedure:
    • This meant only those who scored for one of the core mental disorders (depression, mania, panic disorder, phobia, generalised anxiety etc.) went on to be assessed for psychotic experiences
    • A sample of those who did not score were also assessed for psychotic experiences, so a sample prevalence could be estimated
    • Participants were asked to exclude any psychotic experience associated with sleep or substance use
    • Those who screened positive for a possible diagnosis of psychosis or mania were also excluded from analysis
  • Traumatic experience was also assessed using CIDI criteria, of which there are six categories:
    • Collective violence
    • Bodily harm
    • Interpersonal violence
    • Intimate partner/sexual violence
    • Accidents/injuries
    • Other traumas
  • Participants were asked about age of onset of both trauma and psychotic experience
  • Data analysis attempted to account for different variables in the study by employing several different models, adjusting for different factors.


  • 71.8% (weighted) of respondents reported at least one traumatic event in their lifetime
  • Among the participants with psychotic experiences, 90.5% reported at least one traumatic event
  • Among the participants without psychotic experiences, 70.5% reported at least one traumatic event
  • 75.8% of traumatic events occurred before the onset of psychosis; whereas 19.2% occurred after it. The remainder occurred in the same year
  • More traumatic events increased the odds of later psychotic experiences in a dose-response fashion:
    • those with 1 trauma were at 2.3 times the odds of reporting psychotic experiences,
    • while those with 5 or more episodes of trauma had 7.6 times the odds
  • The effect size associated with each type of trauma showed large differences, In particular, results for being a “civilian in war zone”, and “relief worker in war zone”, suggested that these might be protective of future psychosis
  • Incorporating PTSD or other diagnosed mental disorders into the analyses did not substantially change these findings. It did, however, mean that less trauma types were significantly associated with psychotic experiences (11 in the adjusted models; as opposed to 26 in the non-adjusted).
Respondents with any traumatic events were, on average, three times more likely to subsequently have psychotic experiences compared to other respondents.

Respondents with any traumatic events were, on average, three times more likely to subsequently have psychotic experiences compared to other respondents.


The study provides evidence that traumatic events predict subsequent psychotic experiences. However, this predictive value was not the same for all types of trauma, and seems to be reversed for some traumas.

Strengths and limitations

This study has many methodological strengths. It uses a large sample size, taken from different countries, and a standardised interview tool. The CIDI uses trained interviewers and validated translations of the script. Its statistical analyses were also able to control for associations between traumatic events and mental disorder, enabling observational data to be used to answer a question of causation – that is, whether trauma caused psychosis, and whether it does so independently.

That said, the authors acknowledge there are many other potential factors that could also affect the relationship but were not considered, for example demographic factors like age and social class. Also, some studies have noted cultural differences in the CIDI’s validity in arriving at diagnoses (e.g. Gelaye et al, 2013; Quintana et al, 2012). Misdiagnoses in these countries could have meant PTSD was incompletely adjusted for in analyses, meaning the true effect size may be different to what was reported. Equally, it is also unclear whether this might affect the results’ generalizability to countries that were not surveyed (which included the UK).

Recall bias may be at work whereby participants’ decisions over whether an event is considered trauma or not may be affected by their mental state at the time of the interview. This could yield a spurious association. However, participants with diagnosed psychotic illnesses were removed from the sample, and the CIDI is very specific in its questioning, so this effect is unlikely to be large.

Participation bias may be present too, Surveys tend to recruit ‘hyper-healthy’ participants, as do certain activities and occupations. It is possible that the people who choose to work as relief workers in war zones are especially healthy, with a decreased risk of psychotic experiences, potentially explaining the apparent protective effect of such traumatic experience on risk of psychotic experiences.

Can something as individual as mental disorder really be diagnosed with a standardised interview tool?

Can something as individual as mental disorder really be diagnosed with a standardised interview tool?

Implications for practice

The study’s findings that pre-existing mental disorders need not be considered as an explanation for a relationship between trauma and later psychotic experiences is important for challenging stigmatising narratives that people with mental disorder are somehow “responsible” for their own trauma. It also highlights the importance of advocating for the prevention of trauma to reduce distress and of course future psychotic experiences.

The study also contributes to existing research on how our health can be affected by stress, both in childhood (Danese and Baldwin, 2017) and in later life (Schneiderman et al, 2008). Clarifying the impact other confounders might have on this relationship would help deepen our understanding of it. For example, individual differences between trauma types could be further explored, as could the diversity of psychotic symptoms experienced between participants. The sheer number of such variables that have yet to be explored reflects the complexity of patients we as clinicians see daily in our clinic, and highlights how research into this field has only just begun.

Research into the complex relationship between trauma and psychosis has only just begun.

Research into the complex relationship between trauma and psychosis has only just begun.

Conflicts of interest

None to declare.


Primary paper

McGrath JJ, Saha S, Lim CCW, Aguilar-Gaxiola S, Alonso J, Andrade LH, Bromet EJ, Bruffaerts R, Caldas de Almeida JM, Cardoso G, de Girolamo G, Fayyad J, Florescu S, Gureje O, Haro JM, Kawakami N, Koenen KC, Kovess-Masfety V2, Lee S, Lepine JP, McLaughlin KA, Medina-Mora ME, Navarro-Mateu F, Ojagbemi A, Posada-Villa J, Sampson N, Scott KM, Tachimori H, Ten Have M, Kendler KS, Kessler RC; WHO World Mental Health Survey Collaborators (2017). Trauma and psychotic experiences: transnational data from the World Mental Health Survey. Br J Psychiatry. DOI: 10.1192/bjp.bp.117.205955

Other references

Croft J. The trauma of psychosis: high rates of PTSD in first episode psychosis. The Mental Elf, 18 Aug 2017.

Danese A & Baldwin JR (2017). Hidden Wounds? Inflammatory Links Between Childhood Trauma and Psychopathology. Annual Review of Psychology. 68, 1 [PubMed Abstract]

Gelaye B, Williams MA, Lemma S, Deyessa N, Bahretibeb Y, Shibre T, Wondimagegn D, Lemenih A, Fann JR, Stoep AV, Zhou XH (2013). Diagnostic validity of the composite international diagnostic interview (CIDI) depression module in an East African population. Int J Psychiatry Med. 46(4):387-405.

Martindale B and Summers A (2013). The psychodynamics of psychosis (PDF). Advances in psychiatric treatment. 119, 124-131.

Quintana MI, Mari Jde J, Ribeiro WS, Jorge MR, Andreoli SB (2012). Accuracy of the Composite International Diagnostic Interview (CIDI 2.1) for diagnosis of post-traumatic stress disorder according to DSM-IV criteria. Cad Saude Publica. 2012 Jul;28(7):1312-8. [PubMed abstract]

Roberts AL, Gilman SE, Breslau J, Breslau N, Koenen KC (2011). Race/ethnic differences in exposure to traumatic events, development of post-traumatic stress disorder, and treatment-seeking for post-traumatic stress disorder in the United States. Psychol Med. Jan; 41(1): 71–83.

Salibe E, Cortina-Borja M (2008). Effect of 7 July 2005 terrorist attacks in London on suicide in England and Wales. Br J Psychiatry. 194 (1), 80-85.

Schneiderman N, Ironson G, Siegel SD (2005). Stress and health: psychological, behavioral, and biological determinants. Annu Rev Clin Psychol. 2005;1:607-28.

Photo credits


Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+
Mark as read
Create a personal elf note about this blog
Profile photo of Marcus Tan

Marcus Tan

Marcus is a CAMHS ST6 higher trainee at South London and Maudsley NHS Foundation Trust. He is am passionate about digital mental health, and currently leads a Quality Improvement project aimed at getting professionals to ask more about digital media use in their assessments. He is a committee member of Gaming the Mind: a charity of mental health professionals interested in videogames and psychiatry. He also has interests in cultural psychiatry, and is involved with research on the effect of the wider societal environment on an individual’s mental health.

More posts - Website

Follow me here –

Profile photo of William Lee

William Lee

William is a liaison psychiatrist at Derriford Hospital, Plymouth and a reader in psychiatric epidemiology at Plymouth University Peninsula Schools of Medicine and Dentistry.

More posts

Follow me here –