Search tips
Search criteria 


Logo of schbulschizophrenia bulletinsubscriptionscontact uscurrent issuemy basketarchivemy accountsearchcontact this journaloxford journalsabout this journal
Schizophr Bull. 2017 January; 43(1): 1–2.
Published online 2016 October 31. doi:  10.1093/schbul/sbw153
PMCID: PMC5216867

Investigating Trauma as a Risk Factor for Psychosis

Two articles in this issue of the Bulletin provide new information on the role of trauma in the development of psychotic illnesses. Consideration of the role of trauma has a 30-year history, beginning with observations of psychoses in the children of Holocaust survivors.1 The 2 articles in this issue continue this line of observation and add substantive new evidence for a role of traumatic experiences in psychosis.

An epidemiological study by Okkels et al2 examines the prevalence of trauma preceding the onset of psychosis in Denmark. The comprehensive, population-wide Danish Civil Registration System was used to identify people who presented to outpatient, inpatient, and emergency clinics with moderate to severe levels of acute stress disorder or post-traumatic stress disorder. Nearly 8% of the population met this criterion. The subsequent risk of new onset schizophrenia spectrum disorder was increased 2.3-fold compared to persons without trauma history. Sixteen percent of persons with schizophrenia spectrum disorder have a trauma history. Similar elevated risk for subsequent bipolar disorder was observed. For both conditions, the risk was most markedly elevated within a year of the trauma.

There are a number of strengths in the study. The population-wide estimate of the increased risk has validity that is not achievable in many countries, because they do not have comprehensive reporting of health care encounters. This comprehensive collection of data permits accurate comparison of case and control exposure rates. The data also support attempts to address the weakness of all case-control studies, reverse causality. It is possible that psychosis itself is the cause of increased trauma, rather than the opposite. The database was thus used to assess parental history of psychosis, to attempt to see if trauma results from a vulnerability related to the strong familial risk for psychosis. Such vulnerability could result from a chaotic family environment or from genetic risk that had yet to be recognized as illness. However, there was no effect of familial risk on the occurrence of trauma reactions, although there was a positive interaction between trauma disorder and family history of schizophrenia on the subsequent incidence of schizophrenia in dually affected individuals. The temporal precedent of the trauma, occurring before the psychosis, is also evidence against reverse causality.

Several uncertainties remain, however. First, the possible reasons for the diagnosis of acute or post-traumatic stress disorders cannot be dismissed. The trauma itself is generally considered the predisposing factor, but this study did not detect trauma. Instead it detected stress reactions, some severe enough to result in inpatient treatment. These more severe forms were more highly associated with subsequent psychosis. The data viewed from this perspective raise the possibility that the stress reaction itself is the first sign of the incipient psychosis. Adding to this possibility is that the onset of a psychotic diagnosis generally occurred in about a year, the length of time now given over the prodromal period of psychosis, when symptoms are detectable but do not yet meet criteria for a database diagnosis. A final conundrum is that a focus on trauma is often of clinical interest, because of a belief that treatment of trauma might ameliorate later illness. Yet all these individuals came to clinical attention and presumably were treated appropriately.

The second article from Isovaranu et al draws upon a smaller group of patients with schizophrenia, treated in Belgium and the Netherlands (Genetic Risk and Outcome of Psyhosis or GROUP).3 A comprehensive case record was obtained from over 500 patients, which included the Positive and Negative Symptoms of Schizophrenia (PANSS) and the 5 scales of the Childhood Trauma Questionnaire-Short Form—emotional neglect, physical neglect, emotional abuse, physical abuse, and sexual abuse. Although parents and siblings are included in the larger study, the informants for the Questionnaire are not specified. A second shortcoming is that molestation was eliminated from the analysis, because of translation problems in the Questionnaire.

Emotional neglect had the highest prevalence of 79% with physical abuse at 16%. In many trauma studies, neglect exceeds other forms of emotional abuse in its impact. Only 26% of the sample endorsed emotional abuse, contradicting the still lingering presumption of the role of an abusive schizophrenogenic mother. The major finding was achieved via network analysis, which assessed associations between different traumas and symptoms rated on the PANSS in a comprehensive way to construct an interconnected network. Interestingly, none of the trauma types linked directly to either positive or negative symptoms. All the primary linkages were to general psychopathology. For example, physical neglect was associated with motor retardation. Links to the more specific positive and negative scale then occurred only through this primary link. Physical neglect links to motor retardation, a general symptom, which in turns linked to most specific negative symptoms, such as avolition. For sexual abuse, there is an even more convoluted pathway. Sexual abuse links to emotional abuse, which is then linked to the general symptom of anxiety, which is then linked to the positive symptoms of delusions, hallucinations, and paranoia. Although such links are provocative, these secondary and tertiary linkages must be viewed with some uncertainty, as the more robust effect is that the linkages from trauma to psychopathology are to general symptoms only. Nonetheless, it is a remarkable attempt to use these 2 scales to look at possible psychological mechanisms through which previous trauma influences the later symptoms of psychosis.

Both articles together leave the impression that trauma and traumatic stress reactions are part of a large number of nonspecific factors that contribute to the overlapping risk between schizophrenia and bipolar disorder. The similar effects on the 2 illnesses in the Danish registry study and the finding that primary effects are limited to general psychopathology in the GROUP sample both place trauma into the category of nonspecific risk factors. Nonspecificity is not a criticism of the trauma findings, as even genetic risk factors, presumably the most inherently specific, are overlapping.4 Physiological abnormalities, brain volume, and functional deficits show similar diagnostic overlap.5

Since trauma is a nonspecific factor, it is reasonable to ask how it compares to other risk factors that appear to lack specificity. The odds ratio that a trauma disorder will lead to schizophrenia, approximately 2.3, is similar to the OR 3.4 recorded for individuals exposed to maternal smoking in the prenatal period during their gestation.6 Like trauma, effects of maternal smoking as well as maternal infection are equally applicable to bipolar disorder and schizophrenia.

There are few attempts to look at multiple risk factors together, although most would subscribe to a multifactor model of psychosis that presumes multiple elements of genetic risk and multiple elements of environmental risk. How to combine the effect of these many proposed risk factors remains a challenge. One attempt is a risk calculator for psychosis, analogous to the multifactor risk factor calculators that are now part of the assessment of risk for future heart attack, to improve the accuracy of prediction of conversion to psychosis in high-risk patients. The symptomatic score on the Structured Interview for Psychosis-risk Syndromes was the primary factor. Seven additional factors including trauma were considered, but only decline in social function added a significant increment in predictive accuracy.7 The inclusion of genetic and environmental factors in individual risk calculation remains a challenge for mental disorders.

This year the Bulletin itself recognizes that many risk factors transcend traditional diagnostic boundaries by adding the byline The Journal for Psychoses and Related Disorders to Schizophrenia Bulletin. This expansion of coverage will support not only important cross diagnostic comparisons, but also a new wave of network analyses to look at a number of risk factors and how they act, independently or synergistically, to impact brain function, behavior, and in some individuals to engender the onset of psychosis itself.


1. Link N, Victor B, Binder RL. Psychosis in children of Holocaust survivors: influence of the Holocaust in the choice of themes in their psychoses. J Nerv Ment Dis. 1985;173:115–117. [PubMed]
2. Okkels N, Trabjerg B, Mikkel A, Pedersen C. Traumatic stress disorders and risk of subsequent schizophrenia spectrum disorder or bipolar disorder: a nationwide cohort study [published online ahead of print May 31, 2016]. Schizophr Bull. [PMC free article] [PubMed]
3. Isvoranu A-M, Boyette L, Wigman J, et al. A network approach to psychosis: pathways between childhood trauma and psychotic symptoms [published online ahead of print May 10, 2016]. Schizophr Bull. [PMC free article] [PubMed]
4. Huang J, Perlis RH, Lee PH, et al. Cross-disorder genomewide analysis of schizophrenia, bipolar disorder, and depression. Am J Psychiatry. 2010;167:1254–1263. [PMC free article] [PubMed]
5. Clementz BA, Sweeney JA, Hamm JP, et al. Identification of distinct psychosis biotypes using brain-based biomarkers. Am J Psychiatry. 2016;173:373–384. [PMC free article] [PubMed]
6. Niemelä S, Sourander A, Surcel HM, et al. Prenatal nicotine exposure and risk of schizophrenia among offspring in a national birth cohort. Am J Psychiatry. 2016;173:799–806. [PubMed]
7. Cannon TD, Changhong Y, Addington J, et al. An individualized risk calculator for research in prodromal psychosis. Am J Psychiatry. 2016;173:980–988. [PMC free article] [PubMed]

Articles from Schizophrenia Bulletin are provided here courtesy of Oxford University Press