Abstract
Purpose
This study assessed the distribution of posttraumatic stress disorder (PTSD) symptoms and psychosis indicators among a large sample of individuals with a lifetime diagnosis of PTSD. The identification of a psychotic PTSD subtype was also predicted.
Method
Using data from the National Comorbidity Survey a latent class analysis was conducted on the PTSD symptoms of intrusion, avoidance, and hyperarousal and the psychosis hallucination and delusion indicators.
Results
Results indicated four latent classes, two of which had relatively high probabilities of endorsing the hallucination and delusion indicators. These classes were associated with a broad range of traumatic experiences. One particular class had high probabilities of endorsing both the psychosis indicators and the PTSD symptoms and was associated with a broad range of comorbid psychiatric disorders.
Conclusion
There was a candidate class that met the characteristics expected to be evident in a psychotic PTSD subtype.
Similar content being viewed by others
Introduction
The Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) [1] defines the symptoms of posttraumatic stress disorder (PTSD) in terms of exposure to a traumatic event. However, there is a growing research literature that has also identified associations between trauma (especially early traumatic experience) and diagnoses of psychosis or the occurrence of psychotic-like experiences [2–6]. The high rates of psychotic disorder and psychotic symptoms in PTSD populations [7–9], the phenomenological similarity of some PTSD and psychotic symptoms [10], and the shared etiology associated with both forms of psychopathology [11] have led to suggestions of a ‘psychotic PTSD’ subtype.
Sautter et al. [12] identified a group of war veterans who met the DSM-IV criteria both for psychotic disorder and for PTSD. This psychotic PTSD group had significantly higher scores on measures of depression, aggression, and general psychopathology compared to those with PTSD or psychosis alone. Ivezic et al. [13] reported that war veterans who experienced PTSD with psychotic symptoms had high levels of comorbid depression, delusional disorder, and anxiety disorder. Also, based on a community sample, Sareen et al. [14] reported an increased likelihood of psychotic symptom experience in individuals who had a lifetime diagnosis of PTSD. In addition this group had significantly higher probabilities of meeting the criteria for other psychiatric disorders. These and other studies have shown that psychotic PTSD groups generally have lower levels of functioning and higher levels of other psychiatric problems. While the psychotic PTSD groups in these studies did not require symptoms to present in any particular temporal order, Braakman et al. [15] proposed a new diagnostic entity, PTSD with secondary psychotic features (PTSD-SP), which was characterized by psychotic symptoms which were emergent after the onset of PTSD. This new entity was shown to be associated with a broad range of comorbid diagnoses.
Many studies alluding to a psychotic PTSD subtype have been based on the small samples of war veterans [9, 13, 16, 17] making generalizations to other trauma cohorts difficult. In addition, many studies have used different assessments in order to classify individuals. Furthermore, the criteria for psychotic PTSD have been predefined by the researchers and differ across studies. For example, in some studies psychotic PTSD was based on the diagnosis of both PTSD and psychotic disorder [12], whereas other studies only required some psychotic symptoms to be present [14].
The primary aim of this study was to test if there was a homogeneous group of individuals who could be described in terms of a psychotic PTSD subtype, rather than simply PTSD with comorbid psychosis. The identification of homogenous groups, or latent classes, was based on the latent class analysis (LCA). This analysis allows different classes (groups) to be objectively identified based on the patterns of symptom endorsement rather than a priori definitions imposed by researchers. This study utilized the same community sample as Sareen et al. [14] who created a “PTSD with psychotic symptoms” group based on a diagnosis of PTSD and the endorsement of one or more psychosis indicators. The LCA approach in this study extended this analysis as it (1) included multiple psychosis indicators to assess which particular types of psychotic experience were most commonly endorsed; (2) explicitly tested for a homogeneous group characterized by a high probability of psychosis and PTSD symptom endorsement; (3) allowed for an estimate of the size of such a group; and (4) estimated the probability of endorsing each psychosis and PTSD symptom within the group. On the basis of previous research it was hypothesized that a psychotic PTSD class would, firstly, be characterized by high probabilities of endorsing both PTSD symptoms and psychosis indicators. Importantly, psychotic PTSD classes would have to be qualitatively different, rather than just quantitatively different, from other classes in order to be judged distinct; classes differing only quantitatively would represent groups of individuals who differed only in terms of an underlying continuum of severity.
The second aim of the study was to assess the relationship between the psychotic PTSD subtype and a range of traumatic experiences. Variables representing a broad range of traumatic experiences were used to determine if there were specific traumas that were associated with psychotic PTSD. It was also predicted that a psychotic PTSD subtype would have an increased risk of other comorbid psychiatric diagnoses compared to non-psychotic classes. In order for a psychotic PTSD class to be considered qualitatively, rather than just quantitatively different from non-psychotic PTSD classes, the odds ratios across the classes should not indicate a graded response.
Methods
The National Comorbidity Survey (NCS) [18] was a collaborative epidemiologic investigation (1990–1992) based on a stratified, multi-stage, area probability sample of non-institutionalized persons aged between 15 and 54 years. The NCS was conducted across 48 coterminous states of America and was designed to study the prevalence and correlates of DSM-III-R [19] disorders. The initial survey employed a household sample of over 8,000 respondents and a sub-sample of the original respondents completed the additional NCS Part II survey (N = 5,877) that contained a further detailed risk factor battery and additional diagnoses. Of this sample 48% were male, the average age was 32 years. A full description of the NCS is available [18].
Participants
Based on the NCS Part II survey there were 591 participants with a lifetime diagnosis of PTSD which represented a weighted prevalence of 7.3%. The full NCS population sample had 66 participants with a lifetime diagnosis of psychosis which represented a weighted prevalence of 0.7%. There was a positive association between PTSD and psychosis diagnosis (χ2 = 25.89, df = 1, p = 0.00) with 19 participants receiving a lifetime diagnosis of PTSD and psychosis. This represented 3.2% of the sub-sample of participants who had a lifetime diagnosis of PTSD.
Subsequent analyses were based on those participants with a lifetime diagnosis of PTSD after excluding those with a lifetime diagnosis of psychosis. This sample (N = 568) had a mean age of 33.55 years (SD = 9.97). There were more females (69.2%) than males in the PTSD sample compared to the non-PTSD participants (50.1%) and this was statistically significant (χ2 = 75.05, df = 1, p = 0.00).
Measurements
Information on symptoms of PTSD, psychosis indicators, demographic variables, traumatic experiences and other diagnoses was derived from The Composite International Diagnostic Interview (CIDI) [20]. The CIDI is a fully structured interview that produced diagnoses according to the definitions and criteria of the DSM-III-R [19].
Posttraumatic stress disorder
The Posttraumatic Stress Disorder module of the CIDI (Section U) was used for the diagnosis of lifetime PTSD based on the DSM-III-R criteria. Factor analytic research has suggested that there are eight symptoms, currently associated with the PTSD diagnosis, that are non-specific to PTSD, but instead measure dysphoria [21]. Subsequent studies have reported very high correlations between the dysphoria factor and measures of depression [22] and low associations with measures of trauma [23]. Indeed, when depression scores are controlled for the factor loadings for the dysphoria factor (and the correlation between the dysphoria factor and other PTSD factors) are significantly attenuated [22]. These findings have led to suggestions for the removal of the dysphoria symptoms from the list of PTSD diagnostic criteria [24, 25]. Furthermore, research on the phenomenological similarity of PTSD and psychotic symptomologies has focused on the symptoms of intrusion and hyperarousal [26]. For these reasons, and also to reduce the number of variables in the analysis, the nine symptoms that comprised the intrusion (B1–B5), avoidance (C1, C2), and hyperarousal (D1, D2) symptom clusters were selected.
The items were as follows:
-
1.
(B1) Did you keep remembering the event when you did not want to?
-
2.
(B2) Did you keep having dreams or nightmares about it afterwards?
-
3.
(B3) Did you ever suddenly act or feel as though the event was happening again, even though it wasn’t?
-
4.
(B4) Did you ever get very upset when you were in a situation that reminded you of it?
-
5.
(B5) Did you sweat or did your heart beat fast, or did you tremble when reminded of the upsetting experience?
-
6.
(C1) Did you try hard not to think about it?
-
7.
(C2) Did you go out of your way to avoid situations that might remind you of the event?
-
8.
(D4) Did you become overly concerned about danger or overly careful?
-
9.
(D5) Did you become jumpy or easily startled by ordinary noises or movements?
Psychosis
The psychosis screening section of the CIDI (Section K) contained 13 items related to classic psychotic-like symptoms involving, for example, persecution, thought interference, and hallucinations. Items were responded to using a simple ‘yes’ or ‘no’ response format. The CIDI was used to assess the lifetime prevalence of non-affective psychosis (a summary category made up of schizophrenia, schizophreniform disorder, schizoaffective disorder, delusional disorder, and atypical psychosis). The diagnosis of psychosis was based on the clinical reinterviews administered by experienced clinicians using an adapted version of the Structured Clinical Interview for DSM-III-R (SCID) [27].
For this study eight items were selected that represented hallucinations and delusions. The items were as follows:
-
1.
(Spy) Believed that people were spying on you or following you?
-
2.
(Pois) Believed that you were being secretly tested or experimented on, that someone was plotting against you, or that someone was trying to poison you or hurt you?
-
3.
(Mess) Believed that you were being sent special messages through television or the radio, or that a program had been arranged just for you alone?
-
4.
(Hyp) Felt strange forces working on you, as if you were being hypnotized or magic was being performed on you, or you were being hit by laser beams or X-rays?
-
5.
(Vis) Experience of seeing something or someone that others present could not see—that is, had a vision when you were wide awake?
-
6.
(Aud) Experience of hearing things that other people could not hear, such as noises or a voice?
-
7.
(Olfac) Bothered by strange smells around you that nobody else was able to smell, perhaps even odors coming from your own body?
-
8.
(Tact) Unusual feelings inside or on your body, like being touched when nothing was there or feeling something moving inside your body?
Demographics
The following background variables, or demographic factors, were used in this study: age, sex (0 = male, 1 = female), ethnicity (0 = non-white, 1 = white), education (0 = less than 16 years education, 1 = +16 years education), annual income (0 = $19,000 or more, 1 = less than $19,000), household composition (0 = does not live alone, 1 = lives alone).
Traumatic experiences
During the administration of the PTSD module, participants were provided with a booklet which listed the traumatic experiences. Each trauma was numbered and participants were asked to identify the number of the event rather than naming it. This has been shown to increase participant’s willingness to report such information [28].
The traumatic experiences were as follows:
-
1.
You had direct combat experience in a war
-
2.
You were involved in a life-threatening accident
-
3.
You were involved in a fire, flood, or natural disaster
-
4.
You witnessed someone being badly injured or killed
-
5.
You were raped (someone had sexual intercourse with you when you did not want to by threatening you or using some degree of force)
-
6.
You were sexually molested (someone touched or felt your genitals when you did not want them to)
-
7.
You were seriously physically attacked or assaulted
-
8.
You were physically abused as a child
-
9.
You were seriously neglected as a child
-
10.
You were threatened with a weapon, held captive, or kidnapped
Clinical variables
The CIDI lifetime diagnoses (with hierarchy) of a range of clinical variables were used: alcohol dependence, drug dependence, major depression, bipolar disorder, mania, and generalized anxiety disorder. World Health Organization field trials of the CIDI have documented good inter-rater reliability, [29, 30] test–retest reliability [31–33], and validity of almost all diagnoses [33].
Analysis
Analyses were conducted in three phases. First, a latent class analysis (LCA) was conducted to determine the correct number of classes. LCA is a statistical method used to identify homogeneous groups, or classes, from categorical multivariate data. In this study, LCA was employed to determine the number and nature of classes based on the responses to the nine PTSD items and eight psychosis screening items. All the variables were declared as categorical and the model parameters were estimated using robust maximum likelihood. There was very little missing data (lowest pairwise covariance coverage was 0.97) and this was treated using full information maximum likelihood estimation under the assumption of data missing at random [34]. The fit of five models (a 2-class model through to a 6-class model) was assessed. Selection of the optimal number of latent classes was based on the several statistical fit indices. The statistical fit indices were: Akaike information criterion (AIC) [35], Bayesian information criterion (BIC) [36], sample-size adjusted BIC (ssaBIC) [37], the bootstrapped likelihood ratio test (BSLRT) [38], and entropy measures [39]. The information statistics AIC, BIC, and ssaBIC are goodness-of-fit measures used to compare competing models; lower observed values indicate better fit. The bootstrap likelihood ratio test was used to compare models with increasing numbers of latent classes. When a non-significant value (p > 0.05) occurs it suggests that the model with one less class should be accepted. Entropy is a standardized measure of how accurately participants are classified. Entropy values can range from 0 to 1 with higher values indicating better classification. Recent research has suggested that the BSLRT is the best method for determining the correct number of classes [40, 41]. In order to use the BSLRT no weighting variable was used in the part of the analysis.
Second, covariates (demographic variables and traumas) were added to the model to examine which variables significantly predicted class membership. The parameters linking the covariates with the latent classes were multinomial logistic regression coefficients presented as odds ratios. Third, the classes were used to predict the clinical variables. The posterior probabilities from the LCA model were used to create a variable that assigned each participant to their most likely class. This variable was then dummy-coded and used as a predictor variable in a model with the clinical variables entered as dependent variables. All analyses were conducted using Mplus 4.21 [41] and the appropriate weighting variable was used in the second and third phase of the analyses.
Results
Table 1 shows the frequencies of endorsing the psychosis screening items for the PTSD and the non-PTSD samples. The chi-square results show that endorsement of all psychosis items were significantly higher for the PTSD sample than for the non-PTSD sample. Each psychosis variable was regressed on the PTSD variable using binary logistic regression. The resultant odds ratios are reported. This shows that having a diagnosis of PTSD increased the likelihood of endorsing any of the psychosis items between 2.45 and 4.22 times.
The fit statistics for the LCA are presented in Table 2. The 4-class solution was considered to be the best fitting model. The AIC and ssaBIC information statistics were markedly lower for the 4-class solution compared to the 2- and 3-class solutions, and the BSLRT indicates that the 5-class solution is not significantly better than the 4-class solution (and so the 4-class solution should be preferred on the basis of parsimony). The BIC for the 4-class solution is higher than that for the 3-class solution although the difference is small. The entropy value (0.66) indicates acceptable classification of participants. The latent class profile plot is shown in Fig. 1 and the probabilities in Table 3.
Class 1 (N = 94, 16.4%) was characterized by relatively high probabilities of endorsing two delusional indicators (Spy and Poison), three hallucination indicators (visual, olfactory, and tactile) and all PTSD indicators. This class was labeled ‘High Psychosis-High PTSD’. Class 2 was smaller than class 1 (N = 59, 10.3%) and had a similar profile in terms of probability of endorsing the psychosis indicators. The probabilities were lower than class 1 for all the PTSD indicators. This class was labeled ‘High Psychosis-Low PTSD’. Class 3 (N = 214, 37.4%) was the largest class and was characterized by low probabilities of endorsing the psychosis indicators, but high probabilities associated with all the PTSD indicators (similar to class 1). This class was labeled ‘Low Psychosis-High PTSD’. Class 4 (N = 205, 35.8%) was slightly smaller than class 3 and was characterized by low probabilities of endorsing the psychosis indicators, and relatively low probabilities associated with all the PTSD indicators. This class was labeled ‘Low Psychosis-Low PTSD’. It should be noted that references to ‘high’ and ‘low’ are relative rather than absolute.
Table 4 shows the results from the multinomial logistic regression analysis. The demographic and trauma variables were entered as predictor variables, with class membership entered as the dependent variable. Class 4 (Low Psychosis-Low PTSD) was the reference class. Members of class 1 were more likely to be younger and less likely to live alone. They were also more likely to have witnessed an injury or killing. Members of class 2 were less likely to be white and were also less likely to have been educated for more than 16 years. They were more likely to have experienced a fire, flood, or natural disaster, were more likely to have been raped, and to have been neglected as a child. Members of class 3 were not distinguishable from the reference class on the basis of demographic variables. They were, however, more likely to have experienced a range of traumas including, having been to war, having witnessed an injury or killing, and having been raped.
Table 5 shows the results of the logistic regression model. This model used dummy-coded variables to represent three latent classes. Class 4 (Low Psychosis-Low PTSD) was the reference class therefore the odds ratios associated with the other classes indicated the increased likelihood of a CIDI diagnosis compared to class 4. For class 1 there was a significant increase in the likelihood of a diagnosis of alcohol dependence, major depression, bipolar disorder, mania, and generalized anxiety disorder. For class 2 there was a significant increase in the likelihood of a diagnosis of mania and generalized anxiety disorder. For class 3 there was a significant increase in the likelihood of a diagnosis of bipolar disorder and generalized anxiety disorder.
Discussion
This study aimed to test if there was a homogeneous group of individuals who could be described in terms of a psychotic PTSD subtype. It was further predicted that such a group would have increased risks of other comorbid psychiatric diagnoses. It was found that participants with a lifetime diagnosis of PTSD were significantly more likely to endorse seven of the eight CIDI psychosis indicators compared to the non-PTSD group. The rates of endorsement for this group for visual (19%), auditory (19%), and tactile (19%) hallucinations were all high compared to other general population estimates [42, 43]. However, these were similar to the rates for NCS participants who had been victims of childhood neglect or abuse, or molestation under the age of 16 [4]. These findings are consistent with previous research that reported an association between trauma and the experience of psychosis-like experiences [44] or a diagnosis of psychosis [45].
The LCA indicated that there were four homogenous groups. Classes 1 and 4 differed quantitatively across all the variables, and the differences among the other classes were qualitative: the probabilities of endorsement did not increase or decrease uniformly across classes. Two of these groups had relatively high probabilities associated with the psychosis indicators. For class 1 the probabilities of endorsing the first paranoia (Spy) indicator and three of the hallucination (Vis, Aud, and Tact) indicators were high, indeed higher than the probabilities of endorsing some PTSD symptoms (B2, B3, B5, and D5) for classes 2 and 4. This may be due to the intrusive nature of these psychotic indicators. Class 2 had a similar pattern of probabilities to class 1 for each of the psychosis indicators, although the probabilities of endorsing the PTSD symptoms were lower. The high probabilities of endorsement may be attributable to the phenomenological similarity of intrusive thoughts and flashbacks that are indicative of PTSD and hallucinations/delusions [45, 46]. Many studies [14, 45] have found that individuals with PTSD who reported psychotic symptoms had higher levels of psychopathology than those with a diagnosis of either PTSD or psychosis alone. The findings in this study are consistent with previous research, but only for class 1. Individuals in class 1 were significantly more likely to have a diagnosis of alcohol dependence, depression, bipolar disorder, mania, and GAD compared to the Low PTSD-Low Psychosis class (class 4) while the odds ratios for class 2 were significant for mania and GAD only. On the basis of the profile of endorsement of PTSD symptoms and psychosis indicators, classes 1 and 2 could be argued to represent a psychotic PTSD subtype; however, only class 1 displayed elevated levels of overall psychopathology.
The associations between trauma exposure and class membership did not clearly delineate the classes. Table 4 shows that class 1 membership was significantly predicted only by having witnessed someone being badly injured or killed. However, there were three significant trauma predictors for classes 2 and 3.
Overall, class 1 had two of the main features that would be expected in a psychotic PTSD subtype. The class was characterized by a high probability of endorsing the psychosis indicators and had higher levels of psychopathology. However, there was no clear etiological pathway based on trauma exposure that differentiated this class from the other classes. However, class 1 was consistent with previous descriptions of the psychotic PTSD subtype. First, Kastelan et al. [17] reported that the severity of hyperarousal symptoms reported by PTSD diagnosed war veterans was positively associated with psychotic symptoms. In this study, class 1 had high levels of hyperarousal (D4 and D5) compared to the other high psychosis class. Second, Braakman et al. [15] noted that Hispanic and African Americans displayed a higher incidence of PTSD-SP. In this study the ethnicity variable had a significant effect on class 2 only, indicating an increased likelihood of being non-white. Furthermore, class 1 compared to class 2 had high probabilities of endorsing the two avoidance symptoms (C1, C2). Trauma-related avoidance potentially compromises social relations and promotes social isolation. It has also been suggested that this isolation in turn potentially reduces the possibility for reality testing, thus enhancing vulnerability for psychotic experiences [47, 48]. Additionally, it has been proposed that both the intrusive memories and flashbacks that are characteristic of the re-experiencing and the hyperarousal symptom clusters of PTSD, may also constitute chronic stressors which may worsen symptom experience and enhance psychosis vulnerability [45].
This study has provided a symptom-based description of a homogenous group that could be described in terms of a psychotic PTSD subtype. In addition, it has shown that this group has a clinical profile that is significantly poorer to all other participants. However, the study had some limitations and fails to provide sufficient evidence for a valid diagnostic entity. No information on the etiology, course, treatment response, or biological basis for psychotic PTSD was available. In addition there were limitations in terms of the methodology. First, all measurements used in this study are self-reports based on a structured interview rather than clinical assessments. Second, the measurement of psychosis-like symptoms can be confounded by numerous factors, such as respondents misunderstandings, the nature of the question, or normalizing the experience. In addition, it is not easy from lay-interviews to distinguish reports of odd experiences from true psychotic experiences. Third, whilst self-report measures of psychotic experience may be accurate in clinical samples, they may be falsely denied in the general population due to the perceived stigma associated with such experiences [2, 48, 49]. Finally, the inclusion of those with a lifetime diagnosis of psychosis may have provided a useful comparison group for the psychotic PTSD subtype class in terms of comorbid psychiatric diagnoses.
The current findings may have important implications regarding the conceptualization of trauma-related diagnoses. It is notable that the clinical profile of the High PTSD-High Psychosis class was significantly poorer than the remainder of this PTSD sample. If these findings are indeed representative of trauma psychopathology then traumatized individuals diagnosed with PTSD might also experience psychosis related symptoms, and if these individuals in turn exhibit an elevated probability of clinical comorbidity, then it would seem important that clinicians screen for the presence of symptoms that currently lie beyond the diagnostic boundaries of a PTSD diagnosis.
References
American Psychiatric Association (1994) Diagnostic and statistical manual of mental disorders, 4th edn. American Psychiatric Association, Washington
Bebbington PE, Kuipers L (1992) Social causation of schizophrenia. In: Bhugra D, Leff JP (eds) Principles of social psychiatry. Blackwell, Oxford
Janssen I, Krabbendam L, Bak M, Hanssen M, Vollebergh W, de Graaf R et al (2004) Childhood abuse as a risk factor for psychotic experiences. Acta Psychiatr Scand 109:38–45
Shevlin M, Dorahy M, Adamson G (2007) Childhood traumas and hallucinations: an analysis of the National Comorbidity Survey. J Psychiatr Res 41:222–228
Shevlin M, Dorahy MJ, Adamson G (2007) Trauma and psychosis: an analysis of the National Comorbidity Survey. Am J Psychiatry 164:166–169
Elklit A, Shevlin M (2010) Female sexual victimization predicts psychosis: a case-control study based on the Danish registry system. Schizophr Bull. doi:10.1093/schbul/sbq048
Hamner MB, Frueh BC, Ulmer HG, Arana GW (1999) Psychotic features and illness severity in combat veterans with chronic posttraumatic stress disorder. Biol Psychiatry 45:846–852
Hamner MB, Fossey MD (1993) Psychotic symptoms associated with posttraumatic-stress-disorder. Neuropsychopharmacology 9:S121–S122
Hamner MB (1997) Psychotic features and combat-associated PTSD. Depress Anxiety 5:34–38
Morrison APSC, Read J, Turkington D (2005) Trauma and psychosis: theoretical and clinical implications. Acta Psychiatr Scand 112:327–329
Seedat S, Stein MB, Oosthuizen PP, Emsley RA, Stein DJ (2003) Linking posttraumatic stress disorder and psychosis—a look at epidemiology, phenomenology, and treatment. J Nerv Ment Dis 191:675–681
Sautter FJ, Brailey K, Uddo MM, Hamilton MF, Beard MG, Borges AH (1999) PTSD and comorbid psychotic disorder: comparison with veterans diagnosed with PTSD or psychotic disorder. J Trauma Stress 12:73–88
Ivezic S, Bagaric A, Oruc L, Mimica N, Ljubin T (2000) Psychotic symptoms and comorbid psychiatric disorders in a Croation combat related posttraumatic stress disorder sample. Croat Med J 41:179–183
Sareen J, Cox BJ, Goodwin RD, Asmundson GJG (2005) Co-occurrence of posttraumatic stress disorder with positive psychotic symptoms in a nationally representative sample. J Trauma Stress 18:313–322
Braakman MH, Kortman FAM, van den Brink W (2009) Validity of ‘posttraumatic stress disorder with secondary psychotic features’: a review of the evidence. Acta Psychiatr Scand 199:15–24
David D, Kutcher GS, Jackson EI, Mellman TA (1999) Psychotic symptoms in combat-related posttraumatic stress disorder. J Clin Psychiatry 60:29–32
Kastelan A, Franciskovic T, Moro L, Roncevic-Grzeta I, Grkovic J, Jurcan V et al (2007) Psychotic symptoms in combat-related post-traumatic stress disorder. Mil Med 172:273–277
Kessler RC (1994) The National Comorbidity Survey of the United States. Int Rev Psychiatry 6:365–376
American Psychiatric Association (1987) Diagnostic and statistical manual of mental disorders, 3rd edn, revised (DSM-III-R). American Psychiatric Association, Washington
World Health Organization (1990) Composite International Diagnostic Interview (CIDI), version 1.0. World Health Organization
Simms LJ, Watson D, Doebbeling BN (2002) Confirmatory factor analyses of posttraumatic stress symptoms in deployed and nondeployed veterans of the Gulf War. J Abnorm Psychol 111:637–647
Elklit A, Armour C, Shevlin M (2010) Testing alternative factor models of PTSD and the robustness of the dysphoria factor. J Anxiety Disord 24:147–154
Armour C, Shevlin M (2010) Testing the dimensionality of PTSD and the specificity of the dysphoria factor. J Loss Trauma 15:11–27
Ford JD, Elhai JD, Ruggiero KJ, Frueh BC (2009) Refining posttraumatic stress disorder diagnosis: evaluation of symptom criteria with the National Survey of Adolescents. J Clin Psychiatry 70:748–755
Spitzer RL, First MB, Wakefield JC (2007) Saving PTSD from itself in DSM-V. J Anxiety Disord 21:233–241
Morrison AP, Frame L, Larkin W (2003) Relationships between trauma and psychosis: a review and integration. Br J Clin Psychol 42:331–353
Spitzer RL, Williams JBW, Gibbon M, First MB (1992) The Structured Clinical Interview for DSM-III-R (SCID) I: history, rationale, and description. Arch Gen Psychiatry 49:624–629
Kessler RC, Borges G, Walters EE (1999) Prevalence of and risk factors for lifetime suicide attempts in the national comorbidity survey. Arch Gen Psychiatry 56:17–626
Wittchen HU, Robins LN, Cottler LB, Sartorius N, Burke JD, Regier D (1991) Cross-cultural feasibility, reliability and sources of variance of the Composite International Diagnostic Interview (CIDI). Br J Psychiatry 159:645–653
Cottler LB, Robins LN, Grant BF, Blaine J, Towle LH, Wittchen HU et al (1991) The CIDI-core substance-abuse and dependence questions—cross-cultural and nosological issues. Br J Psychiatry 159:653–658
Wacker HR, Battegay R, Muellejans R, Schloesser C (1990) Using the CIDI-C in the general-population. Psychiatry: A World Perspective 900:138–143
Janca A, Robins LN, Bucholz KK, Early TS, Shayka JJ (1992) Comparison of Composite International Diagnostic Interview and Clinical DSM-III-R Criteria Checklist Diagnoses. Acta Psychiatr Scand 85:440–443
Schafer JL, Graham JW (2002) Missing data: our view of the state of the art. Psychol Methods 7:147–177
Akaike H (1987) Factor analysis and AIC. Psychometrika 52:317–332
Scwartz G (1978) Estimating the dimension of a model. Ann Stat 6:461–464
Sclove SL (1987) Application of model selection criteria to some problems in multivariate analysis. Psychometrika 52:333–343
McLachlan GJ, Peel D (2000) Finite mixture models. New York, Wiley-Interscience
Ramaswamy V, DeSarbo W, Reibstein D, Robinson W (1993) An empirical pooling approach for estimating marketing mix elasticities with PIMS data. Marketing Sci 12:103–124
Nylund KL, Asparoutiov T, Muthen BO (2007) Deciding on the number of classes in latent class analysis and growth mixture modeling: a Monte Carlo simulation study. Struct Equ Model 14:535–569
Muthen BO, Muthen L (1998–2007) Mplus Users Guide, 5th edn. Muthen & Muthen, Los Angeles
Scott J, Welham J, Martin G, Bor W, Najman J, O’Callaghan M et al (2008) Demographic correlates of psychotic-like experiences in young Australian adults. Acta Psychiatr Scand 118:230–237
Johns LC, Hemsley D, Kuipers E (2002) A comparison of auditory hallucinations in a psychiatric and non-psychiatric group. Br J Clin Psychol 41:81–86
Bebbington PE, Bhugra D, Brugha T, Singleton N, Farrell M, Jenkins R et al (2004) Psychosis, victimisation and childhood disadvantage—evidence from the second British National Survey of Psychiatric Morbidity. Br J Psychiatry 185:220–226
Mueser KT, Essock SM, Haines M, Wolfe R, Xie HY (2004) Posttraumatic stress disorder, supported employment, and outcomes in people with severe mental illness. CNS Spectr 9:913–925
Shaw K, McFarlane AC, Bookless C, Air T (2002) The aetiology of postpsychotic posttraumatic stress disorder following a psychotic episode. J Trauma Stress 15:39–47
Bendall S, McGorry P, Krstev H (2006) The trauma of being psychotic: an analysis of posttraumatic stress disorder in response to acute psychosis. In: Larkin W, Morrison AP (eds) Trauma and psychosis: new directions for theory and therapy. Routledge, London, pp 58–74
Cresswell CM, Kuipers L, Power MJ (1992) Social networks and support in long-term psychiatric-patients. Psychol Med 22:1019–1026
Bebbington PE, Kuipers L (1994) The social management of long standing schizophrenia: the deployment of service resources. Clinician 12:17–29
Butzlaff RL, Hooley JM (1998) Expressed emotion and psychiatric relapse—a meta-analysis. Arch Gen Psychiatry 55:547–552
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Shevlin, M., Armour, C., Murphy, J. et al. Evidence for a psychotic posttraumatic stress disorder subtype based on the National Comorbidity Survey. Soc Psychiatry Psychiatr Epidemiol 46, 1069–1078 (2011). https://doi.org/10.1007/s00127-010-0281-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00127-010-0281-4