Trauma exposure, PTSD and psychotic-like symptoms in post-conflict Timor Leste: an epidemiological survey

Knowledge has accrued about the prevalence of psychotic-like symptoms such as hallucinations, paranoia and strange experiences in the general community [1–3]. The relatively high rates of psychotic-like experiences at a community-wide level suggests a distinction between these phenomena and clinical psychotic disorders, which have much lower prevalence rates. A clinical diagnosis of a psychotic disorder requires the simultaneous occurrence and persistence of symptoms, whereas persons experiencing isolated psychotic-like symptoms appear to be more common and are less likely to attract clinical attention [1, 4].

Past studies suggest that psychotic-like symptoms are associated with being female, low levels of social support, alcohol [5] and cannabis use [6], and a family history of mental illness [5, 6]. Psychotic-like symptoms appear to be more prevalent in some ethnic groups [7], but available data are restricted to minorities living in developed countries.

Trauma exposure appears to be associated with psychotic-like symptoms [8–14], with some studies indicating a dose-effect relationship [2, 15]. In addition, posttraumatic stress disorder (PTSD) may mediate the relationship between trauma and psychotic-like symptoms [13].

The relationship between human rights trauma and psychotic symptoms has attracted attention for several decades. Observations following WWII suggested that psychotic-like symptoms and frank forms of psychosis were common amongst survivors of the Holocaust, although selection biases could have influenced these findings [16]. Since then, clinicians in the field of traumatic stress have argued that gross forms of human rights violations can lead to a fundamental loss of trust amongst survivors, resulting in a tendency towards suspicion and even paranoia, phenomena that are not encompassed by contemporary criteria for PTSD [17].

Little is known about the epidemiology of psychotic-like symptoms in low-income countries that are culturally distinct from nations of the west. In particular, there is a dearth of evidence about the prevalence and correlates of psychotic-like symptoms amongst populations exposed to oppression and prolonged conflict. Cultural factors need to be considered in interpreting the reporting of psychotic-like symptoms in settings such as Timor Leste, where belief in supernatural phenomena such as spirit possession is normative. The Psychosis Screening Questionnaire, the measure used in the present study, explicitly inquires whether the experience of psychotic-like symptoms reported are regarded as abnormal by other members of the society. Nevertheless, it remains possible that traditional beliefs, for example concerning curses and communication with spirits, may confound the reporting of psychotic-like symptoms. Even so, the presence of symptoms may still be clinically relevant, especially if they are associated with disability.

The present study is the first to establish the epidemiology of psychotic-like symptoms in a low-income, post-conflict society, Timor Leste. We hypothesized that psychotic-like symptoms would be common, and associated with exposure to the cumulative traumas of conflict. We also postulated that PTSD would mediate the relationship between trauma and psychotic-like symptoms.

Five percent of participants exceeded the 2.0 threshold for PTSD (Table 2), and 12.3% were PSQ screen-positive cases (Table 1). Table 1 shows that paranoia (5.4%) and strange experiences (5.4%) were the most prevalent psychotic-like symptoms. Only two respondents (0.2%) reported hypomania.

Table 4 indicates that both PTSD (OR: 2.1, CI: 1.2-3.7, p=0.0105) and PSQ screen-positive cases (OR: 1.6, 95% CI: 1.1-2.5, p=0.019) were disabled with no interaction effect.

Figure 1 shows the results of the mediation model. PTSD partially mediated the relationship between PTEs and PSQ screen-positive cases (effect = 0.033, 95% CI: 0.014–0.060). However, the direct effect of PTEs was comparatively greater (0.063, 95% CI: 0.034 - 0.093), indicating that the partial mediating effect of PTSD accounted for approximately 34% of the overall relationship of PTEs with PSQ screen-positive caseness.

To our knowledge, this is the first study documenting the population prevalence of PSQ screen-positive cases in a low-income, post-conflict society. There was a substantial prevalence of such cases, more than twice the rate for PTSD. Comparisons with previous studies need to be made with caution because of differences in sampling and indices of psychotic-like symptoms measured. Nevertheless, it is noteworthy that the rate of PSQ screen-positive cases in the British National Morbidity Survey was 4% [2]. Other studies using the same instrument in western countries have found prevalence rates ranging between 4 and 12% [34], although higher rates have been reported in studies using other measures [3]. High rates of PSQ screen-positive cases have been documented amongst ethnic minorities living in western settings. For example, amongst Romanian immigrants living in adverse social circumstances in Italy, 19% were PSQ screen-positive [34]. Similarly, a study of five ethnic groups in the UK found a higher rate of PSQ screen-positive cases in Afro-Caribbeans (12%) than in the majority Caucasian population (6%) [7]. The common characteristic of these groups is their exposure to socio-economic disadvantage, suggesting that adverse environmental conditions may be relevant in the genesis of psychotic-like symptoms.

Our findings suggested a relationship between increasing levels of PTE exposure and PSQ screen-positive cases, even after adjusting for socio-demographic variables. Persons with high levels of trauma exposure had invariably experienced extensive human rights violations such as torture, incarceration and other forms of politically motivated abuses. In that respect, our study echoes the findings of the post-WWII literature, in which associations were noted between gross forms of human rights violation and psychotic-like symptoms, for example, amongst survivors of concentration camps [16].

Recent clinical observations have supported an association between severe PTSD symptoms and psychotic-like experiences including paranoia and hallucinations [35, 36]. Our analysis provided some support for the role of PTSD in mediating the link between PTEs and PSQ screen-positive cases. Nevertheless, the direct path was stronger, suggesting that, in the Timor Leste setting, PTSD played a subordinate role in mediating the effect.

The presence of psychotic-like symptoms in survivors of human rights trauma adds to the evidence supporting the recognition of a more complex form of traumatic stress [37]. Past formulations have included Disorders of Extreme Stress Not Otherwise Specified (DESNOS), considered but not included in the Diagnostic and Statistical Manual of Mental Disorders version 4 (DSM-IV) [37], and the ICD-10 category of Enduring Personality Change After Catastrophic Experiences (EPCACE) [38]. Both constellations include symptoms of mistrust, hostility and alienation, characteristics akin to paranoia and other psychotic-like symptoms.

It seems likely that exposure to prolonged persecution and conflict can attenuate the individual’s sense of trust and security, in a manner that may nevertheless be adaptive. In a minority, however, these responses may create or magnify the susceptibility to paranoia and other psychotic-like symptoms, particularly when ongoing communal divisions continue to foster fear and suspicion. The challenge is to define more clearly the boundaries between normative responses of mistrust and pathological reactions wherein suspicion and fear are transformed into frank and potentially disabling psychotic-like symptoms. In the first paper in this series, we reported estimates of clinical psychotic disorders of 1.35% for the sample as a whole, based on the Structured Clinical Interview for the DSM-IV (SCID) conducted by medical personnel amongst a subset of respondents [18]. This rate clearly is much lower than the prevalence (12.3%) of PSQ screen-positive cases reported here. Hence, as already recognised [2, 18], there is a large discrepancy between PSQ screen-positive cases and the rates of clinical psychotic disorders in the population. Clinical diagnoses require the persistence of a combination of symptoms over a specified time period. In addition, the threshold for endorsing symptoms when applying screening instruments invariably is lower than that imposed by clinicians employing comprehensive clinical assessments to make formal diagnoses. In the present study, only a small number of respondents completed both the PSQ and the SCID, precluding a systematic comparison of the two forms of assessment [18]. Hence, uncertainty remains about the clinical significance of reported psychotic-like symptoms based on instruments such as the PSQ.

The PSQ explicitly inquires whether the psychotic-like experiences endorsed are regarded as alien by others living in the same setting. Nevertheless, cultural factors could still be influential in the expression of symptoms. The traditional animistic faith remains influential in Timor Leste, a system that includes beliefs about curses, being bewitched and communicating directly with ancestors. It is possible that the boundaries between these normative, cultural experiences and psychotic-like symptoms become blurred in states of distress. However, in the univariate analyses, the young, the educated and those living in the city were more likely to be PSQ screen-positive, and yet those sectors of the population are more likely to have adopted a modern, western-based lifestyle no longer grounded in the traditional belief systems. More detailed inquiry is needed, therefore, to assess the influence of culture and religious beliefs on the reporting of PSQ symptoms in Timor Leste and other transcultural settings.

The limitations of the study need to be acknowledged. The study was cross-sectional, cautioning against drawing causal inferences, for example, between PTE exposure and psychotic-like symptoms [12, 39]. The traumas measured were limited to the experiences of mass conflict and natural disasters. As noted, we omitted the item for rape, and this may have attenuated the link between trauma and PSQ screen-positive cases. Nor did we assess early trauma such as childhood sexual abuse, or personality vulnerabilities that may be relevant to the development of psychotic-like symptoms. We do not know what the PSQ screen-positive rate was prior to the conflict in Timor Leste. It also is possible that psychotic-like symptoms preceded trauma exposure in some instances. Nevertheless, the chronology recorded for the list of trauma experiences suggests that the vast majority of these events occurred during the Indonesian occupation (1975-1999) and the humanitarian emergency (1999), that is, at least four years prior to participants completing the PSQ.

Consistent with past research, the HTQ community threshold for PTSD would have identified a wider range of persons as symptomatic than a structured interview assessment conducted by clinicians applying DSM-IV criteria [18, 27]. Even so, the prevalence of PTSD assessed by the HTQ was relatively low in comparison to other post-conflict settings, although within the range of past epidemiological studies undertaken around the world [24]. In that regard, it is noteworthy that conditions in Timor Leste at the time of the study were consistent with those generally associated with lower rates of PTSD in post-conflict settings, particularly the removal of the source of the terror with the expulsion of the Indonesian military, the restoration of peace by the UN, and the passage of some four years since the conflict [24]. An additional consideration is that some studies have suggested that the presence of PTSD may result in increased reporting of past trauma, although that assertion has been challenged by a recent re-analysis of the trauma histories of combat veterans [40].

In summary, our findings support the view that post-conflict mental health research needs to extend its focus to psychiatric outcomes beyond PTSD and depression [24, 41]. Our results suggest the value of pursuing the assessment of psychotic-like symptoms and their association with trauma in post-conflict settings. Further research also is needed to examine the implications of psychotic-like symptoms in post-conflict settings such as Timor Leste, the extent to which cultural factors and ongoing tensions in the community influence endorsement of these phenomena, and the relationship of symptoms to disabling forms of clinical psychosis.