Abstract
Community-based surveys have found that otherwise-well individuals endorse items related to psychotic symptoms [1–3]. For example, a large population-based survey in Australia found that 11.7% of respondents endorsed one or more items designed to identify delusion-like experiences [4]. Most of the community-based studies thus far have used questionnaire items originally designed to screen for psychotic disorders, but more recently self-report instruments have been developed to specifically explore delusion-like experiences. One of the most widely used of these instruments is the Peters Delusional Inventory (PDI) [5], which uses wording designed to capture quasi-delusional as well as putative delusional beliefs (e.g. ‘Do you ever feel as if other people can read your mind?’). Community-based studies using the PDI have confirmed that delusion-like experiences are common in otherwise healthy individuals [6], [7]. Understanding the demographic, phenomenological and neurobiological correlates of delusion-like experiences in the healthy population may provide clues to the mechanisms of action underlying these isolated symptoms.
A number of demographic features have been associated with an increased likelihood of endorsement of isolated psychotic symptoms, including urban residence, lower income, lower educational achievement, younger age and living alone [8]. Using questionnaire items designed to screen for psychosis, Scott et al. confirmed that younger individuals were more likely to report delusion-like experiences [4]. Other factors associated with delusion-like experiences in that Australian study included male sex, unemployment, migrant status (but only those from non-English speaking backgrounds), never-married or divorced/separated status, lower socioeconomic status, and cannabis or alcohol dependence [4], [9].
With respect to the PDI, the association between younger age and higher total PDI scores has been a relatively consistent finding [6], [7], [10]. Neurophysiological studies have also identified that those with higher PDI scores have (i) altered visual scan paths [11]; (ii) altered selective attention to threatening faces [12]; and (iii) altered attribution of self-generated speech [13]. Those with higher scores on the PDI are also more likely to endorse items related to hallucinations [10]. Curiously, in a study of unaffected first-degree relatives of patients with either schizophrenia or bipolar disorder, relatives of those patients with prominent delusions were more likely to have higher PDI scores compared with relatives of patients with fewer delusions [14].
The aim of the present study was to explore these issues using data collected as part of the Brisbane Psychosis Study, which included a wide range of risk factors previously associated with schizophrenia. In particular, we predicted that higher PDI total scores would be associated with male sex, younger age, migrant status, family history of schizophrenia, or family history of any mental illness, increased paternal age [15], and higher scores of various measures of minor physical anomalies [16].
Methods
The Brisbane Psychosis Study was a case–control study designed to examine a wide range of risk factors for psychosis. The controls used in the Brisbane Psychosis Study were the subjects examined in the current study. These subjects were carefully screened to exclude psychotic disorders (but not other psychiatric conditions). The full methodology of this study is described elsewhere [16], [17]. The subjects, aged between 18 and 74 years, were recruited via newspaper and community advertisements. Psychotic disorders were assessed by trained clinical psychologists or research nurses with the Diagnostic Interview for Psychosis [18]. This instrument, which is based on the Operational Criteria for Psychosis (OPCRIT), is a 90-item checklist linked to a computer multidiagnostic algorithm [19]. The OPCRIT includes two broad screening items related to (i) family history of schizophrenia, and (ii) family history of any psychiatric condition, in first- or second-degree biological relatives. Additional items related to paternal age and migrant status were included in a structured questionnaire.
In order to assess delusion-like experiences the participants were given the 40-item version of the PDI [5]. In the present study we used the main total score (range 0–40; higher scores indicating more delusion-like experiences). Full details of the assessment of minor physical anomalies and quantitative measures related to the head and face are provided in detail elsewhere [16]. In the present study we examined the association with the PDI total scores versus three summary measures: (i) total count of qualitative minor physical anomalies; (ii) derived factor score related to the size of cranio-facial quantitative measures; and (ii) derived factor score related to the shape of cranio-facial quantitative measures.
When screening the healthy controls in order to exclude psychosis, we also included 10 items related to hallucinations (auditory, visual, and olfactory). Six of these items were drawn from the Schizotypal Personality Questionnaire [20], and one from the Chapman and Chapman Psychosis Proneness scale [21]; and three new items were included (full details of these items are available from the authors). In addition, we included the Communication Awareness Scale [22], which assesses the frequency of awareness of positive thought disorder.
Because the PDI total scores were not normally distributed, we used non-parametric statistics (Wilcoxon rank–sum test to compare the scores for dichotomous variables; Spearman correlation ρ for continuous variables). For demographic variables showing a significant correlation, we also undertook a secondary analysis that controlled for sex and age between the variables of interest. The PDI total scores were split into tertiles. Logistic regression was used to examine the influence of the predictor variables, when adjusted for age and sex on PDI total score tertiles.
The study was approved by the Wolston Park Hospital Ethics Committee, and all subjects provided written, informed consent.
Results
The sample included 303 individuals (138 women) with a mean age of 40.6 years (SD = 13.0). The total PDI score ranged from 0 to 26 (maximum possible score was 40). The mean, median and modal values were 4.95, 4 and 0, respectively. The distribution of this variable was half-normal. The tertile cut-offs were ≤2; 3–6; and ≥7.
There was no significant difference in PDI total score by sex (W = − 0.85, p = 0.40). There was a trend-level association between younger age and higher PDI total score (spearman ρ = − 0.11, p = 0.051). There was no significant association between the PDI total and (i) family history of schizophrenia (26 subjects, 9%); (ii) paternal age; nor (iii) migrant status (data not shown). Those with a family history of any mental illness (39% of the sample) had significantly higher PDI scores compared to those without such a history (median PDI total = 5 and 3, respectively; W = 2.85, p = 0.004). The association between the PDI total scores and family history of any mental illness persisted when adjusted for age and sex (lowest vs highest PDI total tertile odds ratio = 2.80, 95% confidence interval = 1.54–5.06).
There were significant positive correlations for PDI scores and (i) Communication Awareness Scale score (ρ = 0.36, p < 0.001); and (ii) total hallucinations score (ρ = 0.52, p < 0.001). There was no significant association between PDI total score and any of the three measures related to minor physical anomalies (data not shown).
Discussion
Individuals who were free of psychotic disorders frequently endorsed items related to delusion-like experience. Indeed, more than one-third of the sample endorsed seven or more items on the PDI. We report, for the first time, that individuals who reported a positive family history for any psychiatric disorder had a small but significantly higher total PDI score. While the present study cannot determine the nature of the family history of any psychiatric condition, it seems reasonable to assume that the subjects are mostly reporting the presence of a family history of high prevalence disorders such as depression and anxiety disorders. Only 26 of the 294 subjects reported a family history of schizophrenia, and thus the present study lacked the power to confidently exclude an association with this variable.
It may be that the PDI reflects a genetic predisposition to high prevalence disorders such as depression or anxiety disorders. Other studies have found that high PDI scores predict incident depression at 12 month follow up [23] and that high PDI scores are associated with trait neuroticism [10].
In keeping with previous studies [10], there was a significant positive correlation between PDI scores and the endorsement of hallucination items. We report, for the first time, that self-reported impaired communication was also associated with higher PDI scores. Thus, non-psychotic individuals who report delusion-like experience also report other features of psychosis including hallucinations and awareness of positive thought disorder. There were no significant associations between PDI total score versus minor physical anomaly scores, paternal age, migrant status and sex.
Although the present study did not detect a significant association between age and total PDI score, the finding was at a trend level (p = 0.051), and was in keeping with our directional hypothesis. Neurodevelopmental models of psychosis draw attention to critical periods in brain development, especially the ‘neuronal pruning’ that occurs in late adolescence and early adulthood [24]. Apart from developmentally regulated biological mechanisms, the trend towards higher PDI scores in younger subjects may also reflect factors related to learning and decision making (e.g. older subjects becoming more discriminating in the face of experience), and/or cultural factors (e.g. belief in telepathy may reflect the influence of contemporary media themes).
The present study had several limitations. The sample size was too small to confidently identify associations between PDI scores and several of the variables of interest (e.g. age, family history of schizophrenia). Furthermore, the variable related to family history of any psychiatric history did not provide sufficient information about the nature of the disorder. Apart from the finding that people with psychotic disorders have higher scores on the PDI [25], [26], there is also evidence that higher PDI scores are found in those with mood disorders [8], and in those with cannabis and alcohol use [27]. Information on cannabis and other substance use was not available for the present sample, thus we cannot exclude the possibility that the association between a family history of any psychiatric condition and higher PDI scores could have been mediated by this variable. Similarly, the presence of other psychiatric disorders (e.g. depression, anxiety) was not examined. We are currently exploring these particular issues in a large population-based sample.
Regardless of these issues, this study confirms that delusion-like experiences are relatively common in otherwise-well individuals. Furthermore, endorsement of delusion-like experiences is associated with endorsement of items related to hallucinations and thought disorder, and a family history of mental illness. These findings suggest that some features of the broader phenotype of psychosis can be detected in attenuated form in otherwise-well individuals. Understanding the continuum of psychotic symptoms in the general population may provide clues as to the pathogenesis of psychotic disorders [28].