Statistics from Altmetric.com
The study was undertaken in order (1) to examine the association between DLE and (a) broadly defined anxiety disorders and (b) MDDs; (2) to explore the association between DLE and a range of specific anxiety disorders and (3) to examine if severity of MDD influenced the risk of endorsement of DLE.
Having a lifetime diagnosis of either any anxiety disorder or MDDs was significantly associated with the endorsement of DLE.
The association was found for each of the main anxiety disorders when examined separately.
There was a dose–response relationship between increasing severity of MDD and higher odds of DLE endorsement.
Strengths and limitations of this study
The data were drawn from the nationally representative sample from the Australia general population.
There is now robust evidence indicating that hallucinations and delusional-like experiences (DLE) are common in the general population. In recent years, the field has focused on the demographic and clinical correlates of hallucinations and DLE.1–10 Of particular interest, there is a growing body of evidence reporting an association between DLE endorsement and common mental disorders, such as anxiety disorders and major depressive disorder (MDD). For example, panic attacks during adolescence were significantly associated with increased levels of DLE among young adults.11 In the NEMESIS study, subjects with obsessive–compulsive symptoms were more likely to develop incident psychotic symptoms 3 years later.12 Conversely, a Swiss-based cohort reported that young adults with psychotic-like experiences were significantly more likely to later develop common mental disorders, such as anxiety disorders and MDD.13 A German community-based study found an association between social phobia, social anxiety and DLE,14 while a US primary-care-based sample reported that those who reported psychotic-like experiences were more likely to have generalised anxiety disorders and panic disorders.15
Trauma exposure with or without post-traumatic stress disorder has been associated with DLE.7 Several Australian studies10 16 have found significant associations between DLE and broadly defined anxiety disorders; however, to date, these studies did not report on subtypes of anxiety disorders. In light of the evidence linking DLE with a wide range of different types of anxiety disorders, the evidence suggests that DLE are non-specifically associated with anxiety disorders.
With respect to depression, several studies have found that individuals with depression are significantly more likely to endorse DLE.9 10 16 17 Studies also show that DLE requiring clinical care were progressively more likely to occur with greater levels of affective dysregulation (depressive symptoms and hypomanic symptoms).18 Importantly, there was a significant association between severity of depressive symptoms and persistence of psychotic symptoms.
While longitudinal studies are required to explore the temporal sequence between depression, anxiety and DLE, we had the opportunity to replicate our previous findings with respect to the cross-sectional association between DLE and (1) broadly defined anxiety disorders and (2) MDD.10 Based on our previous studies, we predicted that those with anxiety disorder or major depression disorder would be more likely to endorse DLE. In addition, we were able to explore the association between DLE and a range of specific anxiety disorders. Furthermore, we were able to examine if severity of MDD influenced the risk of endorsement of DLE—we predicted that those with more severe MDD would be more likely to endorse DLE compared with those with milder forms of MDD.
The data were drawn from the 1997 National Survey of Mental Health and Wellbeing conducted in Australia by the Australian Bureau of Statistics from a representative sample (random stratified multistage area sampling) of persons living in private dwellings in all States and Territories of Australia. Details of the survey methodology were published elsewhere.19 In brief, approximately 13 600 private dwellings were initially selected with one person aged 18 years or older from each dwelling invited to participate. In total, 10 641 individuals participated in the survey, representing a response rate of 78%. Interviews were carried out by trained interviewers from the Australian Bureau of Statistics, a statutory body responsible for conducting such surveys using ethical protocols that include written informed consent.
Assessment of DLE and DSM-IV diagnoses
Mental disorders were assessed by a modified version of the Composite International Diagnostic Interview (CIDI),20 which yielded diagnoses of Diagnostic and Statistical Manual of Mental Disorder, Fourth Edition (DSM IV) disorders. Briefly, within the CIDI, there are three items related to identifying individuals who may be psychotic (G Items: ‘screening items’). For those who endorsed the screen item, a follow-up item was used to further explore the delusional-like nature of the experiences (‘probe items’). Full details of the screen and probe items are provided in online appendix 1. The items covered the following features of psychotic disorders: delusions of control, thought interference and passivity (question 1 and 1a); delusions of reference or persecution (question 2 and 2a) and grandiose delusions (question 3 and 3a). There was no item to assess hallucinations.
Based on CIDI-derived DSM-IV criteria, we identified subjects who had lifetime diagnoses of (1) an anxiety disorder and (2) MDD. Anxiety disorders included panic disorder with or without agoraphobia, social phobia, generalised anxiety disorder, obsessive–compulsive disorder and agoraphobia without panic disorder. For those with MDD, allocation to subtypes was based on the total number of particular ‘depressive’ symptoms with the duration of at least 2 weeks. Full details of the symptom list and related rules to deal with multiple episodes can be found in the full report.21 In brief, mild MDD was characterised by the presence of at least four symptoms, moderate MDD with at least six symptoms and severe MDD with at least eight symptoms. These subtypes of MDD were mutually exclusive.
To ascertain trauma exposure, the CIDI elicits responses from 10 questions pertaining to past exposure to traumatic events. Details of the trauma variables have been published previously by our group.7 8 In keeping with our previous analyses,1–10 individuals who screened positively for schizophrenia (ie, respondents who reported ‘Yes’ to the item ‘Had been told at any time by a psychiatrist that they had schizophrenia’) were excluded from the analyses (n=87), leaving a total of 10 554 subjects for this study.
To examine the association between DLE and both anxiety disorders and MDD, logistic models were fitted to the data while adjusting for various confounding factors. Because sex and age are associated with DLE,9 10 22 we included these as covariates in the main analyses. In keeping with our previous studies, we included a range of CIDI-derived potential confounding variables in model 2. These include substance misuse,23 marital status and migrant status,24 educational status, employment status and family income, and trauma exposure.2 6–8 As comorbidity frequently occurs between anxiety disorders and MDD, we also adjusted for the presence of the other psychiatric diagnoses under investigation (ie, the association between MDD and DLE was adjusted for the presence of anxiety disorders, and the association between anxiety disorders and DLE was adjusted for the presence of MDD).
For secondary analyses (a sensitivity analysis), we repeated the main analyses excluding the second screen items (‘Have you ever had a feeling that people were too interested in you?’) because clinical experience suggests that this is a common experience in social anxiety.
The sample was weighted to adjust for differential probabilities of selection within households, oversampling of population subgroups and non-response to match census population distribution on a number of geographic and socio-demographic variables. The initial weights were calibrated against known population estimates. Replicate weight variables were developed using the Jack-knife procedure of replication (ie, the analysis was repeated after one subject was dropped and then the SE was derived from the distribution of results from all ‘minus one’ resamples).25 Analyses were performed using Proc Surveylogistic,26 which is designed to analyse complex survey sample using SAS (V.9.3; SAS Institute). χ2 Test-for-linear trend was used to assess dose–response relationships between the exposure variables and DLE.
Of the 10 554 subjects surveyed, 11.6% (n=1276) positively endorsed one or more DLE items (table 1). There was a weak effect of women being more likely to endorse DLE than men (OR 1.05, 95% CI 1.04 to 1.05). The prevalence of lifetime diagnosis of any anxiety disorder was 4.9% (n=580), and the prevalence of lifetime depressive disorders was 5.3% (n=651).
As predicted, the main analyses showed that those with any anxiety disorder and participants who had lifetime diagnosis of MDD were significantly more likely to endorse DLE. Those with anxiety disorders were two to three times more likely to endorse both DLE screen and probe items (table 2), and those with a diagnosis of MDD were also two to three times more likely to endorse DLE screen and probe items.
Concerning the subtypes of anxiety disorders, each disorder was significantly associated with DLE screen items, and there were no marked differences in the effect sizes between the different disorders (table 3). There was a dose–response relationship between the severity of the MDD and DLE in which severe depression showed twice the odds of endorsement of DLE screen items compared with a diagnosis of mild MDD with a significant linear trend (χ2=44.19, p<0.0001). Broadly similar (but less precise) associations were also found for probe items.
In the secondary analysis, when we conducted the models using two DLE items (G1 and G3), the pattern of significant association for major anxiety and depressive disorders remained unchanged (data not shown).
Individuals with a lifetime diagnosis of MDD or an anxiety disorder were significantly more likely to report DLE compared with those without these disorders. We found that each subtype of anxiety disorder was associated with DLE, and there were no marked differences in the effect sizes for these associations (the CIs around these associations overlapped). Based on this same sample, we have previously demonstrated that trauma exposure without post-traumatic stress disorder was associated with DLE.7 Our new findings add additional weight to the conclusion that a range of disorders with prominent anxiety symptoms are associated with DLE.
As predicted, there was also a dose–response relationship between severity of MDD and DLE. All associations remained significant when adjusted for associated comorbidity with anxiety, alcohol and illicit substance misuse and any traumatic life events, indicating that the associations are independent of comorbid psychiatric illnesses and selected environmental and demographic risk factors.
The mechanisms linking DLE with anxiety disorder and MDD remain unclear. However, there is evidence to suggest that shared familial factors may contribute to these findings.9
In the current study, we were not able to examine the temporal sequence between the variables of interest—for example, we do not know if anxiety or depressive symptoms preceded the onset of DLE or vice versa. Unfortunately, the DLE have no information about age of onset nor presence during the past 12 months. Thus, while the CIDI has some information about the age of onset and the presence of the disorder in the past year, the lack of comparable data for the DLE compromises out ability to infer temporal sequence. Longitudinal studies will be required to explore this particular research question. The reliance on lifetime measures of both DLE and mental disorders is also problematic, as it is known that respondents tend to under-report true lifetime prevalence estimates.27 While the interviewers were trained, the diagnoses of MDD and anxiety disorders were not validated by clinical assessment. However, the CIDI is generally regarded as having good psychometric properties for common mental disorders.28 Comorbidity between anxiety disorders and MDD is common, and while we included adjustments in the models to attempt to account for this feature, the complex nature of the relationships between DLE, MDD and anxiety disorders could reduce the accuracy of the ORs.29 We had a small number of screen and probe items to measure DLE, and there were no items for hallucinations. However, previous general population studies have found a strong association between the presence of DLE and hallucinations.24 30–32
There is now robust and consistent evidence indicating that those with anxiety disorders and MDD have an increased risk of DLE. For example, clinicians involved in the care of those with primary diagnoses of anxiety disorder or depression may not routinely enquire about DLE. In light of the association between DLE and suicidal ideation/behaviour,3 the presence of these experiences may suggest that clinical care plans place greater emphasis on the detection and management of suicidal ideation. A recent study based on adolescents found that most individuals (57%–80% depending on age) who reported psychotic-like experiences (eg, hallucinations and/or DLE) had at least one diagnosable non-psychotic psychiatric disorder.33 We agree with these authors, who note that psychotic symptoms appear to be important risk markers for a wide range of non-psychotic mental health disorders.
Review history and Supplementary material
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online Appendix 1
To cite: Saha S, Scott J, Varghese D, et al. Anxiety and depressive disorders are associated with delusional-like experiences: a replication study based on a National Survey of Mental Health and Wellbeing. BMJ Open 2012;2:e001001. doi:10.1136/bmjopen-2012-001001
Contributors JM, SS and JS have directly participated in the planning and execution of the study. SS analysed the data. All authors have critically read and approved the final version submitted.
Funding This research received no specific grant from any funding agency in public, commercial or not-for-profit sectors.
Competing interests None.
Patient consent Obtained.
Ethics approval We obtained data from the Australian Bureau of Statistics (ABS), which is a government organisation. So, we are not aware about the exact approval authority. However, it is understood that the ABS has followed all the ethical standards to conduct this national survey.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement The data are available from the Australian Bureau of Statistics.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.