Which psychotic experiences are associated with a need for clinical care?

G Model

EURPSY-3207; No. of Pages 7 European Psychiatry xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

European Psychiatry journal homepage: http://www.europsy-journal.com

Original article

Which psychotic experiences are associated with a need for clinical care? C.M.C. Brett a,b, E.R. Peters c,d,*, P.K. McGuire a,d,e a

King’s College London, Institute of Psychiatry, Psychology & Neuroscience (IoPPN), Department of Psychosis Studies, London, United Kingdom Sussex Partnership NHS Foundation Trust, Brighton, United Kingdom c King’s College London, IoPPN, Department of Psychology, PO77, HWB, Institute of Psychiatry, Psychology & Neuroscience, De Crespigny Park, London SE5 8AF, United Kingdom d National Institute for Health Research (NIHR) Biomedical Research Centre for Mental Health at South London and Maudsley NHS Foundation Trust, London, United Kingdom e OASIS, Psychosis Clinical Academic Group, South London and Maudsley NHS Foundation Trust, London, United Kingdom b

A R T I C L E I N F O

A B S T R A C T

Article history: Received 7 May 2014 Received in revised form 11 December 2014 Accepted 12 December 2014 Available online xxx

Background: The aims of this study were to identify (1) the factor structure of anomalous experiences across the psychosis continuum; (2) qualitative and quantitative differences in psychotic experiences (PEs) between ‘‘non need-for-care’’ and two clinical groups: psychosis patients and individuals at ultra high risk (UHR) of psychosis. We aimed to distinguish which types of experiences would be related to malign (need-for-care and/or help-seeking) versus benign outcomes. Methods: Component scores obtained from a Principal Components Analysis of PEs from lifetime scores on the Appraisals of Anomalous Experience Inventory (Brett et al., 2007) were compared across 96 participants: patients diagnosed with a psychotic disorder (n = 37), help-seeking UHR people (n = 21), and non-clinical individuals presenting with enduring PEs (n = 38). Results: A five-component structure provided the best solution, comprising dissociative-type experiences, subjective cognitive deficits, and three separate components relating to ‘‘positive’’ symptoms. All groups reported ‘‘positive’’ experiences, such as ideas of reference and hallucinations, with the non-clinical group displaying more PEs in the Paranormal/Hallucinatory component than both clinical groups. ‘‘Cognitive/Attentional anomalies’’ was the only component where the clinical groups reported significantly more anomalies than the non-clinical group. However psychosis patients reported more frequent first-rank type symptoms and ‘‘hypomanic’’ type PEs than the other groups. Discussion: ‘‘Positive’’ PEs were common across the psychosis spectrum, although first-rank type symptoms were particularly marked in participants diagnosed with a psychotic disorder. Help-seeking and need-for-care were associated with the presence of subjective cognitive disturbances. These findings suggest that anomalies of cognition and attention may be more relevant to poorer outcomes than the presence of anomalous experiences. ß 2015 Published by Elsevier Masson SAS.

Keywords: Checking behavior Eye-tracking Image comparison task Schizophrenia Working memory

1. Introduction There is accumulating evidence to support the existence of a continuum of the psychosis phenotype across the population, which in combination with sensitizing factors may manifest in the clinical presentations of the psychotic disorders [52]. However there is some debate over the exact nature of the continuum [13],

* Corresponding author. King’s College London, IoPPN, Department of Psychology, PO77, HWB, Institute of Psychiatry, Psychology & Neuroscience, De Crespigny Park, London SE5 8AF, United Kingdom. Tel.: +44 0 207 848 0347; fax: +44 0 207 848 5006. E-mail address: [email protected] (E.R. Peters).

the phenotype may not be unitary [23] and the nature of the continuum may be best captured by a combination of categorical and dimensional descriptors [1,14,49]. The heterogeneity of treatment response amongst people with clinically-relevant psychotic symptoms, including the identification of a subgroup that recover without antipsychotic medication, has led to suggestions that a more nuanced understanding of the continuum and subgroups of psychosis is needed [5,12]. Early detection and intervention programmes also depend on the identification of ‘‘At-Risk Mental States’’ (ARMS) to decide which types of presentation are likely to benefit most from identification and intervention [17]. Factor analyses have identified four or five dimensions of psychotic symptoms, including variations on ‘‘manic’’, ‘‘depressive’’,

http://dx.doi.org/10.1016/j.eurpsy.2014.12.005 0924-9338/ß 2015 Published by Elsevier Masson SAS.

Please cite this article in press as: Brett CMC, et al. Which psychotic experiences are associated with a need for clinical care? European Psychiatry (2015), http://dx.doi.org/10.1016/j.eurpsy.2014.12.005

G Model

EURPSY-3207; No. of Pages 7 C.M.C. Brett et al. / European Psychiatry xxx (2015) xxx–xxx

2

‘‘disorganised’’, ‘‘negative’’ and ‘‘reality distortion’’ syndromes, in psychotic [34,38,50], ARMS [15], and healthy but psychosis-prone [11,53] samples. These studies have typically used different measures depending on the sample investigated: interviews assessing symptom severity in patient samples [2,3]; the Comprehensive Assessment of ARMS (CAARMS [59]) in Ultra High Risk (UHR) samples; and questionnaires measuring schizotypal traits and attenuated symptoms in non-clinical samples [33]. It is therefore difficult to compare the presentation and clustering of psychotic experiences across samples. On the one hand, diagnostic interviews can be limited, as only specific, illnessrelated forms of anomalous experience are being elicited. On the other hand, self-report questionnaires include items that are open to individual interpretation. This may be especially problematic in the area of subjective anomalous experiences, which can be subtle or difficult to describe. Furthermore, with some notable exceptions [18], quantitative aspects of the continuum of PEs have rarely been investigated, as questionnaires tend to use probe questions with dichotomous ‘‘yes/no’’ response options. However, there is evidence that more frequent experiences are associated with an increased risk of need-for-care [4,51] and psychotic disorder [25], although it is not known if this relates to specific kinds of anomalies. Negative and disorganized symptoms are predictive of transition to psychosis in ARMS samples [14,48], but it is unclear whether these symptoms also differentiate individuals with PEs with and without a need-for-care. This study sought to identify qualitative and quantitative differences between three groups of individuals, all presenting with anomalous experiences but differing according to their need for clinical care. Non-clinical (NC) participants were specifically selected for the presence of enduring PEs1 but with no accompanying distress or need-for-care2. UHR participants had sought clinical help after developing PEs, and met criteria for an ARMS for psychosis. Clinical (C) participants had a psychotic disorder and were in receipt of mental health services. The Appraisals of Anomalous Experiences (AANEX–Inventory [7]) was used to measure lifetime anomalous experiences in each group. This semi-structured interview elicits and rates anomalous experiences from a purely subjective perspective, and does not depend on objectively observed symptoms or behaviours linked to particular diagnoses. It therefore provided a means to assess the same phenomena in people who varied in their need-for-care. 2. Objectives of the study The aims of the study were to identify:  the factor structure of anomalous experiences occurring across the psychosis spectrum;  whether PEs in a ‘non need-for-care’ group were qualitatively different from those with a psychosis diagnosis or meeting ARMS criteria; 1 We use the terms ‘anomalous experiences’, ‘anomalies’ and ‘PEs’ interchangeably throughout the text. We have deliberately not used the term ‘psychosis-like experiences’ (PLEs) since the experiences we are referring to are not necessarily sub-threshold. 2 We have recruited from this population in a number of previous studies, please see [20] Gaynor K, Ward T, Garety P, Peters E. The role of safety-seeking behaviours in maintaining threat appraisals in psychosis. Behav Res Ther. 2013;51:75-81, [21] Heriot-Maitland C, Knight M, Peters E. A qualitative comparison of psychotic-like phenomena in clinical and non-clinical populations. Brit J Clin Psychol. 2012;51:3753, [32] Lovatt A, Mason O, Brett C, Peters E. Psychotic-Like Experiences, Appraisals, and Trauma. J Nerv Ment Dis. 2010;198:813-9, [54] Ward TA, Gaynor KJ, Hunter MD, Woodruff PWR, Garety PA, Peters ER. Appraisals and Responses to Experimental Symptom Analogues in Clinical and Nonclinical Individuals With Psychotic Experiences. Schizophrenia Bull. 2014;40:845-55. For further description of the nature of this unique sample.

 whether there were quantitative differences in frequency and/or duration of anomalous experiences between the three groups. 3. Method 3.1. Participants The 96 participants comprised three groups of people reporting PEs. Exclusion criteria for all participants included:    

inability to speak and understand fluent English; history of neurological problems, head injury or epilepsy; current substance dependence; estimated current IQ < 70 (based on four WAIS-III [55] subtests). Other data from a subset of this sample have been reported elsewhere [6,7–9].

The Clinical (C) group (n = 37) met DSM-IV criteria for any schizophrenia-spectrum disorder as recorded in medical notes, recruited from the South London and Maudsley NHS Foundation Trust (SLaM). Individuals in their first episode (n = 23; Lambeth Early Onset service; LEO), and with a longer history (n = 14; Psychological Interventions Clinic for outpatients with Psychosis; PICuP) were recruited, to include a range of individuals similar to both the nonclinical group (who are typically in their 40s, with a childhood onset of experiences) and the at-risk group (who are younger with a recent onset of experiences). Their mean age was 32.6 years, 55% were male, and 82.5% were taking psychotropic medication including antipsychotics, mood stabilisers, and antidepressants. The UHR group (n = 21) met the PACE criteria for an ARMS for psychosis [58]. They were recruited from OASIS (Outreach And Support In South London [10]). Their mean age was 23.7 years, 68% were male, and 32% were receiving psychotropic medication. The non-clinical (NC) group (n = 38) comprised participants reporting enduring PEs, who had never sought or received clinical care for their anomalous experiences, and who were not distressed by them. They were recruited from the London area through advertisements on special interest websites, magazines and email groups, to access sub-cultural populations interested in altered states, mediumship, witchcraft etc. Volunteers were screened for suitability using a questionnaire enquiring about the lifetime incidence of a range of anomalous experiences, based on the AANEX-Inventory [7]. Only individuals with at least ‘‘occasional’’ experiences of any Schneiderian symptom3, in the absence of drug use and in clear consciousness, were invited to participate. This criterion ensured that these were participants reporting PEs of direct comparability to clinical phenomena, rather than merely unusual experiences associated with schizotypy. Those who reported any history of clinical intervention for their PEs, or were judged to be in need of care by the experimenter, were excluded. To differentiate them from the UHR group (who have a high risk of developing a psychotic disorder), only participants whose anomalous experiences had commenced more than five years previously, were included. Their mean age was 34 years, 63% were male and none were taking psychotropic medication. A one-way Anova and subsequent post-hoc contrasts (P < .01) showed no age differences between NC and C groups, which were both older than UHR group (UD > UHR: m.d. = 0.35; 99% C.I.’s = 0.17–0.53; P < .001; D > UHR: m.d. = 0.28; 99% C.I.’s = 0.1–0.45; p < .001). There was no association between group and gender (x2(3) = 0.95; P = .81). 3 Ie. symptoms considered indicative of a likely diagnosis of schizophrenia, comprising: delusions of control, thought broadcasting, thought withdrawal, thought insertion, hearing one’s thoughts spoken aloud, second and third person auditory hallucinations.

Please cite this article in press as: Brett CMC, et al. Which psychotic experiences are associated with a need for clinical care? European Psychiatry (2015), http://dx.doi.org/10.1016/j.eurpsy.2014.12.005

G Model

EURPSY-3207; No. of Pages 7 C.M.C. Brett et al. / European Psychiatry xxx (2015) xxx–xxx

3.2. Materials and procedures The AANEX-Inventory [7] was used to elicit and rate selfreports of anomalous experiences over participants’ lifetime. It comprises probe questions and rating scales for 40 anomalous experiences associated with psychosis, across anomalies of perception, cognition and affect (including those non-specific for psychosis but frequently reported in the context of psychosis), as well as Schneiderian symptoms. Participants were interviewed between 30 minutes to 2 hours, with average duration being approximately 1 hour. Each experience was rated between 1-5, based on the frequency or pervasiveness of the experience across an individual’s lifetime. 3.3. Statistics Principal Components Analysis (PCA) was selected as the most suitable method for exploratory analysis, and to provide an empirical summary of the data set for purposes of cross-sectional comparison. PCA is designed to summarise phenomenological information rather than uncover latent factors, and yields a unique mathematical solution through the extraction of maximum variance from the data set with a certain number of orthogonal components. Varimax rotation was selected to improve the interpretability of the components [47]. The data set for the PCA pooled the scores of the three groups, which were recruited on the basis of reporting anomalous experiences, and were differentiated on grounds independent of the phenomenology being submitted to PCA. Missing data were replaced with the mean, which does not affect component structure, but allows for component regression scores to be derived. As in any factor analysis, the selection of the component solution was constrained by expectations based on a qualitative analysis of the data [6], as well as existing literature and statistical heuristics. The analysis was carried out in several stages, due to competing restrictions of sample size and the desirability of maximum representation of the data set. First, a 20-item PCA was performed to explore which components emerged when there was a good ratio of cases to items. This analysis suggested either a four- or fivecomponent solution, based on the Catell scree plot of eigenvalues. Next, a 33-item PCA (including all items except for a set with > 20% missing data) was performed to examine whether the previous 20 items contributed to the same components when four- and fivecomponent solutions were conducted: i.e. the stability of the components. Next, since the items with missing data were nonrandom, a 40-item PCA was performed to examine the stability of the previous components after the addition of these particular items, when four- and five-component solutions were conducted. On the basis of these three sets of analyses, and subsequent investigations to test the reliability of the solutions, including rerunning the analysis excluding the cases with missing data, the five-component solution was chosen as being the most robust and valid (see Table 1). Further information regarding each stage of the analysis is available from the first author on request. Anovas were used for the cross-sectional analyses. Significance values were adjusted by the number of comparisons carried out (P < .05/n), to reduce likelihood of Type 1 errors. 4. Results 4.1. Interpretation of components The PCA method is a potent form of data reduction, and the component solution is valid as a summary of the data, irrespective of interpretability. However, it was considered theoretically

3

important that the identified components were interpretable. The list of items contributing to each component is given in Table 1. Meaning/Reference (MeanRef) component reflected manic or hypomanic states and experiences, including enhanced meaningfulness of events, subjectively profound affective experiences and prominent ‘revelatory’ experiences. Paranormal/Hallucinatory (ParaHall) component reflected alterations in sense of agency, ‘‘magical’’ thinking, and unusual perceptual experiences, excluding auditory hallucinations. Paranormal experiences such as mediumship, clairvoyance and magic, and perception of other entities/energies, would contribute to scores on this component, as would similar phenomena viewed as hallucinatory and/or delusional in the context of clinical psychosis. Cognitive/Attention (CogAtt) component reflected non-specific subjective changes or deficits in thinking and attention, along with emotional disturbance. Dissociative/Perceptual (DissPer) component reflected a second set of non-specific experiences: dissociative experiences and tendencies, along with other global perceptual changes. First-Rank Type Symptoms (FRTS) component reflected experiences of weakened boundaries between self and other, or alterations in the experience of thoughts, in addition to auditory hallucinations. 4.2. Group comparisons 4.2.1. Derivation of component scores Scores for the five components were created using the regression method, which computes regression-like coefficients for weighting standardised variable scores from the factor loading matrix and the original correlation matrix. These are mutually uncorrelated, and have a mean of 0 and a standard deviation of 1 (see Table 2). The component scores consist of the sum of weighted lifetime scores for the contributing anomalies. As such, a higher mean score in one group could either suggest that participants experienced a greater number of different contributing experiences, or that they tended to experience only a few, but more frequently. Therefore two further variables were derived from the components to test for quantitative and qualitative differences separately. Qualitative differences between the groups were tested by computing ‘‘Additive scores’’. This variable summed the number of different anomalies from each component experienced by participants across their lifetime, irrespective of the frequency of occurrence. The ratings for each anomaly were dichotomised into ‘‘present’’ (original score 3–5) or ‘‘absent’’ (original score 1-2). The potential range of scores was 0-8 for MeanRef, ParaHall and DissPer, 0-11 for CogAtt, and 0-5 for FRTS, with items assigned to components by highest item-component correlation (see Table 3). Quantitative differences between the groups were tested by comparing their mean component scores, while controlling for the Additive scores. Since the contribution of different anomalies was weighted in the component scores, the second analysis was a test of the relative ‘‘severity’’ (frequency/duration) of the component itself, between the groups. 4.2.2. Additive scores Since the groups were not matched for age, its association with Additive scores was tested using bivariate correlations. No significant associations were found, and age was not included in subsequent analyses. One-way Anovas (Group  additive score) showed effects of Group for three components: CogAtt (F2,92 = 6.104; p = .003), MeanRef (F2,92 = 5.297; p = .007), and ParaHall (F2,92 = 13.34; P < .001).

Please cite this article in press as: Brett CMC, et al. Which psychotic experiences are associated with a need for clinical care? European Psychiatry (2015), http://dx.doi.org/10.1016/j.eurpsy.2014.12.005

G Model

EURPSY-3207; No. of Pages 7 C.M.C. Brett et al. / European Psychiatry xxx (2015) xxx–xxx

4

Table 1 Summary of loadings for each item on five components. This solution accounts for 43.6% of the variance. Items

MeanRef

ParaHall

CogAtt

DissPer

FRTS

Insights (autochthonous ideas) Sense of having a mission Elation Ideas of reference Loss of boundary between self and other Thought pressure Time distortion Feeling of being monitored Premonitions Somatic experiences/hallucinations Passivity (automatic movements/other external influences on action) Global visual changes Controlling events/others mentally Visual hallucinations Controlled actions Olfactory anomalies Language disturbance Thought blockages Lost automatic skills Captivation/fixation by visual features Disorientation Can’t divide attention Mixed emotions Increased distractability Sense of doom/nihilistic experience Thought interference (intrusive thoughts) Concretism Loss of emotions Depersonalisation Derealisation Oversensitivity Global auditory changes Subjective isolation (feeling cut off from others) Out of body experiences Emotional overreactivity Receptivity: experiencing others’ experiences as if own Thought withdrawal Voices Thought transmission (telepathy) Loud thoughts

.864 .793 .670 .609 .597 .524 .505 .357 .139 .075 .489 .102 .331 –.046 .170 –.102 –.042 .042 –.027 –.199 .293 .134 .226 .365 –.121 .078 .179 .037 .085 .299 –.058 .002 .305 .004 .343 .183 .048 –.193 .311 .054

.042 .107 .419 –.125 .482 .000 .442 –.122 .683 .536 .495 .475 .452 .444 .430 .422 .078 –.110 –.195 .226 .023 –.403 .210 –.192 –.048 –.229 .178 –.047 .184 .169 .144 .084 .007 .367 –.142 .059 .119 .138 .210 –.264

–.111 –.122 –.100 .033 .174 .285 .250 .167 –.142 –.051 .254 .141 –.140 –.013 .262 –.007 .601 .581 .556 .532 .474 .468 .438 .431 .408 .380 .363 .221 .194 .195 .120 .293 .019 –.288 .149 –.048 .349 –.057 .125 –.005

.151 –.051 .129 .180 –.017 .085 .182 –.173 .143 .044 –.052 .394 .097 .210 –.285 .011 .122 .006 .423 .132 .178 .264 .272 .169 .135 .011 .038 .626 .593 .575 .569 .529 .452 .396 .383 .107 –.089 –.007 –.002 .216

.004 .239 –.110 .314 .021 –.153 –.029 .318 .039 .044 .192 .033 .069 .371 .180 –.007 .179 –.031 .088 –.104 .088 –.043 –.106 .020 .058 .088 .031 .000 –.227 –.113 .123 .305 .227 .105 .017 .756 .682 .667 .618 .411

Item-component correlations > 35 are in bold. MeanRef: meaning/reference; ParaHall: paranormal/hallucinatory; CogAtt: cognitive/attention; DissPer: dissociative/ perceptual; FRTS: first-rank type symptoms.

Post-hoc comparisons revealed that the group effect on the CogAtt component was due to the NC group experiencing fewer anomalies than either the C or the UHR groups (NC < C: mean difference = 1.84; 99.4% C.I.’s = 0.06–3.62; P = .005; NC < UHR: mean difference = 2.22; 99.4% C.I.’s = 0.13–4.32; P = .004). The group effect on MeanRef reflected a difference between the NC and UHR groups, with the NC group experiencing more anomalies than the UHR group (mean difference = 1.89; 99.4% C.I.’s = 0.24–3.54; P = .002). The group effect on the ParaHall component resulted from the NC group experiencing more anomalies than both the C and UHR groups (NC > C: mean difference = 1.69; 99.4% C.I.’s = 0.44–2.93; P < .001; NC > UHR: mean difference = 2.48; 99.4% C.I.’s = 1.02– 3.95; P < .001).

4.2.3. Quantitative comparison Ancovas (Group  Component scores, controlling for Additive scores) revealed effects of Group for four of the five components: MeanRef (F2,92 = 6.444; P = .003), CogAtt (F2,92 = 3.715; P = .029), DissPer (F2,92 = 3.157; P = .048), and FRTS (F2,92 = 13.073; P < .001). Post-hoc comparisons showed that the C group had more frequent MeanRef anomalies than the NC group (C > NC: mean difference = 0.49; 99.6% C.I.’s = .01–.90; P = 001), and more frequent FRTS experiences than both the NC and UHR groups (C > NC: mean difference = 0.58; 99.6% C.I.’s = 0.21–0.95; P < .001; C > UHR: mean difference = 0.53; 99.6% C.I.’s = 0.09–0.97; P < .001). There were trends towards more frequent CogAtt anomalies in the UHR group compared to the NC group (P = .009), and towards more

Table 2 Mean factor scores (standard deviation), and range of scores for each group. Group

UD D UHR

MeanRef

ParaHall

CogAtt

DissPer

FRTS

Mean (s.d.)

Range

Mean (s.d.)

Range

Mean (s.d.)

Range

Mean (s.d.)

Range

Mean (s.d.)

Range

.030 (.851) .260 (1.12) –.540 (.863)

3.76

.475 (.637) –.218 (1.20) –.523 (.784)

2.84

–.488 (.734) .236 (.953) .513 (1.15)

2.86

.251 (.949) –.342 (1.00) .144 (.960)

3.84

–.298 (.813) .517 (1.05) –.375 (.859)

2.83

4.43 3.44

5.05 3.15

4.27 3.70

3.83 2.98

4.17 2.92

MeanRef: meaning/reference; ParaHall: paranormal/hallucinatory; CogAtt: cognitive/attention; DissPer: dissociative/perceptual; FRTS: first-rank type symptoms.

Please cite this article in press as: Brett CMC, et al. Which psychotic experiences are associated with a need for clinical care? European Psychiatry (2015), http://dx.doi.org/10.1016/j.eurpsy.2014.12.005

G Model

EURPSY-3207; No. of Pages 7 C.M.C. Brett et al. / European Psychiatry xxx (2015) xxx–xxx

5

Table 3 Mean additive scores (SD’s), and range of scores for each group. Group

UD D AR

MeanRef

ParaHall

CogAtt

DissPer

FRTS

Mean (s.d.)

Range

Mean (s.d.)

Range

Mean (s.d.)

Range

Mean (s.d.)

Range

Mean (s.d.)

Range

5.361 (1.95) 4.85 (2.19) 3.472 (1.87)

0–8

5.281 (1.35) 3.592 (2.23) 2.792 (1.62)

3–7

3.092 (2.10) 4.941 (2.63) 5.321 (3.20)

0–9

4.58 (2.17) 3.56 (2.00) 3.79 (2.32)

0–8

2.85 (1.20) 3.21 (1.27) 2.37 (1.34)

0–5

0–8 1–8

0–7 0–6

0–10 2–11

0–7 0–8

0–5 0–4

Mean scores marked ‘‘1’’ were significantly higher than those marked ‘‘2’’ in post-hoc comparisons (P < .006). MeanRef: meaning/reference; ParaHall: paranormal/ hallucinatory; CogAtt: cognitive/attention; DissPer: dissociative/perceptual; FRTS: first-rank type symptoms.

frequent DissPer anomalies in the UHR group compared to the C group (P = .015). 4.2.4. Overlapping phenomenology? The preceding analyses indicate that, on average, the C group experienced more frequent MeanRef and FRTS anomalies than the NC group. Post-hoc analyses were conducted to explore further the distribution of these specific anomalies in the three groups. First, two new variables were computed to designate the average frequency of anomalies contributing to the MeanRef and FRTS components: component scores were divided by the Additive scores, to give an index of average frequency. Next, the mean ‘average frequency’ index for the C group was calculated for both MeanRef and FRTS. Finally, the number of participants from each group whose ‘average frequency’ index for these traits was equal to or greater than the mean of the C group, was counted. A large subset (45%) of the NC group, and a smaller subset of the AR group (19%) had an ‘‘average frequency’’ score higher than or equal to the mean of the C group on the MeanRef component. A smaller subset of the NC and AR groups (16% and 10% respectively) had an equivalent profile on the FRTS component.

5. Discussion This study examined the factorial structure of anomalous experiences across the psychosis continuum, and compared the nature and frequency of these experiences across groups distinguished by their need for clinical care. 5.1. Identifying the component structure of anomalous experiences The PCA-derived components indicated that phenomena commonly grouped together as ‘‘positive symptoms’’ of psychosis can be subdivided into different clusters of anomalous experience when examined at a more fine-grained level of analysis, unconstrained by diagnostic criteria. Previous research, mainly in patients with psychotic disorders, has often identified a single factor corresponding to ‘‘positive symptoms’’ [30]. In the present study, ‘‘delusional’’ experiences (such as ideas of reference, autochthonous ideas, and experiences of being controlled by, or controlling others), non-auditory hallucinations, and Schneiderian symptoms loaded onto separate components, Meaning/Reference (MeanRef), Paranormal/Hallucinatory (ParaHall) and First-Rank Type Symptoms (FRTS), respectively. This difference may be related to the use of lifetime rather than cross-sectional ratings of phenomenology, and the study of individuals across the psychosis spectrum, rather than of participants diagnosed with psychotic disorders alone. The presence of separate positive PEs components could be suggestive of differing subtypes of psychotic presentation, with the clustering of most of the Schneiderian symptoms together

being partially distinguished from other PEs. While caution must be exercised when inferring the existence of genuine psychopathological syndromes from a PCA (which simply expresses the statistical relationship between items), the observed components appear to have face validity. The emergence of a distinct component relating to cognitive and attentional changes concurs with existing research looking at symptom factors in people with a psychotic disorder [30], schizotypal traits in non-psychotic populations [31,56], and ARMS samples [15,44]. These convergent findings suggest that subjective cognitive disturbances, variously described as ‘‘cognitive disorganization’’ ([33]; schizotypy), ‘‘basic symptoms’’ ([41]; ARMS), and our Cognitive/Attention (CogAtt) component, are separable from other ‘‘positive’’ symptoms of psychosis, and highly relevant to psychosis phenomenology across the continuum. The emergence of the Dissociative/Perceptual (DissPer) component also corroborates the observed association between dissociative symptoms and psychosis [26,40]. 5.2. Group comparisons: qualitative differences Due to the selection criteria for the three groups, some overlap in the range of experiences reported by participants across the spectrum was inevitable. However, the cross-sectional similarities and differences between the groups suggest that many forms of anomalous perceptual and hallucinatory experiences are not necessarily associated with a need-for-care. The only component where both clinical groups reported significantly more anomalies than the NC group was the CogAtt component, which represents self-reported cognitive difficulties, rather than symptoms of psychosis per se. In contrast, individuals with no need-for-care reported comparable numbers of anomalies contributing to DissPer and FRTS to both clinical groups, and a greater number of anomalies contributing to ParaHall than both clinical groups, and to the MeanRef component than the UHR group. The higher scores on the CogAtt component for the clinical groups than the NC group are in line with evidence from prospective studies of ARMS populations [37,29,27], and crosssectional comparisons of non-clinical and clinical samples [32], that alterations in subjective and/or objective cognitive function are associated with help-seeking and an increased risk of clinical care and psychotic disorder [15,43,39,19]. Indeed many of the anomalies that loaded onto CogAtt (such as thought blockages, distractibility, loss of automatic skills) are similar to those linked to psychosis risk in the Schizophrenia Proneness Instrument-Adult version (SPI-A [42]) especially their ‘‘Cognitive Disturbances’’ factor. Our data are also consistent with previous evidence that ‘positive’ dimensions of unusual experiences are less indicative of the presence of a psychotic disorder [24,60] and poor outcome [35].

Please cite this article in press as: Brett CMC, et al. Which psychotic experiences are associated with a need for clinical care? European Psychiatry (2015), http://dx.doi.org/10.1016/j.eurpsy.2014.12.005

G Model

EURPSY-3207; No. of Pages 7 6

C.M.C. Brett et al. / European Psychiatry xxx (2015) xxx–xxx

5.3. Group comparisons: quantitative differences The analyses examining differences in frequency and/or duration of anomalies support a quantitative dimension to the psychosis continuum. The C group reported more frequent or severe anomalies contributing to the MeanRef and FRTS components compared to the NC group and both the UHR and NC groups, respectively. These findings demonstrate the value of exploring individuals’ anomalous experiences using in-depth interviews and ordinal ratings rather than measures using dichotomised response options. They suggest that at least some of the variability in outcome associated with PEs is related to how pervasive these anomalies are in the individual’s life. Our data would suggest that more pervasive or severe first-rank type symptoms are more likely to lead to receipt of diagnosis and treatment than other presentations. However, it was also found that a subset of the NC group reported frequencies of MeanRef, and to a lesser extent, FRTS experiences, that were comparable to the C participants, suggesting that other factors are important in protecting individuals from requiring psychiatric services when they are positioned at the higher end of this quantitative continuum. Such factors are beyond the scope of the current paper but have been explored elsewhere [7–9,6].

6. Conclusions: implications and limitations Overall, the findings contribute to a conceptualisation of the psychosis continuum as having both qualitative and quantitative aspects [23]. Indirectly, the data provide support for psychological models of the development of clinically-relevant psychosis, and a normalising view of anomalous experiences: the extensive overlap in the phenomenology across the groups suggests that such experiences alone do not indicate the presence of a clinical disorder. Their impact on need-for-care may be mediated by other factors, such as changes in cognition and attention, and pervasiveness of the experiences. This is of particular relevance to early detection and intervention programmes in psychosis, in which ‘attenuated’ psychotic symptoms are one of the key criteria of UHR [59]. Our data indicate that the presence of anomalous experiences, particularly those contributing to the MeanRef and ParaHall components, are not in themselves associated with a need for clinical care, while those linked to the CogAtt and FRTS components may be of more clinical significance. The NC group can be considered as representing ‘‘false positives’’ in relation to an UHR group: they would have met the symptomatic criteria for an ARMS at some point in their lives, but did not develop a psychotic disorder or a need-for-care. A subset of the NC group may also have met diagnostic criteria for Transient Psychotic Disorder or Schizophreniform Psychosis [57] at some point. The fact that they did not require care from psychiatric services corroborates other studies showing that a subset of individuals may recover spontaneously from psychotic episodes without the use of antipsychotic medications [5,12,46,45]. As the current study relied on self-report of anomalous experience, in some cases retrospectively, it is not possible to determine conclusively whether individuals would have met diagnostic criteria; we also did not assess the types of non-psychiatric support that may have contributed to the successful recovery of these individuals. Nevertheless, the detailed interview data elicited from participants is indicative of psychosislike crisis in some of them. Further research is indicated into the predictors of good outcomes following psychotic experiences to inform clinical early detection and intervention approaches. This could be considered in the context of other research exploring the

protective factors associated with non-clinical occurrence of PEs [21,6]. In summary, this study suggests a promising avenue of investigation for further exploration of the different dimensions of the psychosis phenotype, and in turn allowing a more sensitive investigation of the development of distress and need-for-care in the context of a distribution of liability to anomalous experiences in the general population. The results need to be interpreted in the light of limitations such as the modest sample size, and the use of ordinal, rather than fully continuous, scores. However, the use of this type of data is not considered an overriding problem, especially if meaningful factors are extracted [47,22]. The strength of this study is that a broad spectrum of specific items describing psychosis-related anomalous experience were included, which were not constrained a priori by diagnostic or clinical constructs. The inclusion of data across the clinical and non-clinical spectrum meant that the association of different kinds of subjective experience was not a function of the ‘‘clinician’s bias’’ but representative of a broader section of the putative phenotypic continuum. The presence of the NC group, members of which have demonstrated a good outcome in the context of PEs, is particularly informative in the cross-sectional analyses in terms of finding associations between the components and ‘‘outcome’’ variables. Nevertheless, this study requires replication in a larger sample in order to establish whether the observed component structure and the group differences are robust. Given the identification of ‘‘depressive’’ factors in other studies, and the association between anxiety, depression, mania and psychosis, it would be interesting to include items rating anxiety and low mood to explore the links between PEs and emotional disorders. A further group of interest, not recruited in the present study, would have been a general population sample reporting distressing PEs, but who are not helpseeking, to further disentangle seeking/receiving clinical treatment from need-for-care [28,36,16]. Finally, it also appears important to investigate further the protective factors characterising the NC group, widening the focus to intrapersonal, social and contextual variables, ideally utilising longitudinal designs. Disclosure of interest The authors declare that they have no conflicts of interest concerning this article. Acknowledgements This work was supported by a studentship from the Joint Research Committee of King’s College Hospital and King’s College London, allocated to the 1st author, and the Guys & St Thomas’ Charity Foundation (which funded the OASIS service). References [1] Allardyce J, Suppes T, Van Os J. Dimensions and the psychosis phenotype. Int J Meth Psych Res 2007;16:S34–40. [2] Andreasen N. Scale for the assessment of negative symptoms (SANS). Iowa City, IA: University of Iowa; 1984. [3] Andreasen N. Scale for the assessment of positive symptoms (SAPS). Iowa City, IA: University of Iowa; 1984. [4] Bak M, Myin-Germeys I, Hanssen M, Bijl R, Vollebergh W, Delespaul P, et al. When does experience of psychosis result in a need-for-care? A prospective general population study. Schizophr Bull 2003;29:349–58. [5] Bola JR. Medication-free research in early episode schizophrenia: evidence of long-term harm? Schizophr Bull 2006;32:288–96. [6] Brett CMC. Transformative Crises. In: Clarke I, editor. Psychosis and spirituality: consolidating the new paradigm. 2nd ed., Chichester: Wiley-Blackwell; 2010. [7] Brett CMC, Peters EP, Johns LC, Tabraham P, Valmaggia LR, McGuire P. Appraisals of anomalous experiences interview (AANEX): a multidimensional measure of psychological responses to anomalies associated with psychosis. Br J Psychiat 2007;191:S23–30.

Please cite this article in press as: Brett CMC, et al. Which psychotic experiences are associated with a need for clinical care? European Psychiatry (2015), http://dx.doi.org/10.1016/j.eurpsy.2014.12.005

G Model

EURPSY-3207; No. of Pages 7 C.M.C. Brett et al. / European Psychiatry xxx (2015) xxx–xxx [8] Brett CMC, Johns LC, Peters EP, McGuire PK. The role of metacognitive beliefs in determining the impact of anomalous experiences: a comparison of helpseeking and non-help-seeking groups of people experiencing psychotic-like anomalies. Psychol Med 2009;39:939–50. [9] Brett C, Heriot-Maitland C, McGuire P, Peters E. Predictors of distress associated with psychotic-like anomalous experiences in clinical and non-clinical populations. Br J Clin Psychol 2014;53:213–27. [10] Broome MR, Woolley JB, Johns LC, Valmaggia LR, Tabraham P, Gafoor R, et al. Outreach and support in south London (OASIS): implementation of a clinical service for prodromal psychosis and the at-risk mental state. Eur Psychiat 2005;20:372–8. [11] Claridge G, McCreery C, Mason O, Bentall R, Boyle G, Slade P, et al. The factor structure of ‘‘schizotypal’ traits: a large replication study. Br J Clin Psychol 1996;35(Pt. 1):103–15. [12] Cullberg J, Levander S, Holmqvist R, Mattsson M, Wieselgren IM. One-year outcome in first episode psychosis patients in the Swedish Parachute project. Acta Psychiat Scand 2002;106:276–85. [13] David AS. Why we need more debate on whether psychotic symptoms lie on a continuum with normality. Psychol Med 2010;40:1935–42. [14] Demjaha A, Morgan K, Morgan C, Landau S, Dean K, Reichenberg A, et al. Combining dimensional and categorical representation of psychosis: the way forward for DSM-V and ICD-11? Psychol Med 2009;39:1943–55. [15] Demjaha A, Valmaggia L, Stahl D, Byrne M, McGuire P. Disorganization/ cognitive and negative symptom dimensions in the at-risk mental state predict subsequent transition to psychosis. Schizophr Bull 2012;38:351–9. [16] DeVylder JE, Oh HY, Corcoran CM, Lukens EP. Treatment seeking and unmet need-for-care among persons reporting psychosis-like experiences. Psychiat Serv 2014;65:774–80. [17] Drake RJ, Lewis SW. Valuing prodromal psychosis: what do we get and what is the price? Schizophr Res 2010;120:38–41. [18] Freeman D, Garety PA, Bebbington PE, Smith B, Rollinson R, Fowler D, et al. Psychological investigation of the structure of paranoia in a non-clinical population. Br J Psychiat 2005;186:427–35. [19] Fusar-Poli P, Deste G, Smieskova R, Barlati S, Yung AR, Howes O, et al. Cognitive functioning in prodromal psychosis a meta-analysis. Arch Gen Psychiat 2012;69:562–71. [20] Gaynor K, Ward T, Garety P, Peters E. The role of safety-seeking behaviours in maintaining threat appraisals in psychosis. Behav Res Ther 2013;51: 75–81. [21] Heriot-Maitland C, Knight M, Peters E. A qualitative comparison of psychoticlike phenomena in clinical and non-clinical populations. Br J Clin Psychol 2012;51:37–53. [22] Hutcheson GD, Sofroniou N. The multivariate social scientist: an introduction to generalised linear models. Sage Publications; 1999. [23] Kaymaz N, van Os J. Extended psychosis phenotype – yes: single continuum – unlikely. Psychol Med 2010;40:1963–6. [24] Kelleher I, Keeley H, Corcoran P, Lynch F, Fitzpatrick C, Devlin N, et al. Clinicopathological significance of psychotic experiences in non-psychotic young people: evidence from four population-based studies. Br J Psychiat 2012;201:26–32. [25] Kendler KS, Gallagher TJ, Abelson JM, Kessler RC. Lifetime prevalence, demographic risk factors, and diagnostic validity of non-affective psychosis as assessed in a US community sample – The National Comorbidity Survey. Arch Gen Psychiat 1996;53:1022–31. [26] Kilcommons AM, Morrison AP. Relationships between trauma and psychosis: an exploration of cognitive and dissociative factors. Acta Psychiat Scand 2005;112:351–9. [27] Klosterkotter J, Hellmich M, Steinmeyer EM, Schultze-Lutter F. Diagnosing schizophrenia in the initial prodromal phase. Arch Gen Psychiat 2001;58: 158–64. [28] Kobayashi H, Nemoto T, Murakami M, Kashima H, Mizuno M. Lack of association between psychosis-like experiences and seeking help from professionals: a case-controlled study. Schizophr Res 2011;132:208–12. [29] Lencz T, Smith CW, McLaughlin D, Auther A, Nakayama E, Hovey L, et al. Generalized and specific neurocognitive deficits in prodromal schizophrenia. Biol Psychiat 2004;59:863–71. [30] Liddle PF. The symptoms of chronic-schizophrenia – a reexamination of the positive-negative dichotomy. Br J Psychiat 1987;151:145–51. [31] Linney YM, Murray RM, Peters ER, MacDonald AM, Rijsdijk F, Sham PC. A quantitative genetic analysis of schizotypal personality traits. Psychol Med 2003;33:803–16. [32] Lovatt A, Mason O, Brett C, Peters E. Psychotic-like experiences, appraisals, and trauma. J Nerv Ment Dis 2010;198:813–9. [33] Mason O, Claridge G, Jackson M. New scales for the assessment of schizotypy. Pers Indiv Diff 1995;18:7–13. [34] McGorry PD, Bell RC, Dudgeon PL, Jackson HJ. The dimensional structure of first episode psychosis: an exploratory factor analysis. Psychol Med 1998;28:935–47.

7

[35] Morita K, Kobayashi H, Takeshi K, Tsujino N, Nemoto T, Mizuno M. Poor outcome associated with symptomatic deterioration among help-seeking individuals at-risk for psychosis: a naturalistic follow-up study. Early Interv Psychiatry 2014;8:24–31. [36] Murphy J, Shevlin M, Houston J, Adamson G. A population-based analysis of subclinical psychosis and help-seeking behavior. Schizophr Bull 2012;38:360– 7. [37] Nelson B, Yuen HP, Wood SJ, Lin A, Spiliotacopoulos D, Bruxner A, et al. Longterm follow-up of a group at ultra high risk (‘‘Prodroma’’) for psychosis. The PACE 400 Study. JAMA Psychiat 2013;70:793–802. [38] Peralta V, Cuesta MJ. How many and which are the psychopathological dimensions in schizophrenia? Issues influencing their ascertainment. Schizophr Res 2001;49:269–85. [39] Ruhrmann S, Schultze-Lutter F, Salokangas RKR, Heinimaa M, Linszen D, Dingemans P, et al. Prediction of psychosis in adolescents and young adults at high risk results from the prospective european prediction of psychosis study. Arch Gen Psychiat 2010;67:241–51. [40] Schafer I, Fisher HL, Aderhold V, Huber B, Hoffmann-Langer L, Golks D, et al. Dissociative symptoms in patients with schizophrenia: relationships with childhood trauma and psychotic symptoms. Compr Psychiat 2012;53:364–71. [41] Schultze-Lutter F. Subjective symptoms of schizophrenia in research and the clinic: the basic symptom concept. Schizophr Bull 2009;35:5–8. [42] Schultze-Lutter F, Addington J, Ruhrmann S, Klosterkotter J. The schizophrenia proneness instrument: adult version (SPI-A). Rome: Giovanni Fiority Editore; 2007. [43] Schultze-Lutter F, Steinmeyer EM, Ruhrmann S, Klosterkotter J. The dimensional structure of self-reported ‘‘prodromal’’ disturbances in schizophrenia. Clin Neuropsychiatr J Treat Eval 2008;5:140–50. [44] Schultze-Lutter F, Ruhrmann S, Berning J, Maier W, Klosterkotter J. Basic symptoms and ultra high risk criteria: symptom development in the initial prodromal state. Schizophr Bull 2010;36:182–91. [45] Seikkula J, Aaltonen J, Rasingkangas A, Alakare B, Holma J, Lehtinen V. Opendialogue approach: treatment principles and preliminary results of a two year follow-up on first episode schizophrenia. Ethical Hum Sci Serv 2003;5:163–82. [46] Seikkula J, Aaltonen J, Alakare B, Haarakangas K, Keranen J, Lehtinen K. Fiveyear experience of first episode non-affective psychosis in open-dialogue approach: Treatment principles, follow-up outcomes, and two case studies. Psychother Res 2006;16:214–28. [47] Tabachnick B, Fidell L. Using multivariate statistics, 2nd ed., New York: Harper Collins; 1989. [48] Valmaggia LR, Stahl D, Yung AR, Nelson B, Fusar-Poli P, McGorry PD, et al. Negative psychotic symptoms and impaired role functioning predict transition outcomes in the at-risk mental state: a latent class cluster analysis study. Psychol Med 2013;43:2311–25. [49] Van Os J, McKenna P. Does schizophrenia exist? Institute of Psychiatry, King’s College London;. 2003. [50] Van Os J, Gilvarry C, Bale R, Van Horn E, Tattan T, White I, et al. A comparison of the utility of dimensional and categorical representations of psychosis. Psychol Med 1999;29:595–606. [51] van Os J, Hanssen M, Bijl RV, Ravelli A. Strauss (1969) revisited: a psychosis continuum in the general population? Schizophr Res 2000;45:11–20. [52] van Os J, Linscott RJ, Myin-Germeys I, Delespaul P, Krabbendam L. A systematic review and meta-analysis of the psychosis continuum: evidence for a psychosis proneness-persistence-impairment model of psychotic disorder. Psychol Med 2009;39:179–95. [53] Vollema MG, Vandenbosch RJ. The multidimensionality of schizotypy. Schizophr Bull 1995;21:19–31. [54] Ward TA, Gaynor KJ, Hunter MD, Woodruff PWR, Garety PA, Peters ER. Appraisals and responses to experimental symptom analogues in clinical and non-clinical individuals with psychotic experiences. Schizophr Bull 2014;40:845–55. [55] Wechsler D. Wechsler adult intelligence scale, 3rd Ed., San Antonio, TX: Harcourt Assessment; 1997. [56] Wickham H, Walsh C, Asherson P, Taylor C, Sigmundson T, Gill M, et al. Familiality of symptom dimensions in schizophrenia. Schizophr Res 2001;47:223–32. [57] World Health Organisation. International Classification of Diseases (ICD-10). Geneva: World Health Organisation; 1992. [58] Yung AR, Phillips LJ, McGorry PD, McFarlane CA, Francey S, Harrigan S, et al. Prediction of psychosis – A step towards indicated prevention of schizophrenia. Br J Psychiat 1998;172:14–20. [59] Yung AR, Yuen HP, McGorry PD, Phillips LJ, Kelly D, Dell’Olio M, et al. Mapping the onset of psychosis: the comprehensive assessment of at-risk mental states. Aust Nz J Psychiat 2005;39:964–71. [60] Yung AR, Nelson B, Baker K, Buckby JA, Baksheev G, Cosgrave EM. Psychoticlike experiences in a community sample of adolescents: implications for the continuum model of psychosis and prediction of schizophrenia. Aust Nz J Psychiat 2009;43:118–28.

Please cite this article in press as: Brett CMC, et al. Which psychotic experiences are associated with a need for clinical care? European Psychiatry (2015), http://dx.doi.org/10.1016/j.eurpsy.2014.12.005