Changes in delusions in the early phase of antipsychotictreatment – An experience sampling study

Suzanne Ho-wai So a,n, Emmanuelle Roisin Peters b,c, Joel Swendsen d,e,Philippa Anne Garety b,c,1, Shitij Kapur c,f,g,1

a Department of Psychology, The Chinese University of Hong Kong, Hong Kong SAR, Chinab Department of Psychology, King0s College London, Institute of Psychiatry, London, United Kingdomc National Institute for Health Research (NIHR) Biomedical Research Centre for Mental Health at South London and Maudsley NHS Foundation Trustand King’s College London, United Kingdomd University of Bordeaux, National Center for Scientific Research, Bordeaux, Francee EPHE-La Sorbonne, Paris, Francef Section on Schizophrenia, Imaging and Therapeutics, King’s College London, Institute of Psychiatry, London, United Kingdomg Department of Psychological Medicine, King’s College London, Institute of Psychiatry, London, United Kingdom

a r t i c l e i n f o

Article history:Received 6 September 2013Received in revised form4 December 2013Accepted 16 December 2013Available online 24 December 2013

a b s t r a c t

It has been suggested that different aspects of delusions (conviction, distress, preoccupation) respond totreatment at different rates, and that the cognitive bias of ‘Jumping to Conclusions’ (JTC) may predicttreatment outcome. This study investigates changes in delusion dimensions using Experience SamplingMethodology (ESM) and the role of JTC as a predictor of change during the initial 2 weeks ofantipsychotic treatment on admission to hospital. Sixteen acute patients with delusions were assessedseven times per day for 14 days using computerised ESM. ESM assessed moment-by-momentexperiences of affect, psychotic symptoms, and delusion dimensions. Clinical ratings were completedat baseline, 1 week and 2 weeks later. The 0beads0 task was used to measure JTC at baseline. Delusiondimensions improved over the two weeks of antipsychotic treatment and admission to hospital. Differentdelusional dimensions changed at different rates, with distress and disruption being more responsivethan conviction and preoccupation on both PSYRATS and ESM ratings. Eight out of 16 participantsshowed a JTC bias on the beads task at baseline. Exploratory analyses showed that JTC predicted changesin the ESM ratings of delusion conviction and distress, suggesting that reasoning biases may predicttreatment response.

& 2013 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Factor analyses have shown that the delusional experienceconsists of a number of dimensions, with the majority of studiesidentifying conviction, distress, preoccupation and disruption tolife (Kendler et al., 1983; Garety and Hemsley, 1987; Appelbaumet al., 1999; Peters et al., 2004; Lincoln, 2007). It has beensuggested that these dimensions respond to cognitive therapydifferently (Brett-Jones et al., 1987; Chadwick and Lowe, 1990,1994). However, less is known about how they respond toantipsychotics. Jørgensen (1995) assessed changes in delusiondimensions in 50 patients with schizophrenia for 8 weeks after

readmission. Conviction, extension (i.e. the areas of life affected bythe delusional belief) and pressure (i.e. the degree to which one ispreoccupied with the delusional belief) were found to be distinctlyindependent dimensions of delusions. Seventy per cent of thepatients showed partial remission of delusional beliefs, wherebyreduction in pressure preceded improvement in conviction.Jørgensen (1995) focused on changes after readmission, and theeffect of antipsychotics was not specifically examined. Mizrahiet al. (2006) assessed dimensions of the principal psychoticexperience with 17 patients during the first 10 weeks of anti-psychotic treatment. They reported that behavioural impact andcognitive and emotional preoccupation changed more markedlyand rapidly than conviction. The measures in these studies werenot commonly used, and the dimensions measured were not thesame as those generated in factor analysis studies. Nevertheless,these studies suggest a differential response of delusion dimen-sions to treatment, with conviction being less amenable to changethan distress and preoccupation. This observation is consistentwith patients’ reports that antipsychotics ‘detach’ them from their

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/psychres

Psychiatry Research

0165-1781/$ - see front matter & 2013 Elsevier Ireland Ltd. All rights reserved.http://dx.doi.org/10.1016/j.psychres.2013.12.033

n Correspondence to: Room 321 Wong Foo Yuan Building, Department ofPsychology, The Chinese University of Hong Kong, Shatin, New Territories, HongKong Special Administrative Region, China. Tel.: þ852 39436211;fax: þ852 26035019.

E-mail address: shwso@psy.cuhk.edu.hk (S.-w. So).1 Joint last authors.

Psychiatry Research 215 (2014) 568–573

psychotic symptoms rather than eradicate or eliminate the symp-toms (Mizrahi et al., 2005).

These findings raise the question of why conviction persistswhile other dimensions improve, and what might predict changein conviction. It has been suggested (Garety et al., 2001; 2007) thatthe way patients form appraisals of experience may be charac-terised by reasoning biases, including ‘Jumping to Conclusions’(JTC) bias and lack of belief flexibility, which may contribute to thedevelopment and maintenance of delusions. The JTC bias refers tothe tendency to reach a decision without gathering sufficient data(Garety et al., 1991; Garety and Freeman, 1999). JTC occurs in onehalf to two-thirds of individuals with delusions (see reviews byFine et al., 2007; Freeman, 2007), as well as in people scoringhighly on delusional ideation scales (e.g. Linney et al., 1998;Colbert and Peters, 2002; Moritz and Woodward, 2005; Van Daelet al., 2006; Warman and Martin, 2006) and in people who haveremitted from delusions (Moritz and Woodward, 2005; Peters andGarety, 2006). In some studies, JTC has also been found to beassociated with severity of delusions and delusional conviction(Garety et al., 2005; Peters and Garety, 2006). JTC may therefore bea stable predisposing factor for delusions (see review by So et al.,2010).

Menon et al. (2008) found that JTC at baseline predictedsubsequent change in psychotic symptoms during the initialweeks of antipsychotic treatment, and argued that JTC was amoderator of treatment outcome. However, they measured delu-sions using clinical ratings and did not measure delusion dimen-sions. Treatments targeting reasoning biases (including JTC) haverecently been developed, with preliminary evidence showing areduction in both JTC and delusion conviction (Waller et al., 2011)and distress (Moritz et al., 2011). Therefore, dimensions of delu-sions are important indicators of treatment outcomes.

There is evidence that, like other psychotic symptoms, delu-sions improve most markedly in the early weeks of antipsychotictreatment, and that initial change predicts overall symptomimprovement (Agid et al., 2003; Leucht et al., 2005). Agid et al.(2003) even found that improvement in psychotic symptoms maybe seen in the very first few hours. Mizrahi et al. (2006) found thatthe different pattern of improvement in delusional dimensionswas evident as early as 2 weeks after the start of treatment.Therefore, the very early stage of treatment is a critical period forthe fine-grained analysis of delusion change and is a good test bedfor the moderators and mediators of such change.

Experience Sampling Methodology (ESM), a structured diarytechnique, has been shown to be a well-suited method forassessing moment-to-moment levels and fluctuations of psychoticexperiences (e.g. Myin-Germeys et al., 2001; Peters et al., 2012). Soet al. (2013) recently showed that ESM assessment using aPersonal Digital Assistant (PDA) is feasible and valid during theacute psychotic stage of psychosis. The present study uses thisapproach to investigate changes in delusion dimensions during theinitial 2 weeks of antipsychotic treatment. It was hypothesisedthat delusion distress and preoccupation, but not conviction,would reduce significantly over 2 weeks of antipsychotic treat-ment. This study also aimed to explore the potential role of the JTCbias in predicting delusion change. It was predicted that therewould be an association between JTC bias at baseline and changein delusion dimensions over time.

2. Methods

2.1. Participants

Ethical approval for the study was granted by the South East London ResearchEthics Committee 4 (ref. 10/H0807/44). In-patients with delusions (scoring 4 orabove on at least one of the PANSS delusion items) and a clinical diagnosis (based

on clinical notes) of schizophrenia spectrum disorder or bipolar disorder wererecruited. Patients were recruited and assessed as soon as they were admitted tothe hospital for an acute psychotic episode, and no longer than 2 weeks after thestart of antipsychotic treatment for the current episode. Patients with drug-inducedpsychosis or organic psychosis, and patients with a primary diagnosis of substancemisuse were excluded.

2.2. Measures

2.2.1. Experience sampling assessmentParticipants were asked to complete self-assessment questionnaires on a

Personal Digital Assistant (PDA) seven times a day on 14 consecutive days. Eachquestionnaire consists of the same set of 20 items on affect, psychotic symptoms,delusion dimensions, and current environment. Details of the methodology and itsvalidity are reported in greater detail in a separate paper (So et al., 2013).

There were four items on delusion dimensions in the ESM questionnaire. Thequestions were preceded with the statement: “The following questions concern theproblem discussed with your researcher, in particular the thought or idea that…”

followed by the exact wording of the ‘problem or main concern’ as described by theparticipant. The wording of the dimensions was similar to that used in Peters et al.(2012) – conviction (“At this moment, to what extent do you believe this concern istrue?”), distress (“At this moment, how much does this concern upset you?”),preoccupation (“At this moment, to what extent does this concern go round andround in your mind?”), disruption (“At this moment, to what extent does thisconcern interfere with what you are doing?”). Participants were asked to rate thesame delusional belief throughout the study period, on a 7-point Likert scale (from1, ‘not at all’, to 7, ‘very much’).

2.2.2. Symptomatology and delusion dimensionsIn addition to ESM, psychotic symptoms at baseline and 2 weeks were

measured using three well-validated symptom scales. The Positive and NegativeSyndrome Scale (PANSS; Kay et al., 1987) consists of a 7-item positive symptomscale, a 7-item negative symptom scale, a 16-item general psychopathology scale,and a total score. Each symptom is rated on a 1–7 scale, with the total score rangingfrom 30 to 210. The Schedule for the Assessment of Positive Symptoms (SAPS;Andreasen, 1984) is a 35-item interview-based scale covering 32 positive symp-toms of schizophrenia, organised into four areas (hallucinations, delusions, bizarrebehaviour and positive formal thought disorder). Each area includes ratings ofspecific symptoms and a global rating (0–5), and total scores range from 0 to 175.The Psychotic Symptom Rating Scales (PSYRATS; Haddock et al., 1999) consists ofan 11-item auditory hallucinations scale and a 6-item delusions scale, measuringmultiple dimensions of the psychotic symptoms. The delusions scale of PSYRATSassesses four dimensions (conviction, amount and intensity of distress, amount andduration of preoccupation, and disruption to life). Items are rated by theinterviewer on a 0–4 ordinal scale. In the present study, ratings of the two distressitems and the two preoccupation items were averaged, yielding a distress score anda preoccupation score respectively.

The inclusion of three symptom scales constitutes a comprehensive assessmentof symptom change. While PANSS provides overall scores for positive, negative andgeneral symptoms, SAPS is a more detailed measure of positive symptoms anddimensional ratings of delusions are unique to PSYRATS.

2.2.3. Jumping to conclusions reasoning biasThe beads task was designed to examine individuals’ data-gathering reasoning

style (Phillips and Edwards, 1966; Garety et al., 1991). In the beads task, individualsare presented with two jars, each containing 100 coloured beads. One jar contains60 beads of one colour (e.g. black) and 40 beads of another (e.g. yellow), while theother jar contains beads in opposite proportions (i.e. 40 black and 60 yellow;Dudley et al., 1997). The jars are then removed from view. Upon request from theparticipant, beads are presented, one at a time, from one of the jars in a seeminglyrandom (but in fact predetermined) order. All the beads drawn are replaced so theproportions of the coloured beads stay the same. Participants can view as manybeads as they want until they decide with certainty from which jar the beads aredrawn. The key variable is the number of beads requested before making adecision, with two beads or fewer classified as a ‘jumping to conclusions’ bias(Garety et al., 2005). The beads task was presented on a laptop computer at thebaseline interview.

2.3. Procedures

All the clinical rating scales were administered by the first author, a clinicalpsychologist trained by experts in the use of these instruments. At the baselineinterview, clinical rating scales were completed. The researcher agreed with theparticipant on the index delusional belief and programmed the individualisedquestionnaire onto a Palm Tungsten E2 PDA (Palm OSs version 5.2.1). Participantswere given the PDA to carry with them and the ESM assessments beganimmediately. The PDA was programmed to emit a beep signal at seven randommoments during the day, and participants were asked to respond to the signals as

S.H.-w. So et al. / Psychiatry Research 215 (2014) 568–573 569

soon as possible. In order to facilitate compliance with the procedure, theresearcher contacted the participant at least twice in the first week to offersupport. Individuals who demonstrated difficulty in understanding assessmentquestions or operating the device were given additional training.

2.4. Statistical analysis

Statistical analysis for this study involved clinical data at two time points andESM data across 14 days, capturing both within- and between-individual differ-ences. The maximum number of potential signals a participant would receive was98 (seven signals per day for 14 days). Only participants who completed at least 30experience sampling activations are included in the analyses.

The ESM data sets contain repeated observations nested within participants.Since observations from the same individual are more similar than observationsfrom different individuals, the residuals are not independent. Therefore, long-itudinal relationships of variables across levels (day and person) were evaluatedusing multilevel linear regression modelling (Goldstein, 1987; Snijders and Bosker,1999), which is a standard method for analysing ESM data (Kimhy et al., 2012). Allresponses were considered in all regression analyses reported. Regression modelswere tested using the multilevel XTREG command in STATA 12 (StataCorp, 2012).Whenever mean levels of ESM scores are presented (e.g. in graphs), scores werecalculated for each participant and then averaged.

The first set of multilevel modelling analysis aimed to investigate the mainhypothesis relating to changes in psychotic symptoms and delusion dimensionsover time. Separate regression models were estimated with each of the symptoms/dimensions as Dependent Variables (DV) respectively and day as an IndependentVariable (IV). For the secondary, exploratory hypothesis relating to the potentialmoderating effect of JTC on outcome, participants were grouped on the basis oftheir performance on the beads task. To test the effect of baseline JTC on changes indelusion dimensions over time, separate multilevel linear regression models wereestimated with JTC, day, and JTC�day interaction as IVs, and the level of eachdelusion dimension (ESM ratings) as DV respectively.

Unilevel analyses of clinical ratings were performed on the IBM SPSS Statisticsfor Windows, Version 19.0 (IBM Corp. Released (2010)). Group differences in thelevel and changes in delusions were tested using Mann–Whitney U tests andwithin-person changes were tested using paired-sample t-tests.

3. Results

3.1. Demographic and clinical data

A total of 26 patients consented to participate in this study,among whom 16 completed at least 30 experience samplingassessments. The mean number of entries per participant was 59(range 34–89). The mean rate of compliance was 70.7% (range40.2–94.6%). The total number of observations available for multi-level models was 1306. There was no significant differencebetween the 16 participants who met the minimum compliancerequirement and the ten participants who did not in age, durationand dosage of medication, number of admissions, or any of theclinical rating scores (p40.10).

Among the 16 completers, nine (56.25%) were male and seven(43.75%) were female. Mean age was 36.1 years (range 20–63years). Psychiatric diagnoses from the case notes were available for15 patients as follows: schizophrenia (n¼5), unspecified psychosis(n¼5), mood disorder with psychotic symptoms (n¼2), schizoaf-fective disorder (n¼1), delusional disorder (n¼1), and acute andtransient psychotic disorder (n¼1). The average number of admis-sions (including the current one) was 1.73 (S.D.¼1.31; range 1–6).Eleven patients (68.8%) were in their first episode of psychosis. Theaverage number of days on antipsychotics at the time of the firstassessment was 4.56 (S.D.¼2.45; range 0–8). The average baselinerating of delusion on the SAPS was four (S.D.¼0.52), indicating amoderate to marked level of severity (Andreasen, 1984). Medica-tion types were as follows: 15 received atypical antipsychotics(Risperidone, Olanzapine, Aripiprazole, Quetiapine, Amisulpiride,and Clozapine), while one received a typical antipsychotic (Piportil).The mean starting dose of antipsychotics in chlorpromazine equiva-lents (Andreasen et al., 2010) was 131.86 mg/d (S.D.¼94.94).

Pattern of extreme responding was checked using ESM ratingsof the key variables (i.e. delusion dimensions). Percentage of

extreme responses (a score of 1 or 7 on a 7-point scale) over 14days was not significantly correlated with age (p¼0.54). However,female participants endorsed a significantly higher percentage ofextreme responses on all four delusion dimensions compared tomale participants (po0.01). Three (out of 16) participants hadextreme responses (either 1 or 7 on the 7-point scale) for allprompts across 14 days on any of the key variables (i.e. delusiondimensions). However, all three participants made variedresponses on other variables. Therefore, we did not exclude dataof any participant due to extreme responding. As a standardprocedure to control for potential confounds, all analyses adjustedfor the effects of age and gender (e.g. Swendsen et al., 2011).

3.2. Jumping to conclusions (JTC) at baseline

Fourteen participants completed the beads task with a meannumber of draws to decision of 5 (range 1–20). Eight participants(57.1%) showed the JTC bias. There was no significant difference atbaseline between the JTC (n¼8) and no-JTC (n¼6) groups onPANSS scale scores, SAPS delusion score, or PSYRATS delusion andconviction scores (see Table 1 for means and S.D.s).

3.3. Changes in psychotic symptoms over 2 weeks

As shown in Table 2, there was a significant improvement inPANSS positive, general and total scores between baseline and

Table 1Mean levels (S.D.) of symptom ratings at baseline in the JTC and no-JTC groups.

JTC group(n¼8)

No-JTC group(n¼6)

Mann–WhitneyU

PANSS positive 22.50 (4.87) 18.83 (2.93) U¼36.00,p¼0.14

PANSS negative 12.75 (5.42) 12.33 (5.50) U¼24.50,p¼1.00

PANSS general 31.38 (11.05) 35.67 (11.20) U¼17.50,p¼0.41

PANSS total 66.63 (17.80) 66.83 (15.99) U¼23.50,p¼0.95

SAPS delusions 4.13 (0.35) 4.00 (0.63) U¼26.50,p¼0.76

PSYRATS delusions 18.63 (2.97) 15.17 (5.00) U¼32.00,p¼0.35

PSYRATS conviction 3.38 (0.52) 3.33 (0.82) U¼23.50,p¼0.95

PSYRATS distress 3.56 (0.42) 2.17 (1.69) U¼34.50,p¼0.18

PSYRATSpreoccupation

2.50 (0.96) 2.08 (1.07) U¼31.50,p¼0.35

PSYRATS disruption 3.13 (0.64) 3.33 (0.52) U¼20.00,p¼0.66

Table 2Mean levels (S.D.) of symptom ratings at baseline and week 2 (N¼16).

Baseline Week 2 Paired-sample t-test

PANSS positive 20.63 (4.52) 16.25 (3.72) t¼4.01, po0.01PANSS negative 11.88 (5.24) 10.00 (3.58) t¼1.57, p¼0.17PANSS general 33.25 (10.51) 24.75 (5.16) t¼3.78, po0.01PANSS total 65.75 (16.10) 51.00 (10.91) t¼4.06, po0.01SAPS delusions 4.00 (0.52) 3.13 (1.15) t¼3.22, po0.01PSYRATS delusions 17.25 (3.92) 13.69 (3.63) t¼3.20, po0.01PSYRATS conviction 3.38 (0.62) 3.25 (1.07) t¼0.46, p¼0.65PSYRATS distress 3.03 (1.24) 2.13 (1.38) t¼2.52, p¼0.02PSYRATS preoccupation 2.31 (0.93) 1.91 (0.69) t¼1.71, p¼0.11PSYRATS disruption 3.19 (0.54) 2.38 (1.03) t¼3.57, po0.01

S.H.-w. So et al. / Psychiatry Research 215 (2014) 568–573570

week 2, as well as in delusions as assessed on the SAPS andPSYRATS.

3.3.1. Delusional distress and preoccupation but not conviction willreduce significantly over 2 weeks of antipsychotic treatment

As shown in Table 2, paired-sample t-tests revealed a signifi-cant reduction in PSYRATS ratings of Distress and Disruption.Change in Conviction or Preoccupation was not significant.

Changes in ESM self-ratings of delusion dimensions are shownin Fig. 1. Multilevel linear regression models showed a significantdecrease in Distress (B¼�0.04, SE¼0.01, po0.01) and Disruption(B¼�0.03, SE¼0.01, p¼0.01), but not in Conviction or Preoccupa-tion (p40.10). In order to examine whether reduction indelusional dimensions was more marked in the earlier days thanin the subsequent days, models with a quadratic term for time(i.e. squared Day) were tested for each delusional dimension. Thequadratic effect of time was not significant (p40.10), suggesting alinear decrease in distress and disruption across the first 14 days.

3.3.2. There will be an association between JTC bias at baselineand change in delusion dimensions over time

Changes in the ESM delusion dimensions by JTC groups arepresented in Fig. 2. In the model with ESM conviction as DV, therewas no main effect of JTC on conviction (B¼0.79, SE¼1.00,p¼0.43) or distress (B¼1.02, SE¼0.83, p¼0.22). However, therewas a significant interaction between JTC and Day (B¼0.10,SE¼0.03, po0.001), indicating that conviction change over timewas different between the two groups, with a decrease in convic-tion in the no-JTC group only (See Fig. 2a). The interactionbetween JTC and Day was also significant for distress (B¼0.06,SE¼0.03, p¼0.02), indicating that distress change over time wasdifferent between the groups, with apparently little change indistress in the JTC group and a small decrease in the no-JTC group(see Fig. 2b).

For preoccupation and disruption there was no significantinteraction between JTC and Day (p40.10). When the interactionterm was taken out of the models, there was a significant maineffect of JTC on level of preoccupation (B¼1.97, SE¼0.91, p¼0.03),with higher preoccupation scores being found in the JTC group(see Fig. 2c). The main effect did not reach significance fordisruption (p40.05; see Fig. 2d).

In summary, the no-JTC group were less preoccupied with theirdelusions overall, and showed a greater improvement in convic-tion and distress over time than the JTC group. JTC did not predictchanges in levels of preoccupation and disruption.

4. Discussion

This study investigated changes over time in delusions among16 inpatients as they began antipsychotic treatment on admissionto hospital, as well as exploring the role of the ‘Jumping toConclusions’ (JTC) bias in predicting delusion change. The mainfindings are that (i) distress and disruption, but not conviction and

1

2

3

4

5

1 2 3 4 5 6 7 8 9 10 11 12 13 14Day

ESM

leve

l (1-

7)

Conviction Preoccupation Distress Disruption

Fig. 1. Changes in delusion dimensions on ESM (N¼16).

1

2

3

4

5

6

7

1 2 3 4 5 6 7 8 9 10 11 12 13 141

2

3

4

5

6

7

1 2 3 4 5 6 7 8 9 10 11 12 13 14

1

2

3

4

5

6

7

1 2 3 4 5 6 7 8 9 10 11 12 13 141

2

3

4

5

6

7

1 2 3 4 5 6 7 8 9 10 11 12 13 14

ESM

(1-7

)

ESM

(1-7

)

ESM

(1-7

)

ESM

(1-7

)

Day Day

Day Day

Delusion conviction Delusion distress

Delusion preoccupation Delusion disruption

Fig. 2. Change over time in delusional dimensions by baseline JTC (N¼14).

S.H.-w. So et al. / Psychiatry Research 215 (2014) 568–573 571

preoccupation, improved on PSYRATS and ESM; and (ii) the JTCbias predicted a difference in change over time in both convictionand distress measured by ESM.

Our sample, which consisted mostly of first-episode patients,showed a rapid improvement in clinical symptoms, with PANSSscores and delusion ratings reducing by more than 20% over2 weeks. This result is consistent with the ‘early onset hypothesis’of antipsychotic action (Agid et al., 2003; Leucht et al., 2005).When assessed multi-dimensionally, both PSYRATS and ESMshowed that the improvement was in delusion distress anddisruption, and not conviction and preoccupation. Although gra-phical representation of ESM ratings seemed to also indicate somedecrease in conviction and preoccupation (Fig. 1), their lack ofsignificance may have been due to their greater variability overtime, while distress and disruption showed a steady (linear)decline. Although the lack of improvement in preoccupation wasnot expected, our finding is consistent with Mizrahi et al. (2006)that conviction does not change reliably in the early phase ofantipsychotic treatment. Our finding is also consistent with thepatients’ perspective that antipsychotics help them cope betterwith, or become less distressed by, their psychotic symptoms,rather than eliminate the symptoms (Mizrahi et al., 2005).

Symptom ratings and multi-dimensional ratings revealed arather different picture of the process of recovery, wherebydelusions were rated as reduced in severity (on the clinical scales)even though the strength of the belief (i.e. conviction) remainedrelatively unchanged on both clinical scales and ESM, suggestingthat (i) clinical ratings were largely affected by the affectivedimensions of delusions, and that (ii) multi-dimensional assess-ment of delusions provides additional and important informationabout recovery that is not represented in the usual clinical ratings.

The ESM data revealed an interesting relationship between JTCand delusion dimensions. Conviction and distress declined in theno-JTC group but not in the JTC group. Preoccupation ratings werehigher overall in the JTC group, but JTC did not predict change ineither preoccupation or disruption. These findings suggest that JTCis associated with introspection and rumination about one’sdelusions, contributing to persistence of delusions, at least interms of conviction and distress (Garety et al., 2005; Menonet al., 2008). While the relationship between JTC and convictionhas been replicated and discussed in other studies (Garety et al.,2005; Fine et al., 2007), it is the first time that such a link wasdemonstrated using a moment-by-moment diary measure. Togetherwith data from reasoning training studies (Moritz et al., 2011; Walleret al., 2011), our finding provides further evidence that the JTC data-gathering style is a moderator of treatment response of delusiondimensions. A recent study found evidence for a correspondencebetween change in JTC and change in delusions following cognitivebehavioural therapy (Sanford et al., 2013). However, the currentstudy measured JTC at baseline only and did not investigate the roleof JTC as a mediator of treatment outcome.

There are several limitations to this study. Firstly, the samplesize was adapted for analysis using the ESM data (assessed atmultiple time points) but was underpowered for group compar-isons on JTC. It is therefore possible that the lack of interactionsfound on the preoccupation and disruption dimensions may havebeen due to a lack of power. However, the fact that significantfindings were obtained for both conviction and distress despite thesmall sample size suggests that they are large effects, although itcannot be concluded that JTC is a moderator of these dimensionsexclusively.

Secondly, patients were not sub-divided based on the type ofantipsychotics they were taking, or their diagnosis. It is possiblethat medication or diagnosis may have been confounding vari-ables. Therefore, findings should be interpreted with caution andreplication of the novel findings is needed.

Lastly, we cannot determine conclusively whether the changesthat were captured are entirely due to a response to the pharma-cological intervention received by the patients. All participantswere recruited at the start of a hospital admission, which offers arange of interventions from a multidisciplinary team. While it isunlikely that psychological interventions would have been offered,or made an impact, in the short time-span of the study, we cannotbe clear on how much of the changes were a result of nursing oroccupational therapy input, or just the removal from a stressfulsituation.

Despite these caveats, it is to our knowledge the first study tohave measured the response of psychotic symptoms in the acutestage using ESM and provides confirmation that aspects of delu-sions improve within the first 2 weeks of admission to hospital,even when viewed from the patient’s experiential perspective, andprovides new evidence that JTC may be a predictor of treatmentresponse.

Acknowledgements

This work was supported by the Croucher Foundation Scholar-ship and the University of London Central Research Fund to SHS.SK receives Grant support from AstraZeneca, Bristol-Myers Squibb(BMS) and Glaxo Smith Kline. He is a consultant/speaker/advisorfor the following companies: AstraZeneca, Bioline, Bristol MeyersSquibb, Eli Lilly, Janssen (Johnson and Johnson), Lundbeck, Otsuka,Organon, Pfizer, Servier, Solvay Wyeth. PG, SK and EP acknowledgesupport for some clinical sessions from the National Institute ofHealth Research (NIHR) Biomedical Research Centre for MentalHealth at the South London and Maudsley NHS Foundation Trustand the Institute of Psychiatry, King’s College London.

References

Agid, O., Kapur, S., Arenovich, T., Zipursky, R.B., 2003. Delayed-onset hypothesis ofantipsychotic action: a hypothesis tested and rejected. Archives of GeneralPsychiatry 60, 1228–1235.

Andreasen, N.C., 1984. Scale for the Assessment of Positive Symptoms (SAPS).Department of Psychiatry, University of Iowa College of Medicine, Iowa City.

Andreasen, N.C., Pressler, M., Nopoulos, P., Miller, D., Ho, B.C., 2010. Antipsychoticdose equivalents and dose-years: a standardized method for comparingexposure to different drugs. Biological Psychiatry 67, 255–262.

Appelbaum, P.S., Robbins, P.C., Roth, L.H., 1999. Dimensional approach to delusions:comparison across types and diagnoses. American Journal of Psychiatry 156,1938–1943.

Brett-Jones, J., Garety, P.A., Hemsley, D., 1987. Measuring delusional experiences: amethod and its application. British Journal of Clinical Psychology 26, 257–265.

Chadwick, P.D., Lowe, C.F., 1990. Measurement and modification of delusionalbeliefs. Journal of Consulting and Clinical Psychology 58, 225–232.

Chadwick, P.D., Lowe, C.F., 1994. A cognitive approach to measuring and modifyingdelusions. Behaviour Research and Therapy 32, 355–367.

Colbert, S.M., Peters, E.R., 2002. Need for closure and jumping-to-conclusions indelusion-prone individuals. Journal of Nervous and Mental Disease 190, 27–31.

Dudley, R.E.J., John, C.H., Young, A.W., Over, D.E., 1997. The effect of self-referentmaterial on the reasoning of people with delusions. British Journal of ClinicalPsychology 36, 575–584.

Fine, C., Gardner, M., Craigie, J., Gold, I., 2007. Hopping, skipping or jumping toconclusions? Clarifying the role of the JTC bias in delusions. CognitiveNeuropsychiatry 12, 46–77.

Freeman, D., 2007. Suspicious minds: the psychology of persecutory delusions.Clinical Psychology Review 27, 425–457.

Garety, P.A., Bebbington, P., Fowler, D., Freeman, D., Kuipers, E., 2007. Implicationsfor neurobiological research of cognitive models of psychosis: a theoreticalpaper. Psychological Medicine 37, 1377–1391.

Garety, P.A., Freeman, D., 1999. Cognitive approaches to delusions: a critical reviewof theories and evidence. British Journal of Clinical Psychology 38, 113–154.

Garety, P.A., Freeman, D., Jolley, S., Dunn, G., Bebbington, P.E., Fowler, D.G., Kuipers,E., Dudley, R., 2005. Reasoning, emotions, and delusional conviction inpsychosis. Journal of Abnormal Psychology 114, 373–384.

Garety, P.A., Hemsley, D.R., 1987. Characteristics of delusional experience. EuropeanArchives of Psychiatry and Neurological Sciences 236, 294–298.

Garety, P.A., Hemsley, D.R., Wessely, S., 1991. Reasoning in deluded schizophrenicand paranoid patients: biases in performance on a probabilistic inference task.Journal of Nervous and Mental Disease 179, 194–201.

S.H.-w. So et al. / Psychiatry Research 215 (2014) 568–573572

Garety, P.A., Kuipers, E., Fowler, D., Freeman, D., Bebbington, P.E., 2001. A cognitivemodel of the positive symptoms of psychosis. Psychological Medicine 31,189–195.

Goldstein, H., 1987. Multilevel Models in Educational and Social Research. CharlesGriffin, London.

Haddock, G., McCarron, J., Tarrier, N., Faragher, E.B., 1999. Scales to measuredimensions of hallucinations and delusions: the psychotic symptom ratingscales (PSYRATS). Psychological Medicine 29, 879–889.

IBM Corp. Released, 2010. IBM SPSS Statistics for Windows, Version, 19. IBM Corp,Armonk, NY p. 0.

Jørgensen, P., 1995. Recovery and insight in schizophrenia. Acta PsychiatricaScandinavica 92, 436–440.

Kay, S.R., Opler, L.A., Fiszbein, A., 1987. Positive and Negative Syndrome Scale (PANSS)Rating Manual. Social and Behavioral Sciences Documents, San Rafael, CA.

Kendler, K.S., Glazer, W.M., Morgenstern, H., 1983. Dimensions of delusionalexperience. American Journal of Psychiatry 140, 466–469.

Kimhy, D., Palmier-Clauss, J., Myin-Germeys, I., Swendsen, J., 2012. Mobile assess-ment in schizophrenia: a research guide. Schizophrenia Bulletin 38, 386–395.

Leucht, S., Busch, R., Hamann, J., Kissling, W., Kane, J.M., 2005. Early-onsethypothesis of antipsychotic drug action: a hypothesis tested, confirmed andextended. Biological Psychiatry 57, 1543–1549.

Lincoln, T.M., 2007. Relevant dimensions of delusions: continuing the continuumversus category debate. Schizophrenia Research 93, 211–220.

Linney, Y.M., Peters, E.R., Ayton, P., 1998. Reasoning biases in delusion-proneindividuals. British Journal of Clinical Psychology 37, 285–302.

Menon, M., Mizrahi, R., Kapur, S., 2008. 0Jumping to conclusions0 and delusions inpsychosis: relationship and response to treatment. Schizophrenia Research 98,225–231.

Mizrahi, R., Bagby, R.M., Zipursky, R.B., Kapur, S., 2005. How antipsychotics work:the patients0 perspective. Progress in Neuro-Psychopharmacology and Biologi-cal Psychiatry 29, 859–864.

Mizrahi, R., Kiang, M., Mamo, D.C., Arenovich, T., Bagby, R.M., Zipursky, R.B., Kapur,S., 2006. The selective effect of antipsychotics on the different dimensions ofthe experience of psychosis in schizophrenia spectrum disorders. Schizophre-nia Research 88, 111–118.

Moritz, S., Kerstan, A., Veckenstedt, R., Randjbar, S., Vitzthum, F., Schmidt, C., Heise,M., Woodward, T.S., 2011. Further evidence for the efficacy of a metacognitivegroup training in schizophrenia. Behaviour Research and Therapy 49, 151–157.

Moritz, S., Woodward, T.S., 2005. Jumping to conclusions in delusional and non-delusional schizophrenic patients. British Journal of Clinical Psychology 44,193–207.

Myin-Germeys, I., Nicolson, N.A., Delespaul, P.A.E.G., 2001. The context of delu-sional experiences in the daily life of patients with schizophrenia. PsychologicalMedicine 31, 489–498.

Peters, E., Garety, P., 2006. Cognitive functioning in delusions: a longitudinalanalysis. Behaviour Research and Therapy 44, 481–514.

Peters, E., Lataster, T., Greenwood, K., Kuipers, E., Scott, J., Williams, S., Garety, P.,Myin-Germeys, I., 2012. Appraisals, psychotic symptoms and affect in daily life.Psychological Medicine 42, 1013–1023.

Peters, E., Joseph, S., Day, S., Garety, P., 2004. Measuring delusional ideation: the 21-item Peters et al. Delusions Inventory (PDI). Schizophrenia Bulletin 30,1005–1022.

Phillips, L.D., Edwards, W., 1966. Conservatism in a simple probability inferencetask. Journal of Experimental Psychology 72, 346–354.

Sanford, N., Lecomte, T., Leclerc, C., Wykes, T., Woodward, T., 2013. Change injumping to conclusions linked to change in delusions in early psychosis.Schizophrenia Research 147, 207–208.

Snijders, T., Bosker, R., 1999. Multilevel Analysis: An Introduction to Basic andAdvanced Multilevel Modelling. Sage Publications, London.

So, S.H., Garety, P.A., Peters, E.R., Kapur, S., 2010. Do antipsychotics improvereasoning biases? A review. Psychosomatic Medicine 72, 681–693.

So, S.H., Peters, E.R., Swendsen, J., Garety, P.A., Kapur, S., 2013. Detecting improve-ments in acute psychotic symptoms using experience sampling methodology.Psychiatry Research 210 (1), 8–88.

StataCorp, 2012. Stata Release, 12. StataCorp LP, College Station, Texas p. 1.Swendsen, J., Ben-Zeev, D., Granholm, E., 2011. Real-time electronic ambulatory

monitoring of substance use and symptom expression in schizophrenia.American Journal of Psychiatry 168, 202–209.

Van Dael, F., Versmissen, D., Janssen, I., Myin-Germeys, I., van Os, J., Krabbendam, L.,2006. Data gathering: biased in psychosis? Schizophrenia Bulletin 32, 341–351.

Waller, H., Freeman, D., Jolley, S., Dunn, G., Garety, P., 2011. Targeting reasoningbiases in delusions: a pilot study of the Maudsley Review Training Programmefor individuals with persistent, high conviction delusions. Journal of BehaviorTherapy and Experimental Psychiatry 42, 414–421.

Warman, D.M., Martin, J.M., 2006. Jumping to conclusions and delusion proneness:the impact of emotionally salient stimuli. Journal of Nervous and MentalDisease 194, 760–765.

S.H.-w. So et al. / Psychiatry Research 215 (2014) 568–573 573