Acting on persecutory delusions: The importance of safety seeking

Transcription

1 Behaviour Research and Therapy 45 (2007) Acting on persecutory delusions: The importance of safety seeking Daniel Freeman a,, Philippa A. Garety a, Elizabeth Kuipers a, David Fowler b, Paul E. Bebbington c, Graham Dunn d a Department of Psychology, Institute of Psychiatry, King s College London, P.O. Box 77, University of London, Denmark Hill, London SE5 8AF, UK b School of Medicine, Health Policy and Practice, University of East Anglia, UK c Department of Mental Health Sciences, Royal Free and University College Medical School, University College London, University of London, UK d Biostatistics Group, Division of Epidemiology & Health Sciences, University of Manchester, UK Received 28 April 2005; received in revised form 13 January 2006; accepted 26 January 2006 Abstract Objective: Acting on delusions is a significant clinical issue. The concept of safety behaviours actions carried out with the intention of reducing perceived threat provides a new way of understanding acting on delusions. A study was conducted with the aim of examining the prevalence and correlates of safety behaviours related to persecutory delusions. Method: One hundred patients with persecutory delusions were assessed for safety behaviours, acting on delusions, anxiety, depression, and psychotic symptoms. Case note data were collected on instances of serious violence or suicide attempts. Results: Ninety-six patients had used safety behaviours in the last month. Greater use of safety behaviours was associated with higher levels of distress. A history of violence or suicide attempts was associated with greater use of safety behaviours. Safety behaviours were significantly associated with acting on delusions, but not with the negative symptoms of psychosis. Conclusion: Safety behaviours are a common form of acting on persecutory delusions. These behaviours have the consequence that they are likely to prevent the processing of disconfirmatory evidence and will therefore contribute to delusion persistence. r 2006 Elsevier Ltd. All rights reserved. Keywords: Schizophrenia; Psychosis; Delusions; Persecutory; Cognitive Introduction In the 1990s an empirical literature emerged indicating that acting on delusions is common (Applebaum, Robbins, & Roth, 1999; Buchanan et al., 1993; Wessely et al., 1993). It was found that persecutory delusions Corresponding author. address: d.freeman@iop.kcl.ac.uk (D. Freeman) /$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi: /j.brat

2 90 D. Freeman et al. / Behaviour Research and Therapy 45 (2007) are those most likely to be acted upon, that the actions are seldom violent, and that the responses are associated with negative emotions such as feeling sad or fearful. However, little is understood about why people act on delusions. One potential route to understanding acting on persecutory delusions is through the concept of safety behaviours. At the heart of persecutory delusions are threat beliefs (Freeman, Garety, Kuipers, Fowler, & Bebbington, 2002; Freeman & Garety, 2004). Individuals believe that they are to suffer physical, social, or psychological harm. Many instances of acting on persecutory delusions may be anxious attempts to seek safety and prevent the perceived threat from occurring. Thus, safety behaviours are acting on delusions, but can be distinguished from the totality of such actions because of their intent. In this article, we investigate the prevalence and correlates of safety seeking in 100 individuals with persecutory delusions, the relationship with an established measure of acting on delusions, and whether the withdrawal often used as a safety behaviour is associated with negative psychotic symptoms. Safety behaviours Individuals who feel threatened often carry out actions designed to prevent their feared catastrophe from occurring; this has been termed safety behaviour (Salkovskis, 1991). When the perceived threat is a misperception, such as in anxiety disorders and paranoia, there are important consequences. Individuals fail to attribute the absence of catastrophe to the incorrectness of their threat beliefs. Rather, they believe that the threat was averted only by their safety behaviours (e.g. The reason I wasn t attacked was because I left the street in time and made it back home ). What are actually instances of the incorrectness of threat beliefs are instead turned into near misses. Threat beliefs are likely to persist partly due to this failure to obtain and process disconfirmatory evidence. Manipulation studies have tested the idea that safety behaviours maintain anxiety disorders (e.g. Salkovskis, Clark, Hackmann, Wells, & Gelder, 1999; Sloan & Telch, 2002; Wells et al., 1995). It has been found that exposure plus decreased use of safety behaviours leads to greater reductions in threat beliefs and anxiety than exposure alone, consistent with the maintenance hypothesis. The concept of safety behaviours was developed in cognitive accounts of anxiety disorders (e.g. Clark, 1999; Salkovskis, Clark, & Gelder, 1996), but has since been applied to persecutory delusions (e.g. Morrison, 1998). Freeman, Garety, and Kuipers (2001) used a semi-structured interview the Safety Behaviours Questionnaire with 25 individuals with current persecutory delusions. It was found that all of the individuals had used safety behaviours in the past month. The most common type of safety behaviour was avoidance. For example, people would avoid going to the local shops or on buses where they feared attack. Apart from avoiding the situations perceived as most dangerous, the individuals also carried out actions to lessen the threat directly. When they felt they were in imminent danger they sought protection (e.g. would only leave the home with a trusted person), took steps to decrease their visibility (e.g. alternated routes and the time of return home), enhanced their vigilance (e.g. looking up and down the street), or acted as if they would resist attack (e.g. prepared to strike out). Further, a smaller proportion of people would try to comply with their persecutors (e.g. trying to do things that they thought the persecutors wanted them to do such as keeping the television volume low) or adopt the opposite behaviour of confronting them (e.g. shouting at neighbours). Greater use of safety behaviours was associated with higher levels of anxiety. The safety seeking appeared to be motivated by fear. The main aim of the current study was to examine the presence and correlates of safety behaviours in a larger sample of individuals with persecutory delusions using the Safety Behaviours Questionnaire. It was predicted that safety behaviours would be present in the majority of individuals with persecutory delusions and would be associated with higher levels of emotional distress. The second aim of the study was to assess the degree of overlap between acting on delusions, as assessed by the main instrument used in the 1990s, the Maudsley Assessment of Delusions Schedule (Wessely et al., 1993), and safety behaviours as assessed by the Safety Behaviours Questionnaire (Freeman et al., 2001). The MADs assesses a broad range of actions associated with delusions, including, for example, whether the person has written to anyone, whether they have lost their temper, and whether their belief has stopped them from watching television or listening to the radio. We predicted significant associations between the

3 two measures. Further, we wished to look in closer detail at two particularly important clinical groups, those with a history of violence and those with a history of suicide or self-harm attempts (e.g. Verdoux et al., 2001), to examine whether they might be more likely to act on their delusions and use safety behaviours. We did not attempt to understand violence or suicide attempts as responses to delusions (see Appelbaum, Robbins, & Monahan, 2000; Walsh et al., 2004). Safety behaviours and the negative symptoms of psychosis A number of factors have been identified as causes (or confounders) of negative symptoms in psychosis. These include, for example, cognitive impairment (e.g. O Leary et al., 2000), depression (e.g. Malla et al., 2002), side effects of medication (e.g. Kelley, van Kammen, & Allen, 1999), and understimulation (e.g. Wing & Brown, 1970). There has long been debate about the relationship between the positive and negative symptoms of psychosis (e.g. Wing, 1989). One argument is that, in some cases, negative symptoms are a secondary consequence of positive symptoms (e.g. Carpenter, Heinrichs, & Alphs, 1985). One plausible route that this might occur is via safety behaviours. Clearly, safety-seeking avoidance, often of interpersonal situations, may present as a loss of interest in activities (anhedonia-asociality) or as physical inactivity (avolition-apathy). Anxious vigilance may create inattention, and compliance with persecutors demands may result in poverty of speech (alogia). Further, the use of withdrawal as a safety mechanism may contribute to negative symptom development via understimulation, analogous to the effects of institutionalisation. The third aim of the current study was to assess the potential relationship between negative symptoms as assessed by the Scale for the Assessment of Negative Symptoms (SANS; Andreasen, 1984b) and safety behaviours. It was predicted that greater use of safety behaviours would be associated with higher levels of negative symptoms. Method Participants D. Freeman et al. / Behaviour Research and Therapy 45 (2007) Individuals with current persecutory delusions participated in the present study, drawn from the first cohort recruited for the Psychological Prevention of Relapse in Psychosis (PRP) Trial (ISRCTN ). The PRP Trial is a UK multi-centre randomised controlled trial of cognitive behaviour therapy and family intervention for psychosis. It is based in four National Health Service Trusts in London and East Anglia. It was designed to answer questions both about outcome and the psychological processes associated with psychosis. Planned studies of psychological processes in psychosis were incorporated into the baseline assessment of participating patients before randomisation into the trial. Recruitment was from specified clinical teams, both inpatient and outpatient services, within each of these trusts. These clinical teams were chosen after agreement that all patients who met the eligibility criteria would be asked to participate in the trial. The aim was to enable recruitment of a representative sample of individuals with psychosis by reducing referral biases. Both inpatient and outpatient services were approached at least fortnightly for patients with psychosis who were relapsing. Patients meeting the eligibility criteria were asked to provide written informed consent for participation by a trial research worker or research clinical psychologist. The inclusion criteria were: a current diagnosis of non-affective psychosis (schizophrenia, schizoaffective psychosis, delusional disorder); age years; a second or subsequent episode starting not more than 3 months before consent to enter the trial; and at least one positive psychotic symptom at first time of meeting. The exclusion criteria were: primary diagnosis of alcohol or substance dependency, organic syndrome or learning disability; inadequate command of English to engage in psychological therapy; unstable residential arrangements. The participants formed the first cohort of 100 individuals with current persecutory delusions (SAPS Persecutory Delusions Mild or above) that consented to complete the Safety Behaviours Questionnaire for the current study in addition to the key trial outcome measures. This was 61% of those eligible (100/165). Those who participated in this study were compared with those who did not: there were no significant differences in age, sex, length of illness, or number of relapses (p4:1).

4 92 D. Freeman et al. / Behaviour Research and Therapy 45 (2007) Measures Safety Behaviours Questionnaire Persecutory Delusions (SBQ) (Freeman et al., 2001). The SBQ is a semistructured interview assessing safety behaviours used in the last month. An action is deemed a safety behaviour if the interviewee reports that it has been carried out with the intention of reducing persecutory threat. Seven types of safety behaviours are assessed: avoidance of situations that the person believes would be dangerous to enter (Avoidance); behaviours carried out when the person feels in a situation of imminent threat (In-situation); escape from situations when the person feels in a situation of imminent threat (Escape); compliance with the demands or wishes of the persecutors (Compliance); seeking help in reducing the threat (Help-seeking); confronting the persecutors (Aggression); and behaviours that were carried out with the belief that they would reduce threat but were judged by the interviewer to have no logical relation to the achievement of this aim (Delusional). After a safety behaviour has been elicited, the participant is asked to rate its frequency over the last month on a four-point scale (1 ¼ behaviour definitely occurred on at least one occasion, 2 ¼ occurred more than once but not frequently, e.g. not more than five or more times, 3 ¼ occurred frequently, e.g. at least five times, 4 ¼ present more or less continuously, at least every day). Scores are then summed for each of the sub-scales, and for the questionnaire as a whole. A further division is made between Avoidance, which can be considered a negative behaviour, and a combined score for the other sub-scales, which reflects active attempts to reduce the threat or positive behaviour (SBQ Positive Actions). The SBQ has been found to have good inter-rater reliability and acceptable test retest reliability. MADs (Wessely et al., 1993). The MADs is a standardised interview assessment of delusions, with good inter-rater reliability but lower test retest reliability, though this latter finding is thought to reflect true changes in the mental state of an acutely ill sample (Taylor et al., 1994). The key sub-scale for the current study was Action on Beliefs. A series of structured questions are asked for the presence of positive behaviours ( Does X make you do anything in particular?) and the presence of negative behaviours ( Has X stopped you from doing things you would normally have done? ). Each behaviour is rated on a three-point scale (did not occur, sometimes, often). Higher scores indicate greater levels of acting on delusions. One item was removed from the positive behaviour scale since it was considered close to the safety behaviour concept ( Have you tried to protect yourself in any way? ). Scale for the Assessment of Positive Symptoms (SAPS; Andreasen, 1984a). The SAPS is a 35-item, six-point (0 5) rating instrument for the assessment of the positive symptoms of psychosis. Symptoms over the last month were rated. Higher scores indicate the presence of greater levels of positive symptoms. SANS (Andreasen, 1984b). The SANS is a 24-item, six-point (0 5) rating instrument for the assessment of the negative symptoms of psychosis. Symptoms over the last month were rated. Five global scores were used (affective flattening, alogia, avolition, anhedonia, attention) and the combined sum of these scores. Higher scores indicate the presence of greater levels of negative symptoms. Psychotic Symptom Rating Scales (Haddock et al., 1999). The PSYRATS is a 17-item, five-point scale (0 4) multi-dimensional measure of delusions and hallucinations. For example, it assesses the conviction, preoccupation, and distress associated with delusions. Symptoms over the last week are rated. For the current study, the key item was the intensity of distress associated with the delusion (no distress, slight distress, moderate distress, marked distress, extreme distress). Beck Anxiety Inventory (BAI; Beck, Epstein, Brown, & Steer, 1988). The BAI is a self-report 21-item, fourpoint (0 3) scale for the assessment of anxiety. Higher scores indicate higher levels of anxiety. Beck Depression Inventory-II (BDI; Beck, Steer, & Brown, 1996). The BDI-II is a self-report 21-item, fourpoint scale (0 3) for the assessment of depression. Higher scores indicate higher depression. Birchwood, Iqbal, Chadwick, and Trower (2000) report a high correlation (r ¼ :91) between the BDI and the interview-based Calgary Depression Scale for Schizophrenia (Addington, Addington, & Maticka-Tyndale, 1993). Clinical case note data Data on age, sex, ethnicity, admissions, contact with services, and a history of violence or suicide were taken from clinical notes as part of the screening procedure for the trial. Details of current medication were taken from patients and their medical notes and recorded as chlorpromazine equivalent doses: low (0 200 mg chlorpromazine), medium ( mg) or high (400+mg).

5 D. Freeman et al. / Behaviour Research and Therapy 45 (2007) Statistical analysis Analyses were conducted using SPSS for Windows (version ) (SPSS, 2001). Significance test results are quoted as two-tailed probabilities. Pearson or Spearman coefficients are reported for correlational analyses. T- tests or ANOVA were used for the comparison of groups on dimensional measures, and 95% confidence intervals (CI) are reported. Multiple regressions were carried out using the Enter method in SPSS. Results Demographic and clinical data Basic demographic and clinical data are displayed in Table 1. The average age of the group was approximately 40, there were more males than females, the typical length of time since first onset of illness was 10 years, and the main diagnosis was schizophrenia. In this acute group there were high levels of delusions and hallucinations, but low levels of negative symptoms, especially affective flattening and inattentiveness. The presence of safety behaviours All but four of the 100 individuals (96%) reported that they had carried out at least one safety behaviour in the last month (see Table 2). The most common safety behaviour was avoidance of threatening situations. Insituation safety behaviours were also present in over half of the group. Other types of safety behaviours were apparent in one-quarter to one-third of the participants. Delusional safety behaviours, actions that could not conceivably reduce the reported threat, were very rare. These findings are highly consistent with those of Freeman et al. (2001). Safety behaviours and emotional distress Pearson correlations are reported. Higher total safety behaviours scores were associated with higher levels of anxiety, n ¼ 93, r ¼ :26, p ¼ :012, and depression, n ¼ 95, r ¼ :25, p ¼ :015. The association of emotional distress was with avoidance safety behaviours rather than other types; higher avoidance scores were significantly associated with anxiety, n ¼ 93, r ¼ :28, p ¼ :008, but higher positive safety behaviour scores were not, n ¼ 93, r ¼ :08, p ¼ :465. These associations were validated using the five-point PSYRATS item measuring the intensity of distress of the delusion. Intensity of distress of the delusion was associated with greater use of safety behaviours, n ¼ 99, Spearman r ¼ :21, p ¼ :040, though in this case there was a significant association with positive safety behaviour scores, n ¼ 99, Spearman r ¼ :34, p ¼ :001, but not avoidance safety behaviour scores, n ¼ 99, Spearman r ¼ :12, p ¼ :241. Safety behaviours and demographic and clinical variables Males and females did not significantly differ in the use of safety behaviours (see Table 3), t ¼ 1:24, df ¼ 98, p ¼ :216, mean difference ¼ 3.78, 95% CI ¼ 9.90, Level of medication was also unrelated to safety behaviour scores, F (2, 87) ¼ 2.01, p ¼ :138. However, individuals with a history of violence reported more safety behaviours than those without, t ¼ 2:84, df ¼ 95, p ¼ :006, mean difference ¼ 9.71, 95% CI ¼ 16.5, This was not accounted for by aggressive safety behaviours. Individuals with a history of violence were not more likely to use aggressive safety behaviours than individuals with no such history of violence, t ¼ :38, df ¼ 95, p ¼ :707, mean difference ¼.1, 95% CI ¼.5,.4. Individuals with a history of suicide attempts or self-harm reported more safety behaviours than individuals with no history of suicide, t ¼ 2:09, df ¼ 96, p ¼ :039, mean difference ¼ 5.87, 95% CI ¼ 11.45,.29. In order to investigate whether the history of violence and history of suicide groups had higher safety behaviour scores because they were currently emotionally distressed, their scores were examined on the BDI and BAI (see Table 4). The history of violence group, compared with those with no history of violence, did not have significantly higher levels of anxiety, t ¼ :67, df ¼ 88, p ¼ :507, mean difference ¼ 2.28, 95% CI ¼ 9.08, 4.52, or depression,

6 94 D. Freeman et al. / Behaviour Research and Therapy 45 (2007) Table 1 Demographic and clinical data Mean age (SD) (11.0) Male/Female /31 Ethnicity: 100 White 80 Black Caribbean 10 Black African 4 Black Other 1 Indian 2 Other 3 Employment: 98 Employed 6 Employed part-time 5 Student 2 Voluntary employment 5 Unemployed 79 Retired 1 Inpatient/Outpatient /45 Mean length of illness (SD) (9.1) Mean number of admissions (SD) (4.9) Diagnosis 100 Schizophrenia 86 Schizo-affective disorder 14 Medication: 93 None 3 Low (0 200 mg chlorpromazine equiv.) 31 Medium ( mg chlorpromazine equiv.) 36 High (400+mg chlorpromazine equiv.) 23 Mean SAPS Global sum (SD) (3.0) Mean Global Delusion (SD) 4.0 (.6) Mean Global Hallucination (SD) 3.0 (1.8) Mean SANS Global sum (SD) (4.5) Mean Global Affective flattening (SD).9 (1.3) Mean Global Alogia (SD) 2.6 (1.4) Mean Global Avolition (SD) 2.5 (1.4) Mean Global Anhedonia (SD) 2.5 (1.4) Mean Global Attention (SD).8 (1.1) Mean BDI (SD) (13.2) Mean BAI (SD) (13.2) N Table 2 The presence of safety behaviours in individuals with persecutory delusions (N ¼ 100) At least one safety behaviour in the last month (%) Mean score SD Avoidance Positive actions: In-situation Escape Compliance Help-seeking Aggressive Delusional Total

7 D. Freeman et al. / Behaviour Research and Therapy 45 (2007) Table 3 Safety behaviours and demographic and clinical variables Variable N Safety behaviour questionnaire Mean score SD Gender Male Female Medication Low Medium High Violence History of violence No history of violence Suicide attempts History of suicide No history of suicide Table 4 Depression and anxiety scores by history of violence or suicide Anxiety Depression N Mean SD N Mean SD Violence History of violence No history of violence Suicide attempts History of suicide No history of suicide t ¼ :58, df ¼ 90, p ¼ :560, mean difference ¼ 1.95, 95% CI ¼ 8.58, History of violence data and BAI scores were entered into a regression analysis with total safety behaviours scores as the dependent variable. History of violence, t ¼ 2:60, p ¼ :011, and anxiety, t ¼ 2:50, p ¼ :014, remained significant predictors. Similarly, the history of suicide group did not have significantly higher levels of anxiety compared with those with no such history, t ¼ 1:04, df ¼ 89, p ¼ :302, mean difference ¼ 2.92, 95% CI ¼ 8.51, 2.67, nor were there differences in depression, t ¼ 1:69, df ¼ 91, p ¼ :094, mean difference ¼ 4.58, 95% CI ¼ 9.95,.79. History of suicide and BAI scores were then entered into a regression analysis with total safety behaviours scores as the dependent variable. Again, anxiety was a significant predictor of safety behaviour scores, t ¼ 2:38, p ¼ :020, but history of suicide did not remain a significant predictor in this analysis, t ¼ 1:84, p ¼ :070. Safety behaviours and acting on delusions There is a significant association between acting on delusions and safety behaviours of medium effect size (see Table 5). Higher levels of safety behaviours are associated with higher levels of acting on delusions as measured by the MADs. Further, avoidance safety behaviours are associated with MADs negative actions and positive safety behaviours are particularly associated with MADs positive actions. A history of violence was not associated with MADs acting on delusions, t ¼ 1:14, df ¼ 95, p ¼ :256, mean difference ¼ 2.0, 95% CI ¼ 5.6, 1.5. Nor was a history of suicide associated with MADs acting on delusions, t ¼ :645, df ¼ 96, p ¼ :520, mean difference ¼.9; 95% CI ¼ 1.9, 3.8.

8 96 D. Freeman et al. / Behaviour Research and Therapy 45 (2007) Table 5 Pearson correlations between safety behaviours and acting on delusions (N ¼ 100) MADs acting on delusions SBQ-Total SBQ-Avoidance SBQ-Positive Actions Total behaviours.34***.25*.36*** po:001 p ¼ :013 po:001 Positive behaviours.17 o.01.41*** p ¼ :098 p ¼ :983 po:001 Negative behaviours.40***.39***.21* po:001 po:001 p ¼ :038 *po:05, **po:01, ***po:001. Table 6 Spearman correlations between safety behaviours, emotion, and negative symptoms Negative symptoms SBQ-Total (n ¼ 100) SBQ-Avoidance (n ¼ 100) SBQ-Positive Actions (n ¼ 100) BAI (n ¼ 93) BDI (n ¼ 95) Affective flattening p ¼ :419 p ¼ :514 p ¼ :836 p ¼ :406 p ¼ :397 Alogia.27**.23*.29** p ¼ :007 p ¼ :022 p ¼ :003 p ¼ :162 p ¼ :872 Avolition-apathy p ¼ :480 p ¼ :461 p ¼ :842 p ¼ :932 p ¼ :636 Anhedonia-asociality *** p ¼ :276 p ¼ :373 p ¼ :539 p ¼ :208 po:001 Attention * p ¼ :262 p ¼ :157 p ¼ :843 p ¼ :090 p ¼ :015 Sum of global scores * p ¼ :341 p ¼ :549 p ¼ :354 p ¼ :254 p ¼ :025 *po:05, **po:01, ***po:001. Safety behaviours and the negative symptoms of psychosis It was predicted that safety behaviours would be positively associated with negative symptoms. There was no evidence for this prediction (see Table 6). Indeed, the only significant finding was in the opposite direction: more safety behaviours were associated with less alogia. Consistent with previous research, however, higher levels of depressive symptoms as measured by the BDI were associated with higher levels of negative symptoms. Discussion The clear finding of the study is that safety behaviours are remarkably common in individuals with persecutory delusions. Ninety-six percent of patients reported that they had used a safety behaviour in the last month. Avoidance of threat was the most frequently used safety strategy as is seen in clinical practice. However, even when patients believed they were within reach of their persecutor they often tried to minimise or prevent the threat from occurring. The use of safety behaviours was modestly associated with higher levels of anxiety and depression. Hence, we think it likely that safety seeking is at least partly motivated by fear and distress. There were significant associations between safety behaviours and the MADs acting on delusion measure. Safety behaviours might explain some components of individuals acting on delusions. The advantage of the safety behaviour concept is that it is a theoretically informed way of understanding responses to delusional beliefs. Acting on delusions is formulated at the level of acts, whereas safety behaviours are formulated in

9 terms of intentions and purposes. Moreover, safety behaviours may help the understanding of delusion maintenance. Such behaviours are likely to prevent the processing of evidence against the persecutory belief, and hence lead to delusion persistence. When making sense of the maintenance of persecutory experience, it is likely to be helpful for clinicians to assess for the presence of safety behaviours. It should also be kept in mind that patients desire for safety may interfere with their engaging in other activities, often in subtler ways than simple avoidance behaviour. We did not examine whether serious previous instances of violence or suicide attempts were safety behaviours. However, individuals with such histories, particularly of violence, reported greater current use of safety behaviours. Perhaps surprisingly, individuals with a history of violence did not report a significantly greater level of use of aggressive safety behaviours in particular, although this may simply reflect the low level of aggressive safety behaviour use in the sample over the last month. Overall, individuals who have been violent or self-harmed have clearly shown a propensity to act on their ideas, and this may reflect failures to inhibit extreme responses, to think of alternative courses of action, and to cope adaptively with strong emotion (e.g. Nestor, 2002; Williams & Pollock, 2000). These processes may make it more likely that such individuals also seek safety from delusional ideas. However, they were not more likely to act on their delusions in nonsafety orientated ways, as measured by the MADs, perhaps indicating the importance of the affective reaction to the threat within the delusions in understanding their responses. The final issue investigated was the hypothesis that safety behaviour is a potential route by which positive symptoms can lead to negative symptoms of psychosis. There was no evidence for such a link. Indeed there was evidence in the opposite direction. Individuals who had higher safety behaviour use had significantly less alogia. Individuals reliant on safety behaviours might be more aroused and activated, and hence exhibit fewer negative symptoms than individuals who do little in response to their delusions. However, there were fairly low levels of negative symptoms in the group. This may have been because they were selected for the clinical trial on the basis of having had a relapse in positive symptoms. Another potential weakness of the study in relation to the study of negative symptoms is that the same raters assessed both safety behaviours and negative symptoms. It cannot be ruled out that the assessors might have under-reported the presence of negative symptoms because of their awareness of the idea of safety behaviours. The reliance on the same rater for all of a patient s assessments may also cause problems for the independence of ratings for the safety behaviour and acting on delusions measures (although the assessors were unaware of the hypothesis to be tested). There are further limitations of the study. All behaviours were self-reported and no corroborative evidence from others was sought. The study was cross-sectional and retrospective, meaning that the causal direction of the associations could not be determined. The group had a reasonable degree of chronicity of illness and it is therefore not apparent whether such a high level of safety behaviour use would be present in younger samples. However, the group was a relapsing sample and therefore clinically important to investigate. It may also be a limitation of the current work that safety behaviours were not considered in the light of the relationship of the person to the persecutor. It has been demonstrated that the relationship of voice hearers to their voices determines subsequent behaviours and reactions (e.g. Birchwood, Iqbal et al., 2000; Birchwood, Meaden, Trower, Gilbert, & Plaistow, 2000; Trower et al., 2004) and that paranoia is associated with submissive behaviours (e.g. Freeman et al., 2005; Gilbert, Boxall, Cheung, & Irons, 2005). Safety behaviours in paranoia may, in part, be understood in terms of the relationship of the individual to the perpetrator of the threat. Overall, we think the study provides a clear demonstration that individuals with persecutory delusions often act on their delusions in a reported attempt to achieve safety, but clearly further research is needed to identify the determinants of safety behaviours other than distress. A further key question remains to be tested: do safety behaviours contribute to the chronicity of patients delusions? Acknowledgements D. Freeman et al. / Behaviour Research and Therapy 45 (2007) The work was supported by a programme grant from the Wellcome Trust (No ). We wish to thank the patients taking part in the trial and the participating clinical teams in the South London and Maudsley NHS Trust, Norfolk Mental Health NHS Trust, Camden & Islington Mental Health and Social Care NHS Trust, and North East London Mental Health NHS Trust.

10 98 D. Freeman et al. / Behaviour Research and Therapy 45 (2007) References Addington, D., Addington, J., & Maticka-Tyndale, E. (1993). Assessing depression in schizophrenia: The calgary depression scale. British Journal of Psychiatry, 163(s22), Andreasen, N. C. (1984a). The scale for the assessment of positive symptoms (SAPS). Iowa City, IA: The University of Iowa. Andreasen, N. C. (1984b). The scale for the assessment of negative symptoms (SANS). Iowa City, IA: The University of Iowa. Appelbaum, P. S., Robbins, P. C., & Monahan, J. (2000). Violence and delusions: Data from the MacArthur violence risk assessment study. American Journal of Psychiatry, 157, Applebaum, P. S., Robbins, P. C., & Roth, L. H. (1999). Dimensional approach to delusions: Comparison across types and diagnoses. American Journal of Psychiatry, 156, Beck, A. T., Epstein, N., Brown, G., & Steer, R. (1988). An inventory for measuring clinical anxiety: Psychometric properties. Journal of Consulting and Clinical Psychology, 56, Beck, A. T., Steer, R. A., & Brown, G. K. (1996). BDI-II manual. San Antonio: The Psychological Corporation. Birchwood, M., Iqbal, Z., Chadwick, P., & Trower, P. (2000). Cognitive approach to depression and suicidal thinking in psychosis: I. Ontogeny of post-psychotic depression. British Journal of Psychiatry, 177, Birchwood, M., Meaden, A., Trower, P., Gilbert, P., & Plaistow, J. (2000). The power and omnipotence of voices: Subordination and entrapment by voices and significant others. Psychological Medicine, 30, Buchanan, A., Reed, A., Wessely, S., Garety, P., Taylor, P., Grubin, D., et al. (1993). Acting on delusions. II: The phenomenological correlates of acting on delusions. British Journal of Psychiatry, 163, Carpenter, W. T., Heinrichs, D. W., & Alphs, L. D. (1985). Treatment of negative symptoms. Schizophrenia Bulletin, 11, Clark, D. M. (1999). Anxiety disorders: Why they persist and how to treat them. Behaviour Research and Therapy, 37, S5 S27. Freeman, D., & Garety, P. A. (2004). Paranoia: The psychology of persecutory delusions. Hove: Psychology Press. Freeman, D., Garety, P. A., Bebbington, P. E., Smith, B., Rollinson, R., Fowler, D., et al. (2005). Psychological investigation of the structure of paranoia in a non-clinical population. British Journal of Psychiatry, 186, Freeman, D., Garety, P. A., & Kuipers, E. (2001). Persecutory delusions: Developing the understanding of belief maintenance and emotional distress. Psychological Medicine, 31, Freeman, D., Garety, P. A., Kuipers, E., Fowler, D., & Bebbington, P. E. (2002). A cognitive model of persecutory delusions. British Journal of Clinical Psychology, 41, Gilbert, P., Boxall, M., Cheung, M., & Irons, C. (2005). The relation of paranoid ideation and social anxiety in a mixed clinical population. Clinical Psychology and Psychotherapy, 12, Haddock, G., McCarron, J., Tarrier, N., & Faragher, E. B. (1999). Scales to measure dimensions of hallucinations and delusions: the psychotic symptom rating scales (PSYRATS). Psychological Medicine, 29, Kelley, M. E., van Kammen, D. P., & Allen, D. N. (1999). Empirical validation of primary negative symptoms. American Journal of Psychiatry, 156, Malla, A. K., Takhar, J. J., Norman, R. M. G., Manchanda, R., Cortese, L., Haricharan, R., et al. (2002). Negative symptoms in first episode non-affective psychosis. Acta Psychiatrica Scandinavica, 105, Morrison, A. P. (1998). Cognitive behaviour therapy for psychotic symptoms in schizophrenia. In N. Tarrier, A. Wells, & G. Haddock (Eds.), Treating complex cases (pp ). Chichester: Wiley. Nestor, P. G. (2002). Mental disorder and violence. American Journal of Psychiatry, 159, O Leary, D. S., Flaum, M., Kesler, M. L., Flashman, L. A., Arndt, S., & Andreasen, N. C. (2000). Cognitive correlates of the negative, disorganized, and psychotic symptom dimensions of schizophrenia. Journal of Neuropsychiatry and Clinical Neurosciences, 12, Salkovskis, P. M. (1991). The importance of behaviour in the maintenance of anxiety and panic: A cognitive account. Behavioural Psychotherapy, 19, Salkovskis, P. M., Clark, D. M., & Gelder, M. G. (1996). Cognition behaviour links in the persistence of panic. Behaviour Research & Therapy, 34, Salkovskis, P. M., Clark, D. M., Hackmann, A., Wells, A., & Gelder, M. G. (1999). An experimental investigation of the role of safetyseeking behaviours in the maintenance of panic disorder with agoraphobia. Behaviour Research & Therapy, 37, Sloan, T., & Telch, M. J. (2002). The effects of safety-seeking behaviour and guided threat reappraisal on fear reduction during exposure: An experimental investigation. Behaviour Research and Therapy, 40, SPSS. (2001). SPSS Base user s guide. Chicago, IL.: SPSS Inc. Taylor, P. J., Garety, P., Buchanan, A., Reed, A., Wessely, S., Ray, K., et al. (1994). Delusions and violence. In J. Monahan, & H. Steadman (Eds.), Violence and mental disorder (pp ). The John D. and Catherine T. MacArthur Foundation series on mental health and development. Trower, P., Birchwood, M., Meaden, A., Byrne, S., Nelson, A., & Ross, K. (2004). Cognitive therapy for command hallucinations: Randomised controlled trial. British Journal of Psychiatry, 184, Verdoux, H., Liraud, F., Gonzales, B., Assens, F., Abalan, F., & van Os, J. (2001). Predictors and outcome characteristics associated with suicidal behaviour in early psychosis. Acta Psychiatrica Scandinavica, 103, Walsh, E., Gilvarry, C., Samele, C., Harvey, K., Manley, C., Tattan, T. & for the UK700 Group, et al. (2004). Predicting violence in schizophrenia: A prospective study. Schizophrenia Research, 67, Wells, A., Clark, D. M., Salkovskis, P., Ludgate, J., Hackmann, A., & Gelder, M. (1995). Social phobia: The role of in-situation safety behaviours in maintaining anxiety and negative beliefs. Behaviour Therapy, 26,

11 D. Freeman et al. / Behaviour Research and Therapy 45 (2007) Wessely, S., Buchanan, A., Reed, A., Cutting, J., Everitt, B., Garety, P., et al. (1993). Acting on delusions (1): Prevalence. British Journal of Psychiatry, 163, Williams, J. M. G., & Pollock, L. R. (2000). The psychology of suicidal behaviour. In K. Hawton, & K. van Heeringen (Eds.), The international handbook of suicide and attempted suicide (pp ). Chichester, UK: Wiley. Wing, J. K. (1989). The concept of negative symptoms. British Journal of Psychiatry, 155(Suppl. 7), Wing, J., & Brown, G. (1970). Institutionalism and schizophrenia. Cambridge: Cambridge University Press.