Cognitive insight and psychotic disorder: The impact of active delusions

Schizophrenia Research 90 (2007) 325 – 333 www.elsevier.com/locate/schres

Cognitive insight and psychotic disorder: The impact of active delusions Debbie M. Warman a,⁎, Paul H. Lysaker b , Joel M. Martin c b

a University of Indianapolis, School of Psychological Sciences, 1400 East Hanna Avenue, Indianapolis, IN 46227, USA Roudebush VA Medical Center and Indiana School of Medicine, Department of Psychiatry, 1481 West 10th Street, Indianapolis, IN 46202, USA c Butler University, 4600 Sunset Avenue, Indianapolis, IN 46208, USA

Received 30 May 2006; received in revised form 13 September 2006; accepted 18 September 2006 Available online 7 November 2006

Abstract While several studies have determined the Beck Cognitive Insight Scale (BCIS; [Beck, A.T., Baruch, E., Balter, J.M., Steer, R.A., Warman, D.M., 2004. A new instrument for measuring insight: The Beck Cognitive Insight Scale. Schizophr. Res. 68, 319–329] is a useful measure of cognitive insight, a number of questions have remained unanswered. While individuals with psychotic disorders have been shown to have impaired cognitive insight compared to a psychiatric comparison group, it has remained unclear how the cognitive insight of individuals with psychotic disorders compares to healthy individuals. Further, as previous studies have classified participants based on diagnostic classification, it has remained unknown if individuals with delusions and individuals with psychotic disorders without active delusions score differently on this measure. To examine these questions, we assessed the cognitive insight of healthy individuals and individuals with psychotic disorders, both with and without active delusions. Results indicated that individuals with psychotic disorders had impaired cognitive insight relative to healthy controls ( p = .005), though individuals with active delusions and individuals with psychotic disorders without delusions had impairments in different domains. Individuals with delusions were overly confident in their own judgment relative to healthy controls and those without delusions ( p = .011), though their self-reflectiveness was the same as normal controls. Individuals without delusions reported low self-reflectiveness relative to healthy controls and individuals with delusions ( p = .004), though they were not overconfident in their judgment. These results are discussed in terms of existing research on cognitive insight, decision making, and psychosis. © 2006 Elsevier B.V. All rights reserved. Keywords: Cognitive insight; Psychosis; Schizophrenia; Delusions

1. Introduction It is widely recognized that many persons with schizophrenia have poor insight or significant difficulties

⁎ Corresponding author. Tel.: +1 317 788 2102; fax: +1 317 788 2120. E-mail addresses: [email protected] (D.M. Warman), [email protected] (P.H. Lysaker), [email protected] (J.M. Martin). 0920-9964/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.schres.2006.09.011

recognizing their deficits and need for treatment (Amador et al., 1993). More recently, Beck et al. (2004) have noted that schizophrenia may involve a compromise in an additional kind of insight — cognitive insight, or the ability to self-reflect, acknowledge the possibility of being mistaken, be open to feedback, and to refrain from overconfidence. This form of insight has been thought to have great clinical significance as it taps directly into how individuals with delusions think,

326

D.M. Warman et al. / Schizophrenia Research 90 (2007) 325–333

making it particularly relevant for facilitating change in psychotherapy (Beck and Warman, 2004). To assess cognitive insight, the Beck Cognitive Insight Scale (BCIS; Beck et al., 2004) was recently developed. The BCIS, a 15-item self-report instrument, is comprised of two factors, the first labeled SelfReflectiveness, which includes items related to the willingness to acknowledge fallibility, the recognition of having jumped to conclusions at times, and the possibility of having misinterpreted unusual experiences. The second factor, labeled Self-Certainty, includes items related to overconfidence. It is speculated that overconfidence may serve to impair self-reflectiveness, therefore the Composite Index, the measure of the person's overall cognitive insight, is determined by subtracting the individual's Self-Certainty score from his/her SelfReflectiveness score. To date, the BCIS has been investigated in three studies, all which examined different populations (Beck et al., 2004; Pedrelli et al., 2004; Warman et al., 2004). The initial study (Beck et al., 2004), which determined the factor structure of the BCIS and its psychometric properties, investigated a group of inpatients who had psychotic disorders of various diagnoses and also patients with major depressive disorder without psychosis. Pedrelli et al. (2004) examined the BCIS with a sample of middle to older adult outpatients, all of whom had a diagnosis of schizophrenia and schizoaffective disorder and who had, on average, mild to moderate psychotic symptoms, and determined that the factor structure of the measure was the same as it was in the study by Beck et al. In addition, Warman et al. (2004) found that the basic factor structure of the BCIS was the same in a normal population of university students (none of whom had a psychotic disorder) as it was in the original Beck et al. study of inpatients with psychiatric diagnoses. In sum, the studies to date have determined that the BCIS is a useful measure for inpatients with various diagnoses, outpatients with psychotic disorders, and also healthy university students. While these initial investigations indicate the BCIS is a valuable measure of cognitive insight, a number of points remain unclear. First, how cognitive insight relates to active delusions, as opposed to psychotic disorder diagnosis has not yet been determined. Beck et al. (2004) found that individuals with psychotic disorders were significantly less self-reflective and more overconfident than individuals without psychotic disorders; as group classification was based on diagnosis, not active delusional status at the time of testing, the relationship to active delusions cannot be determined. While Pedrelli et al. (2004) did not directly examine individuals with active

delusions compared to individuals without active delusions, they noted that severity of positive symptoms was not related to self-reflectiveness or to overall cognitive insight. Severity of positive symptoms was, however, positively (though weakly, r = .24) correlated with overconfidence. These results suggest that active psychosis may impair certainty judgments more than it does selfreflectiveness, though a direct assessment of individuals with active delusions versus individuals with psychotic disorders without active delusions has not yet been done. A further remaining question is how the cognitive insight of individuals with psychotic disorders and/or active delusions compares to a normal population. While it has been determined that individuals with psychotic disorders have impaired cognitive insight relative to those with major depressive disorder without psychosis, the cognitive insight of individuals with psychotic disorders has not yet been compared to healthy individuals. Specifically, no studies have included both a population of individuals with psychotic disorders and a population of healthy individuals with no psychotic disorder diagnosis. Despite the lack of such direct comparisons, by examining the results of the three studies that have examined the BCIS, it is possible to develop some tentative conclusions (see Table 1). Overall, it appears that individuals with psychotic disorders and healthy individuals score approximately the same on the Self-Reflectiveness subscale and that both groups are less self-reflective than individuals with severe depression. In terms of overconfidence, it appears that healthy individuals score midway between individuals with severe depression without psychosis and those with psychotic disorders. It appears that the relationship between cognitive insight and psychosis may not be a simple one. Further highlighting the need to determine how the cognitive insight of individuals with delusions compares to healthy controls, a recent study determined that individuals who were highly delusion prone (i.e., members of the general, non-clinical population who had no psychotic disorder but who scored high on a measure assessing presence of unusual beliefs) were both more self-reflective and more overconfident on the BCIS than those who were low in delusion proneness (Warman and Martin, 2006). While this seems counterintuitive, since no study has compared individuals with active delusions to healthy controls, it cannot be ruled out that the results that were found for delusion prone individuals actually mirror those that would be revealed between delusional and healthy populations. It is possible that individuals with delusions recognize that their thinking may have been faulty at times, but that overconfidence in current judgment serves to maintain delusional beliefs.

D.M. Warman et al. / Schizophrenia Research 90 (2007) 325–333

327

Table 1 Means and standard deviations of cognitive insight scores from previous studies Beck et al. (2004)

Self-Reflectiveness Self-Certainty Composite Index

Psychotic disorders

Depression no psychosis

13.19 (5.14) 7.45 (3.78) 5.74 (5.74)

15.29 (4.07) 6.29 (2.92) 9.00 (4.51)

Pedrelli et al. (2004)

Warman et al. (2004)

12.46 (4.58) 7.24 (3.06) 5.22 (5.78)

11.69 (2.92) 6.77 (2.68) 4.90 (4.08)

Note: Beck et al.'s participants were from an inpatient setting and all had a diagnosis of schizophrenia, schizoaffective disorder, major depressive disorder with psychotic features (all of whom were classified as “Psychotic disorders” above), or had a diagnosis of major depressive disorder without psychosis; participants of Pedrelli et al. all had a diagnosis of schizophrenia or schizoaffective disorder and were in an outpatient setting; participants of Warman et al. were all university students and were ‘healthy’ in that none had a current or historical diagnosis of psychotic disorder.

The present study aimed to answer a number of questions that have been unanswered in research to date. First, while the BCIS is a useful measure to understand cognitive insight, the role of active delusions has yet to be investigated. In addition, the cognitive insight of individuals with psychotic disorders, both with and without active delusions has not yet been compared to healthy controls. The present study assessed the cognitive insight of individuals with psychotic disorders, both with and without active delusions, and also assessed the cognitive insight of a group of healthy individuals enrolled in a university. Pedrelli et al. (2004) found no relationship between severity of positive symptoms and self-reflectiveness in their study of outpatients with psychotic disorders, and Warman et al. (2004) reported means for healthy individuals on the self-reflectiveness scale that were similar to those reported by studies with individuals who had psychotic disorders (Beck et al., 2004; Pedrelli et al., 2004); Warman and Martin (2006), however, found that individuals who were delusion prone had higher selfreflectiveness than individuals who were not delusion prone. Given these findings, it was expected that all participants would score similarly on the Self-Reflectiveness subscale, with the possibility that individuals with delusions would score somewhat higher than healthy controls. As Pedrelli et al. found that the severity of symptoms was related to overconfidence, a finding that has been extended to individuals who are delusion prone (Warman and Martin, 2006), it was expected that individuals with active delusions would be more confident in their judgment (i.e., they would score higher on the SelfCertainty subscale) than those with psychotic disorders without active delusions and also healthy controls. 2. Method 2.1. Participants In order to participate in the present study, individuals needed to be at least 18 years of age. Participants

were recruited from two locations: 1) a VA Medical Center 2) a small liberal arts college in the American Midwest. At the VA, individuals with SCID confirmed diagnoses of Schizophrenia or Schizoaffective Disorder who were enrolled in a larger study investigating anxiety were invited to participate. All the participants recruited from the VA were outpatients at the time of testing. A portion of the participants were also involved in a decision making study, the results of which will be reported elsewhere. At the college, undergraduate students who elected to participate in the present study for course credit were included as long as they met the inclusion criteria of being at least 18 years old and had the ability to read and understand English. The university students were enrolled in a larger study investigating decision making. No standardized assessment was administered to the university students to determine presence or absence of a psychotic disorder; rather, participants were asked as part of the protocol if they had ever been diagnosed with a psychological disorder. No participant from the university sample reported a history of psychotic disorder diagnosis. All of the participants with psychotic disorders were from the VA sample. Each participant provided informed consent and voluntarily agreed to participate. Participant characteristics are summarized in Table 2. 2.2. Materials 2.2.1. Assessment of cognitive insight The Beck Cognitive Insight Scale (BCIS; Beck et al., 2004) is a 15-item self-report measure that assesses how individuals evaluate their own judgment. The scale has two factors, Self-Reflectiveness and Self-Certainty. To compute a Composite Index, the Self-Certainty score is subtracted from the Self-Reflectiveness score. The original validation study of the BCIS reported a coefficient α for the Self-Reflectiveness scale of 0.68 and for Self-Certainty 0.60 (Beck et al., 2004). The BCIS has also been examined with a normal population;

328

D.M. Warman et al. / Schizophrenia Research 90 (2007) 325–333

Table 2 Participant characteristics Psychotic disorders Active delusions

Gender Male Female Race Caucasian African American Other Age

No active delusions

Healthy controls

N

%⁎

N

%⁎

N

%⁎

33 4

86.8 10.5

11 1

84.6 7.7

15 45

25.0 75.0

18 12

47.4 31.6

5 7

38.5 53.9

49 4

81.67 6.67

3 7.9 Mean (S.D.) 48.08 (6.03)

1 7.7 Mean (S.D.) 50.46 (7.87)

7 11.67 Mean (S.D.) 21.35 (5.53)

Note: 1 participant in the psychotic disorder without active delusions group did not report gender; 5 participants in the active delusions group did not report their race; 1 participant in the active delusions group did not report age. ⁎ Percentages are reported for number of participants within each group.

the factor loadings and internal consistencies of the SelfReflectiveness and Self-Certainty scales were similar in the normal population as with the original inpatient sample (Warman et al., 2004). 2.2.2. Measure of depression The relationship of depression to cognitive insight has not been investigated extensively. Two studies found no relationship between the BCIS and depression scores for individuals with psychotic disorders (Beck et al., 2004; Pedrelli et al., 2004), though Beck and colleagues found moderate positive correlations between depression and the Self-Reflectiveness and SelfCertainty scales for individuals diagnosed with depression without psychosis (Beck et al., 2004). Granholm et al. (2005), in an investigation of psychotherapy for individuals with chronic schizophrenia or schizoaffective disorder, found a relationship between increased cognitive insight and increased depression midway through treatment. Two studies that investigated the BCIS with a normal population did not include a measure of depression (Warman et al., 2004; Warman and Martin, 2006), making it difficult to draw general conclusions about the relationship between the BCIS and depression. As the present study's goal was to examine the unique relationship of cognitive insight to group classification (psychotic disorder or no psychotic disorder; active delusions vs psychotic disorder with no active delusions vs healthy control), a measure of depression was administered so that effects, if they

emerged, could be partialed out. The Beck Depression Inventory-II (BDI-II; Beck et al., 1996) is a 21-item selfreport instrument developed to measure the severity of depression in adults and adolescents. Each item consists of four statements reflecting increasing levels of severity for a particular symptom of depression. This measure was modified for the current study in that two items, “Pessimism” and “Suicidal Thoughts and Dying” were omitted due to their sensitive nature for a non-clinical sample. The BDI has demonstrated high reliability (α = .91) and convergent validity (r = .93; Dozois et al., 1998). 2.2.3. Measure of psychotic symptomology The Positive and Negative Syndrome Scale (PANSS; Kay et al., 1987) is a 30-item rating scale completed by clinically trained research staff at the conclusion of chart review and a semi-structured interview. It is one of the most widely used semi-structured interviews for assessing the wide range of psychopathology in schizophrenia. For the present study, if a participant scored a 3 or above on the Delusions item, he or she was classified as having delusions. Individuals who scored a 2 or below on this item were classified as not having active delusions. 2.3. Procedure The present study had full approval from the Institutional Review Boards of their respective universities/institutions. Participants were tested individually. Research assistants with extensive training in the procedures of the study administered the assessments. To test the questions of the current study, participants completed the BCIS; the BDI-II was administered to determine if this was related to cognitive insight and, thus, may need to be considered as a covariate. The individuals with psychotic disorders were administered the PANSS by one of the members of the research team with extensive training in administration of this measure. 3. Results Results are presented in two major sections. In the first section, individuals with schizophrenia or schizoaffective disorder are compared to the individuals in the nonclinical sample. For the first section of the results, these groups are labeled psychotic disorder and healthy controls, respectively. The analyses in the first section were done to be consistent with the original study of Beck et al. (2004) which distinguished participants based on diagnostic classification, not by delusional status at the time of testing. In the second

D.M. Warman et al. / Schizophrenia Research 90 (2007) 325–333

section, individuals with schizophrenia or schizoaffective disorder are distinguished based on their delusional status at time of testing (i.e., their PANSS score on the delusions item) and are compared to the nonclinical sample. Thus, for the second section three groups are compared: schizophrenia/schizoaffective disorder with delusions, schizophrenia/schizoaffective disorder without delusions, and nonclinical sample. For section two of the results section, these groups are labeled psychotic disorder with current delusions (CD), psychotic disorder with no current delusions (ND), and healthy controls (HC), respectively. 3.1. Psychotic disorder and healthy control comparisons 3.1.1. Preliminary analyses To determine whether participants with psychotic disorders differed from healthy controls in terms of demographic variables (age, gender, and race) and depression (as measured by the BDI-II) a series of tests were conducted. No differences were found between individuals with psychotic disorder and healthy controls in terms of depression (t (85.39) = − 1.83, p = .071). The two groups did differ, however, in terms of age, t (96.22) = − 23.37, p b .001, gender, χ2 (1, N = 109) = 45.61, p b .001, and race χ2 (2, N = 106) = 18.43, p b .001. Those with psychotic disorders were significantly older (M = 48.7, S.D. = 6.56) than healthy controls (M = 21.35, S.D. = 5.53). Further, females were far more represented in the healthy control sample than in the sample with psychotic disorders (75% versus 10% of the samples, respectively). Last, the psychotic disorder sample had significantly more minorities than the healthy control sample. Specifically, while nearly 82% of the healthy sample was Caucasian, only 39% of the sample with psychotic disorders was Caucasian. To determine whether demographic variables and depression were related to the dependent variables, a series of analyses were conducted. Depression was related to the Self-Reflectiveness factor (r= .194, p = .042) and to the Composite Index (r= .198, p = .038), but was not related to the Self-Certainty factor (r= −.073, p= .449). Neither age nor race was related to scores on the BCIS Self-Reflectiveness factor (r= −.093, p = .335; F (2, 103)= 1.12, p = .330, respectively), the SelfCertainty factor (r = .163, p=.088; F(2, 103) = 2.013, p = .139, respectively) or the Composite Index (r= −.170, p = .076; F(2, 103)= 1.318, p =.272, respectively). Gender was related to scores on the Self-Reflectiveness factor, (t (99.22) =−2.323, p = .022), the Self-Certainty factor, (t (107) = 2.537, p = .013), and the Composite Index, (t (107) =−3.363, p= .001). It is clear that the strong relationship between gender and cognitive insight is an

329

artifact of females being almost exclusively in the healthy control group, making gender redundant with psychotic disorder versus healthy control classification. Because gender is redundant with psychotic disorder, it will not be considered a mitigating factor in further analyses and, like the other demographic variables that did not relate to the dependent variables, will not be discussed further. Thus, for the following analyses, when Self-Reflectiveness and the Composite Index are examined only BDI score will be entered as a covariate. 3.1.2. Self-Reflectiveness A one-way ANOVA was conducted with psychotic disorder versus healthy control group classification as the independent variable, and Self-Reflectiveness score on the BCIS as the dependent variable. As preliminary analyses indicated that depression was related to selfreflectiveness, BDI score was entered as a covariate. No significant effect emerged, F (1, 107) = 2.47, p = .119, η2 = .023. Those in the psychotic disorder group (M = 12.69, S.E. = .564) and those in the healthy control group (M = 13.90, S.E. = .514) had similar self-reflectiveness (see Fig. 1). 3.1.3. Self-Certainty A one-way ANOVA was conducted with psychotic disorder versus healthy control group classification as the independent variable and Self-Certainty score on the BCIS as the dependent variable. A significant effect

Fig. 1. Scores on the Self-Reflectiveness, Self-Certainty, and Composite Index of the Beck Cognitive Insight Scale as a function of psychotic disorder versus healthy control.

330

D.M. Warman et al. / Schizophrenia Research 90 (2007) 325–333

emerged, F (1, 109) = 7.03, p = .009, η2 = .061. Those with psychotic disorders were significantly more certain in their own judgment (M = 8.06, S.D. = 3.34) than were healthy controls (M = 6.62, S.D. = 2.37) (see Fig. 1). 3.1.4. Composite Index A one-way ANOVA was conducted with psychotic disorder versus healthy control group classification as the independent variable, and participants' Composite Index score on the BCIS as the dependent variable. As preliminary analyses indicated that depression was related to Composite Index score, BDI score was entered as a covariate. A significant effect emerged, F (1, 107) = 8.35, p = .005, η2 = .072. Those with psychotic disorders had lower overall cognitive insight (M = 4.64, S.E. = .684) than did healthy controls (M = 7.34, S.E. = .624) (see Fig. 1). 3.2. Comparisons between individuals with psychotic disorder with current delusions (CD), those with psychotic disorder with no current delusions (ND) and healthy controls (HC) 3.2.1. Preliminary analyses To determine whether those with psychotic disorder with current delusions, those with psychotic disorder with no current delusions, and healthy controls differed in terms of demographic variables (age, gender, and race) and depression (as assessed by the BDI-II), a series of analyses were conducted using those variables as dependent variables and delusion status classification (CD, ND, or HC) as the independent variable. A significant effect emerged for depression, F (2, 107) = 7.14, p = .001. Those in the CD group (M = 17.92, S.D. = 12.40) endorsed significantly more depression than did those in the ND (M = 7.69, S.D. =7.05) or HC groups (M = 11.55, S.D. = 12.40); the ND and HC groups did not differ from each other. Age was significantly related to delusion status classification, F (2, 107) = 283.83, p b .001. Follow up tests revealed that those in the CD and ND groups (M = 50.46, S.D. = 7.87; M = 48.08, S.D. = 6.03, respectively) were significantly older than those in the HC group (M = 21.35, S.D. = 5.53). In addition, gender (χ2(2, N = 109) = 45.61, p b .001) and race (χ2 (4, N = 106) = 20.15, p b .001) were significantly related to delusion status classification. A closer examination of these effects suggests that, as discussed previously, the differential rate of gender and race is redundant with the disordered versus healthy distinction; no differences emerged between the CD and ND groups in terms of gender (χ2 (1, N = 49) = .06, p = .81) or race (χ2 (2, N = 46) = 1.19, p = .55). The relationships

of demographic variables and depression to the dependent variables has already been reported; examining the results by three groups does not change the relationships. Thus, for the following analyses, BDI score will be entered as a covariate when Self-Reflectiveness and the Composite Index are examined. 3.2.2. Self-Reflectiveness A one-way ANOVA was conducted with delusion status (CD, ND, or HC) as the independent variable, and Self-Reflectiveness score on the BCIS as the dependent variable. As preliminary analyses indicated that depression was related to self-reflectiveness, BDI score was entered as a covariate. A significant effect emerged, F (2, 106) = 3.40, p = .021, η2 = .070. Follow up tests (based on estimated marginal means) indicated that those in the ND group (M = 10.49, S.E. = 1.10) demonstrated significantly less self-reflectiveness than did those in the CD (M = 13.53, S.E. = .662) and the HC groups (M = 13.85, S.E. = .504), who did not differ from one another (see Fig. 2). 3.2.3. Self-Certainty A one way ANOVA was conducted with delusion status (CD, ND, or HC) as the independent variable and Self-Certainty score on the BCIS as the dependent variable. A significant effect emerged, F (2, 108) = 5.93, p = .004, η2 = .099). Follow up tests indicated that those in the CD group (M = 8.55, S.D. = 2.92) were significantly more confident in their judgment than were those

Fig. 2. Scores on the Self-Reflectiveness, Self-Certainty, and Composite Index of the Beck Cognitive Insight Scale as a function of delusion status versus healthy control.

D.M. Warman et al. / Schizophrenia Research 90 (2007) 325–333

in the ND (M = 6.62, S.D. = 4.15) and the HC groups (M = 6.62, S.D. = 2.37), who did not differ from one another (see Fig. 2). 3.2.4. Composite Index A one-way ANOVA was conducted with delusion status (CD, ND, or HC) as the independent variable, and Composite Index score on the BCIS as the dependent variable. As preliminary analyses indicated that depression was related to Composite Index score, BDI score was entered as a covariate. A significant effect emerged, F (2, 106) = 4.22, p = .017, η2 = .074. Follow up tests (based on estimated marginal means) indicated that those in the HC group (M = 7.33, S.E. = .627) demonstrated significantly more cognitive insight than did those in the CD (M = 4.82, S.E. = .822) and the ND groups (M = 4.17, S.E. = 1.36), who did not differ from one another (see Fig. 2). 4. Discussion The present study was an investigation into the relationship between cognitive insight, psychotic disorder, and active delusions. While individuals with psychotic disorders, regardless of delusion status, had less overall cognitive insight than healthy controls, the factor involved in decreased insight differed depending on delusion status. Individuals with active delusions demonstrated overconfidence relative to both healthy controls and individuals with psychotic disorders without delusions. Individuals with psychotic disorders without active delusions demonstrated impaired self-reflectiveness relative to both healthy controls and individuals with current delusions. Thus, it appears that having delusions affects certain aspects of cognitive insight, while other aspects remain intact. Likewise, it appears that in the absence of delusions, cognitive insight is impaired for individuals with psychotic disorders, though it is affected by different processes relative to individuals with delusions. Individuals with psychotic disorders have previously been demonstrated to have high certainty in their own judgment when compared to individuals with severe depression without psychosis (Beck et al., 2004). Such findings suggest that it is psychotic disorder, not psychiatric disorder, that impairs cognitive insight. Until the present study, it was unclear whether individuals with psychotic disorders demonstrated differences in selfcertainty relative to healthy controls; the present findings indicate that individuals with psychotic disorders are, indeed, overly certain in their own judgment. When examined relative to delusion status, however, it was evident that overconfidence was limited to those with

331

active delusions. Individuals with psychotic disorders without active delusions had Self-Certainty scores that were the same as healthy controls; thus, individuals with psychotic disorders without active delusions do not appear to be overconfident in their judgment. The present finding adds to previous studies that have demonstrated the relationship between delusional ideation and overconfidence in judgment. Huq et al. (1988), in a study of probabilistic reasoning, found that individuals with delusions made decisions after gathering minimal relevant information, yet were quite confident about the decisions they made. Moritz et al. (2005) suggest that confidence in errors may be partly responsible for delusions; they found, in an experimental task of learning and recognition, that individuals with delusions were quite confident when they were inaccurate. Supporting the notion that overconfidence is related to acquisition of delusions, individuals who are delusion prone (i.e., individuals in the general population who do not have a psychotic disorder but who score high on a measure assessing presence of unusual beliefs) have been found to be more confident in their judgment than individuals who are not delusion prone (Warman and Martin, 2006). Thus, it appears that certainty in judgment is a particularly relevant issue for individuals with delusions. Previous research determined that individuals with psychotic disorders are significantly less self-reflective than individuals with major depression without psychosis (Beck et al., 2004). The present study did not find that individuals with psychotic disorders, overall, have impaired self-reflectiveness when compared to healthy controls. It appears, then, that individuals who are depressed may self-reflect at a level that is higher than healthy controls. Supporting that depression and selfreflection are related, in the present study higher depression was associated with higher scores on the SelfReflectiveness scale. Interestingly, Warman and Martin (2006) recently demonstrated that individuals in the general population who did not have a psychotic disorder but who scored high on a measure of delusion proneness were more self-reflective than individuals who scored lower on this measure. It is possible that high selfreflectiveness for individuals who are delusion prone is related to a risk for developing depression; Verdoux et al. (1999) found that individuals who had no history of depression but who scored high on a measure of delusion proneness were more likely to demonstrate clinically significant depression 1 year after their delusion prone assessment than individuals who scored lower on this measure. Thus, it appears that self-reflection may be healthy to a point, after which high levels may be associated with depressive thinking.

332

D.M. Warman et al. / Schizophrenia Research 90 (2007) 325–333

While higher than normal self-reflection may be associated with depression, the present study demonstrated that self-reflection significantly lower than normal was associated with psychotic disorder, but only in the absence of delusions. Individuals with psychotic disorders without delusions were significantly less self-reflective than either the healthy controls or individuals with psychotic disorders with active delusions. This relationship was unexpected and results should be interpreted with caution given the low number of participants in the present study who had psychotic disorders without delusions. However, how this finding may relate to research on decision making by individuals with delusions will be explored. A recent longitudinal study of individuals with delusions determined that after delusions remitted, decision making strategies did not change; individuals requested equivalent amounts of information before making a decision on a probabilistic reasoning task when their delusions remitted as they did when their delusions were active (Peters and Garety, 2006). Also using a probabilistic reasoning task, Moritz and Woodward (2005) found that individuals with psychotic disorders with and without current delusions requested equivalent amounts of information before making a decision and that both groups requested less information than psychiatric and healthy controls. It is possible that the strategies that lead to what appears to be similar decision making by those with and without active delusions are affected by different processes. Perhaps when individuals are actively delusional, they make judgments prematurely because they are influenced by their high confidence in their decisions. For individuals without current delusions, however, perhaps their lack of recognition of having jumped to conclusions and their lack of recognition of fallibility (i.e., self-reflectiveness) influences their hasty judgments. Moritz and Woodward found that individuals with delusions changed their answers when exposed to evidence that contradicted their original hypothesis; this finding was not demonstrated for individuals with schizophrenia without active delusions. One explanation for this finding is that it is the result of the jumping to conclusions bias, in that individuals with delusions jump to a new conclusion quickly (Peters and Garety, 2006). Alternately, perhaps this is an expression of the selfreflectiveness that is intact for those with delusions; perhaps their recognition of fallibility results in their changing their minds, though their high overconfidence continues to lead to poor decision making. It may seem counterintuitive that a person could have intact self-reflectiveness, demonstrating that they are open to feedback and recognize they could be mistaken, yet have high overconfidence. A closer examination of

the items on the BCIS, however, may aid in understanding this apparent contradiction. All of the items on the Self-Certainty subscale can be understood as relating to assessment of current judgment, such as “My interpretations of my experiences are definitely right,” and “I can trust my own judgment at all times.” In contrast, 4 of the 9 items on the Self-Reflectiveness subscale can be understood as reflecting back on previous experiences, such as “At times, I have misunderstood other people's attitudes towards me,” “I have jumped to conclusions too fast,” and “Some of the ideas I was certain were true turned out to be false.” Perhaps individuals with active delusions can recognize they have made errors in the past, but due to their overconfidence in current judgment they do not make accurate decisions. Those who did not have current delusions in the present study were able to refrain from overconfidence in current judgment, yet did not appear to recognize some past errors in judgment. It may be important to examine whether those with no delusions who have low self-reflectiveness are at higher risk for future delusional thinking than those with high self-reflectiveness; perhaps a mistaken assessment of previous judgment is a marker for relapse. A number of methodological issues limit the conclusions that can be drawn from the present study. First, the sample size for the current study was low, making any conclusions drawn only preliminary. In addition, while the purpose of the study was to compare those with psychotic disorders to healthy controls, the two groups differed in many ways other than just their psychiatric status. For example, the healthy control group was significantly younger, had far more females, and had far less diversity in terms of race representation than the group with psychotic disorders. A better comparison sample would be one that mirrors the demographics of the individuals studied with psychotic disorders. It is possible that differences that were expected to be redundant with psychotic disorder in the present study, such as gender, are important variables to study in their own right and may be related to cognitive insight. In addition, since the healthy control sample was recruited from a university setting, all of the healthy controls had at least some advanced education; while the education level of the individuals with psychotic disorders was not specifically reported, it is unlikely that all of them had some college education. It cannot be ruled out that education is related to cognitive insight, a question that may be answered in future research. In summary, individuals with psychotic disorders, regardless of whether they had active delusions or not, had impaired cognitive insight relative to healthy controls. The process “responsible” for this impaired insight, however, varied. For individuals with delusions, self-

D.M. Warman et al. / Schizophrenia Research 90 (2007) 325–333

certainty was so high that cognitive insight was impaired despite having intact self-reflectiveness. For individuals with psychotic disorders without active delusions, certainty in judgment was similar to healthy controls; their self-reflectiveness, however was impaired relative to healthy controls. The finding that individuals with psychotic disorders with and without delusions both have impairments in judgment, but that the impairments seem to be quite different ones, may have significant clinical relevance. Perhaps when treating patients with delusions, their strengths at self-reflection could be utilized and their overconfidence in judgment could be targeted for change. It is important to note that while it appears that decreasing overconfidence in judgment would likely be an important psychotherapy target, Granholm et al. (2005) found that patients whose selfcertainty improved (i.e., decreased) in treatment showed increases in depression in the middle of treatment. While their depression resolved by the end of treatment, this relationship is an important one for practitioners to be aware of so that depression can be targeted in treatment and resolved appropriately. In contrast to individuals with delusions, perhaps the ability of individuals without delusions to refrain from overconfidence could be utilized to increase their skills at self-reflection. It remains to be seen how cognitive insight is related to another significant positive symptom — hallucinations. It is possible that investigations of cognitive insight and hallucinations will help clarify the relationship between active psychosis and belief flexibility, which may lead to improved treatments for psychosis. As individuals with schizophrenia are not a homogenous group, assessing patients' cognitive insight may prove useful for determining strengths and weaknesses that could be useful in treatment. References Amador, X.F., Strauss, D.H., Yale, S.A., Flaum, M.M., Endicott, J., Gorman, J.M., 1993. Assessment of insight in psychosis. Am. J. Psychiatry 150, 873–879.

333

Beck, A.T., Warman, D.M., 2004. Cognitive insight: theory and assessment. In: Amador, X., David, A. (Eds.), Insight and Psychosis: Awareness of Illness in Schizophrenia and Related Disorders, 2nd edition. Oxford University Press, New York, pp. 79–87. Beck, A.T., Steer, R.A., Brown, G.K., 1996. The Beck Depression Inventory, Second Edition. The Psychological Corporation, San Antonio. Beck, A.T., Baruch, E., Balter, J.M., Steer, R.A., Warman, D.M., 2004. A new instrument for measuring insight: the Beck Cognitive Insight Scale. Schizophr. Res. 68, 319–329. Dozois, D.J.A., Dobson, K.S., Ahnberg, J.L., 1998. A psychometric evaluation of the Beck Depression Inventory-II. Psychol. Assess. 10, 83–89. Granholm, E., McQuaid, J.R., McClure, F.S., Auslander, L.A., Perivoliotis, D., Pedrelli, P., Patterson, T., Jeste, D.V., 2005. A randomized, controlled trial of cognitive behavioral social skills training for middle-aged and older outpatients with chronic schizophrenia. Am. J. Psychiatry 16, 520–529. Huq, S.F., Garety, P.A., Hemsley, D.R., 1988. Probabilistic judgments in deluded and non-deluded subjects. Q. J. Exp. Psychol. 40A, 801–812. Kay, S.R., Fiszbein, A., Opler, L.A., 1987. The positive and negative syndrome scale for schizophrenia. Schizophr. Bull. 10, 388–398. Moritz, S., Woodward, T.S., 2005. Jumping to conclusions in delusional and non-delusional schizophrenic patients. Br. J. Clin. Psychol. 44, 193–207. Moritz, S., Woodward, T.S., Whitman, J.C., Cuttler, C., 2005. Confidence in errors as a possible basis for delusions in schizophrenia. J. Nerv. Ment. Dis. 193, 9–16. Pedrelli, P., McQuaid, J.R., Granholm, E., Patterson, T.L., McClure, F., Beck, A.T., Jeste, D.V., 2004. Measuring cognitive insight in middle-aged and older patients with psychotic disorders. Schizophr. Res. 71, 297–305. Peters, E., Garety, P., 2006. Cognitive functioning in delusions: a longitudinal analysis. Behav. Res. Ther. 44, 481–514. Verdoux, H., van Os, J., Maurice-Tison, S., Gay, B., Salamon, R., Bourgeois, M.L., 1999. Increased occurrence of depression in psychosis-prone subjects: a follow-up study in primary care settings. Compr. Psychiatry 40, 462–468. Warman, D.M., Martin, J.M., 2006. Cognitive insight and delusion proneness: an investigation using the Beck Cognitive Insight Scale. Schizophr. Res. 84, 297–304. Warman, D.M., Dunahue, S., Martin, J.M., Beck, A.T., 2004. An investigation of the Beck Cognitive Insight Scale in the general population. Poster presented at the 38th annual meeting of the Association for the Advancement of Behavior Therapy, New Orleans, LA. November.