|Home | About | Journals | Submit | Contact Us | Français|
Factor analytic studies have shown that in schizophrenia patients, disorganization (conceptual disorganization, bizarre behavior) is a separate dimension from other types of positive symptoms such as reality distortion (delusions and hallucinations). Although some studies have found that disorganization is more strongly linked to neurocognitive deficits and poor functional outcomes than reality distortion, the findings are not always consistent.
A meta-analysis of 104 studies (combined n = 8,015) was conducted to determine the magnitude of the relationship between neurocognition and disorganization as compared to reality distortion. Additional analyses were conducted to determine whether the strength of these relationships differed depending on the neurocognitive domain under investigation.
The relationship between reality distortion and neurocognition was weak (r = -.04; p=.03) as compared to the moderate association between disorganization and neurocognition (r = -.23; p <.01). In each of the six neurocognitive domains that were examined, disorganization was more strongly related to neurocognition (r’s range from -.20 to -.26) than to reality distortion (r’s range from .01 to -.12).
The effect size of the relationship between neurocognition and disorganization was significantly larger than the effect size of the relationship between neurocognition and reality distortion. These results hold across several neurocognitive domains. These findings support a dimensional view of positive symptoms distinguishing disorganization from reality distortion.
Researchers studying the effects of “first generation” antipsychotic medications have usually relied on positive symptom severity, typically defined as “psychotic relapse,” to evaluate the efficacy of new medications. In many of those studies, positive symptoms such as hallucinations and delusions were combined with conceptual disorganization to form a positive symptom factor (Guy 1976). Over time, several factor analytic studies supported using a three factor model that included positive symptoms, sometimes referred to as reality distortion, negative symptoms, and disorganization. As a result, disorganization has emerged as separate domain worthy of consideration (Bilder et al. 1985; Liddle 1987; Arndt et al. 1991; Toomey et al. 1997; Brekke et al. 2005; Cuesta et al. 2007). In fact, several researchers have suggested that we should consider completely separating symptom dimensions in studying the course and outcome of schizophrenia, i.e., viewing reality distortion (delusions and hallucinations) as independent from disorganization (e.g., conceptual disorganization, bizarre behavior). Furthermore, symptoms of disorganization have been identified as risk factors for a worse course of illness (Shenton et al. 1992; Reed et al. 2002; Metsanen et al. 2004; Metsanen et al. 2006). Providing additional support for a separate examination of positive symptoms, some evidence suggests that disorganization might be a stronger predictor of community functioning than reality distortion (Norman et al. 1999; Ventura et al. 2009).
Several cross-sectional studies have suggested that performance on neurocognitive tests is only weakly correlated with positive symptoms (Roy and DeVriendt 1994; Davidson and McGlashan 1997; Rund et al. 1997; Addington and Addington 1999; Addington and Addington 2000; Brazo et al. 2002; Brazo et al. 2005; Ventura et al. 2009). Studies examining disorganization have found stronger ties to neurocognition than for reality distortion (Nieuwenstein et al. 2001; Dibben et al. 2008; de Gracia Dominguez et al. 2009). Neurocognitive domains such as executive functions have been more robustly linked to disorganization than reality distortion (Aleman et al. 1999; Nieuwenstein et al. 2001; Dibben et al. 2008). Understanding how the relationship of disorganization to neurocognition is different from that of reality distortion might help in understanding the factors that can contribute to a patient’s poor cognitive functioning and ultimately poor functional outcome.
Given the importance of neurocognition to the course and outcome of schizophrenia, hypotheses about the existence of a symptom-based “cognitive disorganization” factor have been proposed. For instance, Bryson and colleagues (Bryson et al. 1999) found that performance on neurocognitive tests was related to cognitive disorganization when using data collected with the Positive and Negative Syndrome Scale (PANSS). Furthermore, in a 5-factor model created using PANSS data, investigators identified a “cognitive factor” (Lindenmayer et al. 1995). Yet, several of the PANSS symptom items that load on this cognitive factor, e.g., conceptual disorganization, mannerisms and posturing, are typically thought of as symptoms associated with disorganization in schizophrenia. The interchangeable use of such labels as “disorganization” and “cognitive disorganization” underscores the importance of clarifying the construct of disorganization and differentiating these components from other types of positive symptoms in relationship to neurocognitive functioning.
Several meta-analytic studies have found differential relationships between positive symptoms and neurocognitive functioning when examining disorganization versus reality distortion, or when combining both symptom clusters. Aleman and colleagues (Aleman et al. 1999) conducted a meta-analysis of 70 studies that compared the performance of schizophrenic patients and healthy controls on measures of verbal and nonverbal memory impairment. The authors concluded that positive symptoms did not have a moderating effect on memory in schizophrenia. However, positive symptoms that were comprised of reality distortion and disorganization were combined and the separate effects of disorganization were not reported. In a 2001 meta-analysis, disorganization was shown to have a significant positive correlation with Wisconsin Card Sorting Test (WCST) perseverations (average r = .25), but not with attention (average r = .06) as measured by the Continuous Performance Test (CPT) (Nieuwenstein, Aleman et al. 2001). However, symptoms of reality distortion did not correlate significantly with either of the two neurocognitive measures. In a meta-analysis of 88 studies (Dibben, Rice et al. 2008) a small-to-moderate effect was found for the relationship between executive function and disorganization (effect size r = -.17) as compared to reality distortion (effect size r = .01). Recently, (de Gracia Dominguez et al. 2009) in a meta-analysis of 58 studies found that IQ and several neurocognitive domains were more correlated with disorganization than was reality distortion. These meta-analytic studies consistently found that disorganization was related to neurocognition, but that reality distortion was not. However, relatively little is known about how these two types of symptoms affect various domains of neurocognition such as those defined in the MATRICS project (Nuechterlein et al. 2004). Those domains were identified by examining several factor analytic studies and include working memory, attention/vigilance, verbal memory, visual memory, reasoning and problem solving, and speed of processing, each of which are worthy of separate study.
Several studies that employed attribution scales and attentional or information processing tasks derived from human experimental psychology provide evidence suggesting that cognitive functions are linked with neurocognitive processes that are associated with positive symptoms (Frith et al. 1992; Hemsley 1993; Bentall et al. 2001; Blackwood et al. 2001; Dibben et al. 2008; Guillem et al. 2008). There are links between certain information processing abnormalities such as poor signal detection, and cognitive misattributional processes, such as “overgeneralization” that underlie delusional thinking. Information processing or attentional disturbances are theoretically related not only to the formation, but the maintenance of delusional beliefs. Patterns of performance of acute schizophrenic patients in these experiments are consistent with cognitive “psychological” models. Knowing more about the differential magnitude of the relationship between reality distortion and neurocognition (as measured by objective tests) could help explain the interrelationship of these variables.
The aim of this meta-analysis was to replicate and expand the examination of the relative strength of the relationship of neurocognition and reality distortion as compared to disorganization across a wide range of neurocognitive domains. Seven domains of cognitive functioning were identified after a thorough review of published factor analyses (Nuechterlein et al. 2004). We aimed to systematically examine relationships between six domains of neurocognition (verbal memory, attention/vigilance, reasoning and problem solving, speed of processing, visual memory, and working memory) and domains of positive symptoms (reality distortion and disorganization). We hypothesize a larger effect size for the relationship between disorganization and neurocognition as compared to the effect size of the relationship between reality distortion and neurocognition. Furthermore, we expected that combining reality distortion and disorganization into a single dimension would show an intermediate effect size between the relatively pure domains (i.e., reality distortion and disorganization).
We conducted a literature search of the following databases: PsychInfo, PsychAbstracts, EBSCOhost, PubMed, and Google Scholar covering the period from January 1, 1977 to December 31, 2008. Searches were restricted to articles published in the English language. We used the following key search terms (some terms were combined): neurocognition, neuropsychology, schizophrenia, disorganization symptoms, positive symptoms, formal thought disorder. We also used the search options in PubMed and Google Scholar that allow for a search of papers with related topics. In addition, the reference lists of published articles identified by this method were then screened to locate additional relevant studies.
Using these methods, 176 articles were identified as potentially relevant to this topic. These studies were then evaluated using the following inclusion criteria: (1) Study must have used empirical methods and been published in a peer reviewed journal; (2) study must have contained descriptions of study measures and operational definitions of variables; (3) study must have used structured assessments of symptoms with established scales or standardized methods of symptom assessment; (4) study must have assessed neurocognitive functioning using standardized batteries; (5) study must have been cross-sectional (as defined by an assessment interval of 90 days or less); (6) all participants in the study must have been diagnosed with schizophrenia or schizoaffective disorder according to DSM criteria; (7) statistics reported must have been correlation coefficients or other statistics that could be converted into correlations so that an effect size and z score could be calculated; (8) study data must not have been included or published previously in another paper.
Seventy-two papers did not meet these criteria and were excluded from the study. A total of 104 studies met the inclusion criteria (see Table 1), with a combined total sample of 8,015 patients. The aggregate sample characteristics were as follows: 69% of patients were male, the mean age was 37.5 years, and the mean education was 12.1 years. From the data base of 104 studies, we identified those that reported relationships between neurocognition and reality distortion (N = 50) and studies that reported relationships between neurocognition and disorganization (N = 69) (see Table 1). Additionally, we identified 40 studies that combined reality distortion with disorganization, e.g., the PANSS positive symptoms scale. Studies that combined reality distortion and disorganization were coded and analyzed separately to determine whether those studies reported stronger effect sizes between positive symptoms (broadly defined) and neurocognition than reality distortion alone. Studies of both inpatients and outpatients were included. Data from all of the 104 studies were compiled in a database containing: (1) the author(s) and year of publication; (2) demographic information; (3) description of the neuropsychological tests used, e.g., California Verbal Learning Test (CVLT), and the neurocognitive domain assessed by each test; (4) symptom measures, e.g., Scale for the Assessment of Positive Symptoms (SAPS), and the symptom types examined by these measures, e.g. disorganization; and (4) study statistics, e.g., correlation coefficients.
For the current study, neurocognition was operationally defined as cognitive functions, such as verbal memory and working memory that are objectively measurable with standardized neuropsychological tests, such as the WAIS Digit Span Test (Table 2). One of the primary goals of the NIMH Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS) was to develop a reliable and valid consensus cognitive battery. An essential step in the process was to identify the major separable cognitive deficits in schizophrenia. After evaluating the empirical evidence that consisted of the examination of factor analytic studies, seven separable domains were identified that were replicable across studies (Nuechterlein et al. 2004). The current study included 6 of the 7 MATRICS domains of cognitive functioning: Speed of Processing, Attention/Vigilance, Working Memory, Verbal Memory, Reasoning and Problem Solving, and Visual Memory. Social cognition was excluded from the current analyses because it is not a traditional domain of neurocognition, and at this point there are relatively fewer studies that address the relationship between positive symptoms and social cognition. The dimension of reality distortion included positive symptoms consisting of delusions, e.g., suspiciousness, and hallucinations, e.g., auditory, as measured by items from structured symptom scales, e.g. PANSS, BPRS (Table 2). The dimension of disorganization included positive symptom items, such as conceptual disorganization, formal thought disorder, mannerisms and posturing, and bizarre behavior, as measured by structured assessment scales, e.g., PANSS, SAPS (for a review, see (Bryson et al. 1999). The combined category was comprised of studies of neurocognition in which symptoms of reality distortion and disorganization were combined, e.g., by using the PANSS positive symptom scale which includes delusions, hallucinations, and conceptual disorganization, or the SAPS total score which combines delusions, hallucinations, and formal thought disorder.
For the main analysis we combined all six domains of neurocognitive functioning and created one composite neurocognitive variable to represent neurocognition. We examined the relationship between disorganization and neurocognition, and then reality distortion and neurocognition. We were primarily interested in the differences in effect size magnitude between reality distortion and disorganization on neurocognition. We also examined the relationship between neurocognition and combined positive symptoms, i.e., reality distortion and disorganization. We note that for several published studies, combining the categories of reality distortion and disorganization amounted to adding only one item to the cluster of symptoms, e.g., conceptual disorganization.
The first step for these analyses was to transform the observed (published) correlations in each study using Fisher’s r-to-z transformation. Where indicated, multiple results were averaged from the same domain, e.g., several tests of working memory were combined into a single observation for a given study. The correlation coefficients were then combined into a single estimate of the population correlation by averaging them weighted by sample size (Hedges and Olkin 1984). Based on these combined correlation coefficients, the studies were then tested for homogeneity by calculating a Q-statistic. When examining separate neurocognitive domains, heterogeneity was evident in most of the domains (see appendix). Heterogeneity of measures is a known problem in the field. We tested for association of the heterogeneity for the following variables: age, education, gender ratio, inpatient or outpatient status, and sample size. The heterogeneity of the studies was related in expected ways to heterogeneity of education, gender, and sample size. In the appendix the alternative estimates for the overall effect sizes based on the homogeneous subset studies are shown. Although a few studies were identified as outliers in some of the neurocognitive domains, we found it difficult to justify excluding one study for not being a valid source of information for a given domain, and yet valid for another. Using this rational, the study results are based on parameter estimates from all studies. For comparison purposes, parameter estimates based on the homogeneous subset studies are provided (see appendix). Although the significance of the reported p-values is potentially inflated, the data presented here can be considered as being a reasonably robust representation of the relationships between the variables of interest.
To address the primary question posed in this study, we examined separately the relationship between neurocognition and reality distortion, and the relationship between neurocognition and disorganization (see Table 3). The cross-sectional relationship between reality distortion and the composite neurocognition score was statistically significant but the effect size was weak (r = -.04, p=.03). The effect size of the cross-sectional relationship between disorganization and the composite neurocognition score was moderate and statistically significant (r = -.23, p < .01). We also examined the relationship between neurocognition and the combined category of reality distortion and disorganization symptoms. A meta-analysis of those studies (N = 40, subjects = 4,654) showed a statistically significant but weak effect for positive symptoms and neurocognition (r = -.05, p < .01). Contrary to our hypothesis, the combined category of positive symptoms did not yield intermediate correlations with the composite neurocognition score or the separate domains of neurocognition.
A number of published studies included in this study examined specific relationships between symptoms and several key domains of neurocognitive functioning. Here is a breakdown of the total number of studies in each domain: speed of processing (N = 60), reasoning and problem solving (N = 58), working memory (N = 33), visual memory (N=18), verbal memory (N = 32), and attention/vigilance (N = 29). The results indicate that more domains of neurocognition were related to disorganization as compared to reality distortion (Table 3). In each of the six MATRICS domains that were examined, neurocognition was more highly correlated with disorganization (r’s range from -.20 to -.26) than was the case for reality distortion (r’s range from = .01 to -.12). In addition, for disorganization, moderate effects were found for attention/vigilance (r = -.25), reasoning and problem solving (r = -.24), and speed of processing (r= -.26) compared to weaker relationships with reality distortion (r = -.12, r = -.06, r= -.03, respectively). Of the six MATRICS neurocognitive domains that were examined, attention/vigilance was the most highly correlated to reality distortion (r = -.12, p < .01), still a small effect size but larger than the effects of the remaining domains which were near zero (r’s range from .01 to -.06; Table 3).
Meta-analytic techniques were used to examine studies of positive symptoms defined as reality distortion and disorganization in relationship to neuropsychological functioning in schizophrenia. Consistent with the previous literature, our results demonstrated a moderate (r = -.23) relationship between disorganization and neurocognition, while the relationship between neurocognition and reality distortion was relatively weak (r = -.04). We found evidence that disorganization was related to all domains of cognitive functioning we examined, whereas reality distortion showed no such broad association. Our findings suggest that disorganization represents a separate set of positive symptoms from reality distortion with independent links to neurocognition. Our findings support theory and results from several studies in schizophrenia suggesting that positive symptom factors should be considered separately (Grube et al. 1998; Peralta and Cuesta 1999; Bell and Mishara 2006).
Our analyses suggest that for many of the previous studies which combined different types of positive symptoms, the observed relationships between positive symptom factors and the other study variables might have obscured important findings. For example, the PANSS positive symptom cluster combines hallucinations, delusions, excitement, and hostility, with conceptual disorganization. Yet, several investigators (Toomey et al. 1997; Grube et al. 1998; Peralta and Cuesta 1999; Stuart et al. 1999; Peralta and Cuesta 2001) have suggested that perhaps studies should consider correlating variables of interest with single symptom items, such as hallucinations. Our analysis support the necessity of at least distinguishing between the two broad constructs of reality distortion and disorganization. In addition, disorganization and cognitive disorganization appear to be related symptom concepts that overlap. Disorganization usually includes conceptual disorganization, mannerisms and posturing, disorientation, and bizarre behavior whereas cognitive disorganization includes those items and stereotyped thinking, difficulties in abstract thinking, poor attention, and inappropriate affect. The two symptom groups are most likely highly correlated even though they contain slightly different symptoms most of which are correlated with neurocognitive deficits. To validate the concept of cognitive disorganization, future meta-analyses would need to separately examine these two clusters of symptoms to determine their intercorrelation and relationship to neurocognition.
The notion that disorganization is a separate positive symptom factor is also supported by studies examining the relationship between positive symptoms and community-based functioning. Several studies indicated that the relationship of disorganization to community-based functioning is relatively strong (Reed et al. 2002; Smith et al. 2002; Evans et al. 2004; Takahashi et al. 2005), while a meta-analysis showed that the relationship between reality distortion and community-based functioning was weak (estimated r = -.03; (Ventura et al. 2009). Thus, while symptoms such as hallucinations and delusions might not consistently interfere with a person’s ability to socialize or to perform at work, the data suggest that disorganization symptoms might be more closely linked to impairments in day-to-day functioning. Patients might learn to compensate for reality distortion symptoms in various ways, e.g., ignoring beliefs about aliens during a social interaction or while working in a retail clothing store, but disorganization symptoms might cause more of a disruption in daily functioning, e.g. disorganized speech interfering with functional aspects of communication.
We found that reality distortion was relatively more highly correlated with attentional deficits (r = -.12, p <.01) than other domains of neurocognition (r’s ranging from .01 to -.06). This association between attentional deficits and reality distortion is consistent with findings and theory from a series of studies conducted in the late 1980s and early 1990s by British researchers such as Richard Bentall, Chris Firth, and David Hemsley (for a review see, Blackwood et al. 2001). According to experiments in human psychology, positive symptoms such as delusions and/or hallucinations, are believed to be associated with attentional and information processing deficits. These sorts of distortions in cognitive processing are thought to be related to misattributions and altered perceptions of environmental cues, such as reliance on internal states, and situational context. Delusions are believed to be associated with an attentional bias toward threatening information and a tendency to misattribute meaning, which usually occurs in accordance with a patient’s set of rigid expectations (Frith 1979; Magaro 1980; Bentall et al. 1991; Frith et al. 1992; Hemsley 1993). For example, delusional patients have been shown to be overconfident in their responses and require less information before “jumping to a conclusion”(Garety et al. 1991). In addition, hallucinations have been associated with impairments in attentional processes and perceptions that lead to the intrusion of unintended information into conscious awareness (Bentall and Slade 1985; Bentall et al. 1991). Finally, reality distortion was not associated with deficits in most of the neurocognitive domains we examined, which is in accordance with the British theorists and predictions by Liddle and colleagues (Liddle and Morris 1991).
There are several limitations to this study that warrant mentioning, some of which are common to all meta-analytic investigations (Rosenthal 1991; Lipsey and Wilson 2001). For example, the study sample was not randomly selected. Also, we believe based on theoretical considerations that several domains of neurocognition are an underlying cause of the severity of disorganization, but the data examined in this meta-analysis was cross-sectional rather than longitudinal in design, thus the choice regarding which variables are conceptualized as “cause” and which to consider an “effect” is essentially arbitrary. Additionally, neurocognition is not a homogenous concept and its definition in this meta-analysis is influenced by how common a particular set of neurocognitive tests appear in the published literature. For example, our reasoning and problem solving domain is defined largely by the WCST. Also, we note the possibility that measurement overlap resulted in an inflated correlation between neurocognition and disorganization symptoms. For example, the PANSS includes an item that assesses difficulties in abstract thinking in the disorganization factor. Despite the fact that each of these study limitations suggest that caution should be used in interpreting the results, our findings still provide some direction for future research on neurocognition and disorganization. We want to emphasize that a meta-analysis cannot replace focused empirical research.
The strong relationship between neurocognition and disorganization would profit from further examination. If causality is validated through longitudinal research, this relationship would have implications for intervention in schizophrenia. Considering the central role that neurocognitive deficits play in relationship to daily functioning in schizophrenia, it is not surprising that cognitive deficits have emerged as important targets for new treatments (Green and Nuechterlein 1999; Carpenter and Gold 2002; Carpenter 2004; Gold 2004). If neurocognition is related to disorganization, then perhaps improvement in community functioning could be mediated by improvements in disorganization. Also, our finding that attentional deficits were most closely related to reality distortion is supportive of interventions which improve attentional control, e.g., attentional training, or interventions that specifically target the cognitive component of delusions and hallucinations, such as Cognitive Behavior Theory for psychosis.
This research was supported in part by National Institute of Mental Health Grants R21MH07391 (PI: Joseph Ventura, Robert Bilder, Co-PI, Steven Reise, Co-PI), MH37705 (P.I.: Keith H. Nuechterlein, Ph.D.), and P50 MH066286 (P.I.: Keith H. Nuechterlein, Ph.D.).
Funding Source The funding source did not play role in the design, implementation, results, or publication of this paper.
|Reality distortion||r||Type of N||Critical N||Q statistic||p||Significant study characteristics||Number of Studies dropped for homogeneity||r||P|
|Reasoning and Problem solving||-.06||FD||485||32.05||.23|
|Speed of processing||-.03||S||3033||67.59||<.01||Gender, Sample Size||Bell et al., 1994||.01||.76|
|Combined Category||r||Type of N||Critical N||Q statistic||p||Significant study characteristics||Number of Studies dropped for homogeneity||r||P|
|Working Memory||-.06||FD||1895||30.58||<.01||Education||Bressi et al., 1996|
Pukrop et al., 2003
|Reasoning and Problem solving||-.03||S||898||54.44||<.01||(*)||Braff et al., 1991|
Bryson, Bell, Grieg, 1999
|Speed of processing||-.02||S||8439||26.05||.19|
|Disorganization||r||Type of N||Critical N||Q statistic||p||Significant study characteristics||Number of Studies dropped for homogeneity||r||P|
|Verbal Memory||-.20||FD||22907||45.94||<.01||(*)||Eckman and Shean, 2000, Mohamed et al., 1999||-.19||<.01|
|Visual Memory||-.20||FD||9286||27.95||.01||Gender||Mohamed et al.,1991||-.23||<.01|
|Reasoning and Problem solving||-.24||FD||74399||77.03||<.01||Sample Size||Franke et al., 1992,|
Klingberg, Wittorf, Wiedemann, 2006,
Nestor et al., 1998
|Speed of processing||-.26||FD||101862||64.68||.01||(*)||Eckman & Shean, 2000,|
Mohamed et al., 1991,
Hofer et al., 2007
|Attention||-.25||FD||31184||40.32||<.01||(*)||Mohamed et al., 1999||-.27||<.01|
Note: 1) Only 2 studies were outliers in more than one of the neurocognitive domains: Eckman and Shean, 2000 (2 domains) and Mohamed et al., 1999 (4 domains) whereas all other studies showed discrepancies only in one domain. 2) for Type of N, S = number of studies with the same effect size required to get a significant result and FD = number of unpublished studies with an r of 0 required to make this result not significant (file drawer problem). 3) for Significant Study Characteristics an (*) indicates that study heterogeneity was not associated with any of the study parameters.
Joseph Ventura conceived of the study design, data analysis plan, conducted literature searches, supervised the conduct of the study, and wrote the manuscript. Dr. Hellemann conducted the data analysis and commented on drafts of the manuscript. Dr. Thames preformed literature searches, created tables, and commented on drafts of the manuscript. Ms. Wood preformed literature searches, created tables, and commented on drafts of the manuscript. Ms. Guzik conducted literature searches, organized study papers, and created tables. All authors have contributed to and approved the final manuscript.
The authors report no conflict of interest
The findings from this meta-analysis were presented in part at the 11th bi-annual meeting of the International Congress on Schizophrenia Research, Ventura, J., Thames, A.D., Hellemann, G.S. Disorganization in Schizophrenia: Positive Symptom or Neurocognitive Deficit. March 28 – April 1, 2007, San Diego, California.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.