اختلال در یادگیری مبتنی بر احساسات انعطاف پذیر در افراد در معرض و هذیان

Deficits in emotion-based learning are implicated in many psychiatric disorders. Research conducted with patients with schizophrenia using one of the most popular tasks for the investigation of emotion-based learning, the Iowa Gambling Task (IGT), has largely been inconclusive. The present study employed a novel, contingency-shifting variant IGT with hallucination- and delusion-prone university students to determine whether previous findings were due merely to the presence of psychosis. Following initial screening of a sample of 253 students (mean age = 20.13 years, S.D. = 3.27), 28 high (10 male, 18 female) and 27 low (12 male, 15 female) hallucination-prone and 27 high (7 male, 20 female) and 26 low (11 male, 15 female) delusion-prone individuals completed the contingency-shifting variant IGT. Results showed no significant differences between the performances of high and low hallucination- and delusion-prone individuals during the original phase of the task. Differences only emerged following the onset of the contingency-shift phases, with individuals high in hallucination- and delusion-proneness having impaired performance compared with low hallucination- and delusion-prone individuals. Overall, the present findings demonstrate that impairments associated with hallucination- and delusion-proneness are specific to the shift phase of the contingency-shifting variant IGT, which supports previous findings with patients with schizophrenia.

Deficits in emotion-based learning or emotional decision-making are implicated in several psychiatric disorders, including schizophrenia (e.g., Lawrence et al., 2006, Martino et al., 2007, Must et al., 2006 and Sevy et al., 2007). The evidence from research conducted using the Iowa Gambling Task (IGT; Bechara et al., 1994 and Bechara et al., 2000) with patients with schizophrenia is, however, largely inconclusive. Some studies ( Evans et al., 2005 and Ritter et al., 2004) have shown that patients perform at levels comparable to healthy participants, while other studies ( Lee et al., 2007, Martino et al., 2007 and Shurman et al., 2005) have shown that patients with schizophrenia engage in disadvantageous decision-making compared to healthy controls. In seeking to explain these findings, it is important to acknowledge the contribution of factors such as the relatively small sample sizes, the influence of medication, comordid diagnoses, and the heterogeneity of symptoms within the diagnosis of schizophrenia itself ( Dunn et al., 2006 and Sevy et al., 2007). Recently, Turnbull et al. (2006) suggested that people with schizophrenia might not show consistent deficits on the IGT because the original task does not adequately tap flexibility in emotion-based learning. Until now, researchers have relied on tasks that separately index set-shifting and reversal learning ability to infer the role of flexible emotion-based learning in IGT performance, with a number of studies showing that people with schizophrenia perform relatively poorly (Pantelis et al., 1999 and Waltz and Gold, 2007). Recently, Rodriguez-Sanchez et al. (2005) reported that first episode schizophrenia patients have unimpaired IGT performance, yet have impaired performance on one of the most widely used measures of executive functioning and set-shifting ability: the Wisconsin Card Sorting Test (WCST; see also Prentice et al., 2008). These authors also found that performance on the IGT was not correlated with WCST performance. Lee et al. (2007) reported impaired IGT performance and, similar to Rodriguez-Sanchez et al. (2005), an absence of correlations between WCST ability and IGT performance. Lee et al. (2007) also found that performance on the Simple Reversal Learning Task (SRLT; Fellows and Farah, 2003) was impaired in people with schizophrenia, relative to healthy controls, but was not associated with performance on the IGT in either of the groups. Both the Rodriguez-Sanchez et al. (2005) and Lee et al. (2007) studies failed to find correlations between set-shifting ability, as measured by the WCST, and reversal learning ability, as measured with the SRLT, and IGT performance. Turnbull et al. (2006) recently developed a novel, contingency-shifting version of the IGT. In the contingency-shifting version, the reinforcement contingencies of the card decks were shifted following initial exposure to the original IGT trials such that card decks that had previously been advantageous became disadvantageous, and vice versa. Turnbull et al. (2006) compared a group of patients with schizophrenia who were classified as either high or low in positive and negative symptomatology with a healthy control group. Results showed that patients high in positive and negative symptoms initially learned at levels comparable to the healthy controls during the original IGT, supporting some previous studies (e.g., Rodriguez-Sanchez et al., 2005 and Ritter et al., 2004). During the contingency-shift phase of the task, however, those patients high in negative symptoms exhibited markedly poorer performance in adjusting to the changing contingencies relative to both healthy controls and those patients high in positive symptoms, suggesting that deficits associated with schizophrenia are specific to the shift phases. In this way, the contingency-shifting variant IGT may be useful for research on emotion-based learning in the schizophrenia spectrum because it adds a cognitive component consistently shown to be impaired in schizophrenia (Waltz and Gold, 2007). Differentiating between emotion-based learning deficits that may reflect core pathological processes in schizophrenia and the impact of symptomatology on the reported behavioural deficits is important in understanding the emotional and cognitive determinants of psychopathology. An intriguing means of addressing this question is to examine flexible emotion-based learning with the contingency-shifting variant IGT in non-clinical samples that have elevated psychosis-proneness scores (Johns and van Os, 2001 and Verdoux and van Os, 2002). Hallucination- and delusion-proneness are two of the most prominent features of psychosis-proneness that appear to be dimensionally distributed across the general population (Johns and van Os, 2001). Previous studies have investigated how measures of proneness to psychosis relate to measures of set-shifting, such as the WCST (e.g., Nieuwenstein et al., 2001, Suhr, 1997 and Suhr and Spitznagel, 2001). Performance on the contingency-shifting IGT and WCST is, however, likely to be reliant on a number of relatively diverse cognitive processes, such as working memory, attention and response inhibition, and some caution is therefore necessary when interpreting previous findings. Nonetheless, it seems plausible to suggest impairment in IGT contingency-shifting performance in individuals high in hallucination- and delusion-proneness. The aim of the present study, therefore, was to extend the findings of Turnbull et al. (2006) with the contingency-shifting variant IGT by examining the performance of non-clinical groups high on hallucination- and delusion-proneness. We hypothesised that high and low psychosis-prone individuals would not differ in their performance during the original IGT trial blocks but would differ significantly during the contingency-shifting phases.