حافظه کاذب در بیماران اسکیزوفرنی با و بدون هذیان

Delusions are fixed ‘false beliefs’ and, although a hallmark feature of schizophrenia, no previous study has examined if delusions might be related to ‘false memories’. We used the classic Deese–Roediger–McDermott (DRM) paradigm to compare false memory production in schizophrenia patients who were currently experiencing delusions (ED), patients not experiencing delusions (ND) and healthy control participants. The ED group recalled twice as many false-positive memories (i.e., memory for words not previously seen) as both the controls and crucially, the ND group. Both patient groups also recognised fewer correct words than the healthy controls and both showed greater confidence in their false memories; however, on the recognition task, the ED group made more false-negative (i.e. rejecting previously seen words) high confidence responses than the ND group.

Fixed false beliefs (i.e., delusions) are a hallmark feature of schizophrenia presenting in virtually all persons with schizophrenia at some time in the course of their illness, (Moritz and Woodward, 2002). Understanding the cognitive basis of schizophrenia symptoms is a major goal of psychological schizophrenia research that could have implications for treatment. One important class of schizophrenia symptom is delusions — false, often bizarre beliefs about persecution, ideas of grandeur and so on. Although delusions are perhaps the most common and central feature of schizophrenia, to date, little consensus exists about their psychological or neuropsychological basis (see Gilleen and David, 2005). Nevertheless, parallels may be drawn with the fact that healthy participants can be induced to make false recollections. The notion of false memories relates to either remembering events that never happened or remembering them quite differently from the way they happened (Roediger and McDermott, 1995). Researchers investigating false memories have used numerous techniques to demonstrate the phenomenon, although one method that has been widely used is the Deese–Roediger–McDermott (DRM) paradigm (Deese, 1959 and Roediger and McDermott, 1995). This method involves presenting participants with lists of words and later testing their memories for the words. Each list of words has strong associates to a critical, target word (often the strongest associate with the list), e.g., needle was the critical unpresented target word for the following presented list: . Following presentation, participants are given a free recall test and after all presentations, they receive a recognition test. Roediger and McDermott found that participants recalled the unseen target word in 40% of the lists, and other unseen words in 14% of the lists. Laws and Bhatt (2005) used the same DRM technique in healthy participants dichotomised into high and low ‘delusion-proneness’ groups (based on the Peters Delusional Inventory: PDI: Peters et al., 1999). They found that, as the score on this scale increased so did the number of false memories in the DRM paradigm and that participants with high delusional ideation also attached greater confidence to false- positive errors. Memory dysfunctions have been repeatedly shown within schizophrenia (Aleman et al., 1999). Specifically, studies using the DRM paradigm have established that simple memory intrusions are quite common in patients with schizophrenia (Huron and Danion, 2002, Moritz and Woodward, 2002, Elvevåg et al., 2004 and Moritz et al., 2005) and that they may exhibit increased memory confidence for false memories (Moritz and Woodward, 2002 and Moritz et al., 2005, 2006). In Moritz and Woodward's (2002) study, the findings suggested that the ‘memory responses rated with high confidence by patients with schizophrenia contain a large number of intrusions’. More recent studies report that patients produce a greater percentage of high confidence responses that are errors (this is referred to as knowledge corruption1 by Moritz and colleagues) and significantly so for false-negative errors (Moritz et al., 2004 and Moritz et al., 2006). Moritz et al. (2005) proposed that paranoid schizophrenia patients display a stronger tendency to trust information that is actually incorrect, whereas healthy controls are more cautious in their evaluation of information that turns out to be incorrect. Overconfidence in errors was thought to arise from the impaired ability to cast doubt on fallible information, whereas healthy controls were able to attach a so-called ‘not trustworthy’ tag to such representations. Finally, at the neurological level, patients with frontal lobe lesions show a well-documented susceptibility to false recognition/recall and problems with memory monitoring processes (e.g., Janowsky et al., 1989 and Schacter and Slotnick, 2004). Given that patients with schizophrenia display poor performance on tests of frontal lobe function (e.g., the Wisconsin Card Sort Test: see Laws 1999) and moreover, functional neuroimaging studies have revealed hypofrontality in patients with schizophrenia (for a review, see Hill et al., 2004), such factors may also underpin such memory distortions in patients with schizophrenia. In the current study, we used the Roediger and McDermott (1995) paradigm to investigate recall, recognition and memory confidence in schizophrenia patients with and without delusions and healthy controls. Laws and Bhatt (2005) speculated whether their findings would relate to those of studies of patients with schizophrenia. A review of studies (using the same DRM paradigm) appears inconsistent with Laws and Bhatt (2005) insofar as schizophrenia patients display no greater incidence of false-positive errors than healthy controls (Huron and Danion, 2002, Elvevåg et al., 2004, Moritz et al., 2004 and Lee et al., 2006). All of these studies examined schizophrenia patients per se rather than focussing specifically on comparing schizophrenia patients who are experiencing delusions (ED) or not experiencing delusions (ND). Our main aim was to determine whether schizophrenia patients with delusions show a greater tendency to create false memories compared both to schizophrenia patients without delusions and to healthy controls. Furthermore, we aimed to determine if schizophrenia patients with delusions display greater knowledge corruption than schizophrenia patients without delusions and healthy controls.

Descriptive statistics of the total, correct, incorrect, target and other words recalled were computed for each of the three groups (See Table 3). Table 3. Recall performance for both patient groups and healthy controls. ED ND Control (C) ANOVA M (SD) M (SD) M (SD) Recalled correctly (n = 120) 41.23 (15.18) 37.67 (17.88) 83.45 (13.51) ED = ND < C Target words (n = 8) 5.31 (1.25) 2.67 (1.50) 2.55 (2.01) ED > ND = C Other words 8.08 (7.87) 12.00 (15.83) 2.45 (3.71) ED = ND > C M = mean, SD = Standard Deviation. Table options One way ANOVAs were conducted to investigate differences among the three groups in the mean number of correct words, other words and target words recalled.2 The results revealed significant group effect for correct words, F(2, 42) = 45.39, P < 0.001. The pairwise comparisons revealed two contrasts that were significant for the number of correct words recalled (both P < 0.001: between the healthy controls and both the ED and ND groups. A significant group effect also emerged for target words (false memories) recalled F(2,42) = 11.89, P < 0.001. A follow-up contrast analysis revealed no difference between controls and ND patients, while the ED patients were worse than both controls and the ND (see Table 3). Finally, for other words, the result also revealed a significant group difference, F(2, 42) =4.05, P < 0.03, but with ND group making more other errors than healthy controls (d = 1.5, P = 0.008). These analyses confirmed that any false memories were not solely a product of poor memory since both patient groups showed no significant difference in mean correct or other words recalled, although the ED group made more target errors.