حافظه حقیقت گویی و کاذب برای مواد دیدنی در اختلال استرس پس از سانحه

The question whether memory aberrations in posttraumatic stress disorder (PTSD) also manifest as an increased production of false memories is important for both theoretical and practical reasons, but is yet unsolved. Therefore, for the present study we investigated veridical and false recognition in PTSD with a new scenic variant of the Deese–Roediger–McDermott (DRM) paradigm, which was administered to traumatized individuals with PTSD (n = 32), traumatized individuals without PTSD (n = 30), and non-traumatized controls (n = 30). The PTSD group neither produced higher rates of false memories nor expressed more confidence in errors, but did show inferior memory sensitivity. Whereas depressive symptoms did not correlate with veridical nor false recognition, state dissociation was positively associated with false memories.

Memory aberrations, like intrusive vivid recollections of a traumatic event, represent core symptoms of posttraumatic stress disorder (PTSD, American Psychiatric Association, 1994). Current cognitive (e.g., Brewin et al., 1996, Ehlers and Clark, 2000 and Foa et al., 1989) and neurobiological (e.g., Elzinga and Bremner, 2002 and van der Kolk, 1994) models of PTSD emphasize the key role of mnestic dysfunctions in the development and maintenance of the disorder (for reviews see Brewin and Holmes, 2003 and Dalgleish, 2004). Furthermore, numerous studies have shown that impaired memory functioning is demonstrated not only for autobiographical content, but also for non-autobiographical material (for reviews see Brewin et al., 2007 and Vasterling and Brailey, 2005). There are several pieces of evidence to assume that mnestic aberrations in PTSD also manifest as an increased production of false memories. For example, neuropsychological studies demonstrated impairments in initial learning (Bremner et al., 1993 and Vasterling et al., 2002), a heightened sensitivity to retroactive interference (Vasterling et al., 1998 and Yehuda et al., 1995), as well as a greater difficulty to inhibit inaccurate responses (i.e., high rates of commission errors and false positive responses) in PTSD (Vasterling et al., 1998). The latter authors also found an association between false positive responses and PTSD re-experiencing symptoms. Generally, impairments of working memory (i.e., the active maintenance and manipulation of information) have been associated with poor monitoring abilities and shown to correlate with false memory production in healthy subjects (e.g., Peters et al., 2006 and Peters et al., 2007). In addition, the impaired retrieval of specific versus overgeneral autobiographical memories, which has been observed in PTSD (for a review see Moore & Zoellner, 2007), points to deficient source monitoring (for reviews on cognitive abnormalities in PTSD see, e.g., Isaac et al., 2006 and McNally, 2006). Notably, another body of research has suggested an increased production of false memories in PTSD due to dissociation, (e.g., Bremner et al., 1992 and Dancu et al., 1996). The umbrella term “dissociation” includes alterations in memory, identity, and consciousness (Lynn and Rhue, 1994 and van der Hart and Horst, 1989), which would all be expected to naturally interfere with monitoring processes (Johnson, 2006 and Zoellner et al., 2000) making dissociation one of the most discussed correlates of memory suggestibility (Wright & Livingston-Raper, 2002). In fact, dissociation has been associated with false memory production in both healthy and patient samples (e.g., Clancy et al., 2000, Dehon et al., 2008 and Winograd et al., 1998; but see also Geraerts et al., 2005 and Wright et al., 2005). Finally, there is an ongoing debate in the literature concerning the relationship of trauma and false autobiographical memories, which was first sparked by a considerable controversy particularly concerning the validity of recovered memories of childhood sexual abuse (CSA) in the 1990s (see, e.g., Geraerts and McNally, 2008 and Loftus and Davis, 2006). In fact, the question of false memory propensity in PTSD is important for both the understanding of the disorder and from a practical (i.e., juridical) perspective. Traumatic incidences are often witnessed only by the alleged perpetrator and the victim, making the validity of the victim’s memory crucial in finding the truth. As memory lapses for the traumatic experience are common in PTSD, the credibility of the victim is often undermined in the eye of juries, judges, or lawyers. Knowing more about the memory signature of PTSD therefore has very practical implications and may remove common misconceptions. 1.1. Deese–Roediger–McDermott paradigm, false memories and PTSD An experimental paradigm allowing the concurrent investigation of veridical and false memory production is the Deese–Roediger–McDermott (DRM) paradigm (Deese, 1959 and Roediger and McDermott, 1995). Numerous DRM studies have consistently demonstrated that after learning word lists (e.g., hill, climb, valley, summit, top), previously not presented but highly semantically associated items (“critical lures“, in the example: mountain) were falsely recalled or recognized in a subsequent memory test (for an extensive review see Gallo, 2006). Five previous studies have used variants of the DRM paradigm to investigate whether PTSD is associated with an increased production of false memories. These studies yielded mixed results: Bremner, Shobe, and Kihlstrom (2000) found higher false and lower correct recognition of neutral DRM word lists in a sample of women with PTSD following CSA compared to women with CSA and without PTSD, as well as non-traumatized controls. In a study by Brennen, Dybdahl, and Kapidzic (2007), a higher production of false memories for trauma-related but not for neutral DRM word lists was demonstrated in a PTSD sample relative to traumatized controls. Zoellner et al. (2000) found that both a PTSD and a trauma-exposed non-PTSD group produced higher rates of false recognition for neutral DRM word lists than non-traumatized controls. A recent study (Goodman et al., 2011) investigated false memory for trauma-related, negative, neutral, and positive DRM word lists in adults and adolescents with and without PTSD following CSA. Regression analyses showed a relation of PTSD severity with less proficient memory monitoring for trauma-lists. Only one study (Jelinek, Hottenrott, Randjbar, Peters, & Moritz, 2009) used a visual variant of the DRM paradigm to compare a PTSD sample with traumatized and non-traumatized controls. In this visual variant, false memories were assessed for critical lures from complex pictures (lures were related to but not shown in the picture). Here, groups did not differ with regard to false memory. At trend level, the PTSD group showed lower corrected recognition than non-traumatized controls. In summary, although there is some evidence for an association of PTSD and false memories from verbal DRM tasks, results do not yet allow a definite conclusion. Besides inconsistent findings, the extent to which the results of studies using (verbal) variants of the DRM paradigm may be generalized to the production of real-life autobiographical memories is being controversially discussed in literature (e.g., Pezdek & Lam, 2007; but see also Wade et al., 2007; for a review see Gallo, 2010). This question, however, is of utter importance in the context of PTSD, as the validity of trauma memories is repeatedly challenged (e.g., in court). Though classical DRM studies have greatly contributed to our basic understanding of false memories, in order to advance our comprehension of memory functioning in PTSD, this line of research should be extended to more ecologically valid settings. 1.2. Moderating factors of veridical and false memory production Veridical and false memory is affected by various factors which may have led to inconsistent results in the literature. As outlined above, dissociation is being discussed as one of these factors. Surprisingly, none of the aforementioned DRM studies with PTSD samples reported a significant relationship between dissociation and false memory production. Whereas four of five studies (Bremner et al., 2000, Goodman et al., 2011, Jelinek et al., 2009 and Zoellner et al., 2000) included a trait measure for dissociation, a state measure has not yet been employed, even though, as a state characteristic, dissociation is particularly likely to influence encoding processes. Further possible moderators for false memory production are valence of the stimulus material as well as arousal during encoding. However, findings vary with regard to whether more or less false memories occur for negative, positive, or neutral stimuli. While in some DRM studies with healthy participants affective (i.e., negative, positive) word lists produced more false memories for critical lures than neutral lists (e.g., Brainerd et al., 2010, Dehon et al., 2010 and Howe et al., 2010), other studies reported less false memories for emotional lists (e.g., Palmer & Dodson, 2009). Similarly, with regard to arousal, more (e.g., Corson & Verrier, 2007) or less (e.g., Smeets et al., 2006) source monitoring errors have been found under conditions of high versus low arousal, independent of stimulus valence. Thus, evidence is contradictory for whether stimulus valence and arousal enhance or disrupt memory accuracy (for a review of fMRI and behavioural studies see Mitchell & Johnson, 2009). Results from the eyewitness literature even point to interactions between valence of information (emotional/unemotional), type of detail information (central/peripheral), retention time (immediate/delayed), or type (free recall/recognition) of test (for a review see Christianson, 1992). Findings on how valence and arousal influence false memory production in PTSD are still scarce. There is some evidence pointing towards an interaction between valence and test type with trauma-related or negative material producing more false memories in free recall (Brennen et al., 2007 and Goodman et al., 2011) but better memory sensitivity in recognition tests (Goodman et al., 2011). In these studies, no separate measures of arousal were administered. Evidence from information-processing studies in PTSD suggests a heightened accessibility, thus superior correct recognition, of trauma-relevant relative to neutral material in individuals with PTSD (e.g., Golier et al., 2003, Litz et al., 1996 and Paunovic et al., 2002). However, it remains unclear whether this would transfer to higher or lower rates of false recognition for trauma-related or other arousing material in DRM paradigms. A further important factor linked to false memory production is metamemory, which was also investigated in some of the aforementioned studies involving PTSD patients, for example, by assessing memory confidence (e.g., item was “probably” or “definitely” presented/not presented before). Although individuals with PTSD were expected to display greater confidence (i.e., more high confident responses) for their false memories, no significant group differences were evident (Brennen et al., 2007, Jelinek et al., 2009 and Zoellner et al., 2000). 1.3. The present study The primary aim of the present study was to examine veridical and false memory in PTSD by means of a scenic variant of the DRM paradigm using stimuli of varying emotional content, including trauma-related material. In addition, the role of possible underlying mechanisms (state and trait dissociation, depression, arousal), as well as memory confidence was investigated. On the basis of the majority of findings from the literature, we hypothesized that participants with PTSD show impaired memory accuracy resulting in (1) higher false recognition and (2) lower correct recognition rates compared to traumatized and non-traumatized controls. Furthermore, we expected an association of false memory production and dissociation. Finally, we aimed to explore the role of stimulus valence, subjective arousal during encoding, as well as memory confidence in the recognition test.

In line with the literature, the present study demonstrated a small impairment of visual memory sensitivity in PTSD. However, results neither provide evidence for increased visual false memories nor higher confidence herein in individuals with PTSD, thus, speaking against a greater memory suggestibility as a disorder-specific characteristic. Our results indicate that dissociation, a PTSD-unspecific but related factor, is linked with false memory in particular state dissociation at the time of encoding the information. Future studies should further investigate visual memory in PTSD, explicitly examining the role of state dissociation in false memory production with a potential focus on PTSD subgroups.