خاطرات مزاحم در افسردگی و اختلال استرس پس از سانحه

Abstract This study compared the stressors and consequent intrusive memories reported by matched samples of patients with posttraumatic stress disorder (PTSD) and major depression. Although intrusive memories were slightly more common among PTSD patients, both quantitative and qualitative measures revealed few differences between the groups. PTSD patients were more likely to have experienced personal illness or assault, and depressed patients family deaths and illness, and interpersonal events. Factor analysis of the associated emotions and memory characteristics suggested the existence of specific links between fear and reliving, and helplessness and out-of-body experiences. Possible inhibitory relationships between fear and sadness, and between guilt and anger, were also noted.

1. Introduction Systematic empirical research on posttraumatic stress disorder (PTSD) is of relatively recent origin, and has been heavily influenced by definitions of the disorder contained in the diagnostic and statistical manuals produced by the American Psychiatric Association. Important though these have been, they inevitably rested on some untested assumptions that are now coming under greater scrutiny. Many important diagnostic issues relating to PTSD remain to be clarified (Davidson & Foa, 1991), and our understanding of the disorder is undergoing important changes (e.g. Yehuda & McFarlane, 1995). In this article we examine the assumption that intrusive memories of stressful or traumatic events are a distinctive feature of PTSD, by conducting qualitative and quantitative comparisons of intrusive memories in PTSD and major depressive disorder. Although DSM-IV lists re-experiencing symptoms as a feature of PTSD and does not identify them as a component of depression, previous work has suggested that depressed patients experience intrusive memories at a level equivalent to PTSD patients (Kuyken & Brewin, 1994; Brewin, Hunter, Carroll, & Tata, 1996). However, no study has directly compared this phenomenon in matched samples of patients with PTSD and depression. PTSD shares a number of clinical characteristics with other psychiatric disorders and is rarely diagnosed in isolation. Comorbidity with depression is especially common, whether in samples of inpatients with PTSD (Green, Lindy, & Grace, 1985), community samples (Davidson, Hughes, Blazer, & George, 1991; Helzer, Robins, & McEvoy, 1987; Shore, Vollmer, & Tatum, 1989), Vietnam veterans (Davidson & Foa, 1991), or underground train drivers (Farmer, Tranah, O'Donnell, & Catalan, 1992). Several authors, noting the high comorbidity of depression and PTSD, have pointed to the substantial symptom overlap between the two, with many depressive symptoms appearing in sections C and D of the diagnostic and statistical manual's (DSM-III-R: APA, 1987) criteria for PTSD (Farmer et al., 1992; McNally, 1992). Other symptoms such as guilt are commonly found in both conditions. There are also striking similarities between depression and PTSD in cognitive processing. Depression is, in general, associated with an increased access to negative memories and a decreased access to positive memories (e.g. Williams, Watts, MacLeod, & Mathews, 1997). Anxiety disorders, in contrast, have been linked by Williams et al. to attentional rather than to memory biases. PTSD, however, is by definition a disorder involving repeated unwanted access to memories of a traumatic incident, and in this respect shares important characteristics with depression. Another characteristic and reliable aspect of memory functioning in depression is overgeneral memory, the inability to retrieve a specific autobiographical memory to cue words such as `successful' or `lonely'. Instead of retrieving an episode that occurred at a particular time and place, the depressed tend to produce memories relating to a series of incidents or to a whole period in their life. This problem of overgeneral recall has been repeatedly found in depressed patients (Moore, Watts, & Williams, 1988; Williams & Scott, 1988; Williams, 1992; Kuyken & Brewin, 1995), and also appears to be a feature of PTSD (McNally, Litz, Prassas, Shin, & Weathers, 1994; McNally, Lasko, Macklin, & Pitman, 1995). Consistent with DSM-IV, the intrusive symptoms of PTSD have generally been regarded as characteristic of anxiety disorders, rather than depression (e.g. Davidson & Foa, 1991). McNally (1992) suggested that the distinctive features of PTSD are the exaggerated startle, the re-experiencing symptoms (such as nightmares or intrusive memories of the trauma) and physiological reactivity to trauma-related cues. However, intrusive waking memories, nightmares, and physiological arousal are typical of normal responses to major life stressors including bereavement (see Brewin, Dalgleish, & Joseph, 1996, for a review). Moreover, recent studies of depressed psychiatric patients have found that around 86% describe experiencing repetitive intrusive memories (Kuyken & Brewin, 1994; Brewin et al., 1996b). In the first study to address this issue, (Kuyken & Brewin, 1994) interviewed depressed women about experiences meeting research criteria for childhood physical or sexual abuse and asked whether they had been experiencing intrusive memories of the abuse in the week prior to interview. Women completed the impact of event scale (IES: Horowitz et al., 1979) in respect of these memories. The IES is a short questionnaire much used in research on posttraumatic stress disorder and has two subscales, one measuring the extent to which memories intrude, and one the extent to which the individual has been trying to exclude the memories from consciousness. Kuyken and Brewin found that scores on this measure were equivalent to those reported in published studies of patients diagnosed with PTSD. Moreover, higher levels of intrusion and avoidance were related to indices of more severe abuse. Women with abuse and high scores on the IES were significantly more depressed than women not reporting abuse. Similar findings were obtained by Brewin et al. (1996b) with a mixed-sex sample of depressed patients. In this study patients were asked about a variety of recent and past life events, and 27/31 reported corresponding intrusive memories. On average each patient reported between 2–3 memories, with abnormally high levels of intrusion and avoidance for each. Other studies have confirmed that depressed patients are significantly more likely to experience intrusive memories than controls matched on level of stress, and that these memories were in most cases either not present before the individual became depressed, or were not so intrusive (Spenceley & Jerrom, 1997; Brewin, Watson, McCarthy, Hyman, & Dayson, 1998). Given that depression is associated with a history of severe life events, which may include physical and sexual abuse in childhood (see Brewin, Andrews, & Gotlib, 1993, for a review) and marital violence (Andrews & Brown, 1988), it is perhaps not surprising that some patients should experience high frequency intrusive memories. A critical question, then, is whether memories of traumatic and nontraumatic events reported by the depressed differ qualitatively or quantitatively from those experienced by patients with PTSD. If they do, this would support the claim that intrusive memories are a distinctive symptom of PTSD. If they do not, it may be necessary to revise our conceptualization of PTSD, perhaps downplaying the importance that is usually given to intrusive memories as a characteristic feature. In either case, the investigation is likely to illuminate further the nature of both depression and PTSD. Numerous authors have claimed that memories for highly emotional events, and traumatic memories in particular, have different properties to those of other memories. In some cases individuals retain vivid, detailed images for long periods (`flashbulb memories': Brown & Kulik, 1977), although this is not invariably the case and is related to several factors including the personal significance of the event (Conway, 1995). Research on individuals with PTSD has also found that their intrusive memories are characterized by great vividness and persistence, and additionally that they are accompanied by high levels of fear, are very distressing and contain strong sensory and perceptual features. Some of the memories (commonly known as `flashbacks') described by patients with PTSD differ from ordinary autobiographical memories in that they often consist of parts or fragments rather than whole memories, and they are often experienced as a reliving of the event in the present (Terr, 1991; Foa, Molnar, & Cashman, 1995; Van der Kolk & Fisler, 1995; Brewin et al., 1996a). These features are believed to reflect an underlying dissociative process (Van der Kolk & Fisler, 1995), and other dissociative phenomena such as out-of-body experiences are sometimes reported. It should be noted, however, that traumatic memories, particularly in nonclinical samples, may not always share these characteristics. One study found that rape memories, compared to other unpleasant memories, were less clear and vivid, involved less visual detail, and were less well-remembered (Tromp, Koss, Figueredo, & Tharan, 1995). Studies of combat veterans have also noted that intrusive memories are as often accompanied by emotions of sadness or anger as they are by fear (Pitman, Orr, Forgue, de Jong, & Claiborn, 1987; Pitman et al., 1990). Drawing on previous theories of psychological change (Brewin, 1989 and Brewin, 1996), Brewin et al.'s dual representation theory of PTSD explained these varied findings by proposing that there are two distinct types of intrusive memory (Brewin et al., 1996a). One type is based on verbally accessible representations, and consists essentially of ordinary autobiographical memories that can be retrieved and edited at will. These memories are likely to typify nonclinical samples, such as that of Tromp et al. (1995). They are believed to be associated with secondary emotions such as sadness, anger and guilt, that reflect the individual's conscious appraisal of the causes of the trauma and of its implications for future aspirations and goals. The second type of intrusive memory is based on representations that cannot be deliberately accessed, instead being automatically accessed by internal or external cues. These representations, they suggested, underlie intrusive memories consisting of the involuntary, intense reliving of all or part of the traumatic event (the flashbacks). These memories are believed to be associated with the experience of emotions (predominantly fear, helplessness and horror) that were felt while the trauma was actually occurring, rather than later. Brewin, Dalgleish, and Joseph argued that this second type of intrusive memory had not been clearly identified in the literature on response to stressful life events, and that it was the presence of this type of memory that ought to distinguish PTSD from other conditions, such as abnormal grief reactions or depression. PTSD and depression also differ, however, in that the former is meant to be a response to death, serious injury, or threat to physical integrity, whereas the antecedents of depression are not formally specified. Thus, for diagnostic reasons, the events preceding the two disorders are likely to differ, and this fact alone might be responsible for quantitative or qualitative differences in patients' intrusive memories. In comparing disorders, therefore, it would seem essential to analyze any additional effects attributable to the events featuring in patients' memories. Whereas depressed patients frequently do not meet diagnostic criteria for PTSD, the degree of comorbidity for depression in PTSD patients means that it would be neither practicable nor representative to obtain a sample of PTSD patients who are uniformly nondepressed. In this study we therefore compared the stressful experiences and intrusive memories described by samples of patients with PTSD (with or without depression) and depression (without PTSD). Following DSM-IV (APA, 1994), we expected that the PTSD group, relative to the depressed group, would more often report experiencing, witnessing, or being confronted with events that involved actual or threatened death or serious injury, or a threat to the physical integrity of self or others, and that their intrusive memories would more often be characterized by fear, helplessness and horror. In accordance with the theoretical accounts provided by Foa, Steketee, and Rothbaum (1989), Van der Kolk & Fisler (1995) and Brewin et al. (1996a), we also predicted that the intrusive memories of the PTSD patients would be more vivid and distressing, and that they would more often involve physical sensations, feelings of reliving the event and out-of-body experiences. Finally, we predicted that the memory characteristics of vividness, reliving, distress, physical sensations and out of body experiences would be associated with the emotions of fear, helplessness and horror, but not with the emotions of sadness, anger and guilt. Other memory characteristics recorded, but about which no specific predictions were made, included their frequency and duration, as well as the extent of intrusion and avoidance as measured by the impact of event scale.

. Results 3.1. Number of memories, intrusion and avoidance Forty-two of the 43 patients with PTSD and 45 of the 62 depressed patients reported experiencing intrusive memories in the past week. The PTSD group reported a mean of 1.47 memories and the depressed group a mean of 1.10 memories. This difference was significant, t(103)=2.61, p<0.01. For the most prominent intrusive memory, the IES Intrusion scores were significantly higher in the PTSD group (x̄=26.5) than in the depressed group (x̄=23.1), t(83)=2.25, p<0.05. There were no differences in the IES avoidance scores for the first memory (PTSD x̄=24.1, depressed x̄=22.7), t(83)<1, or in the length of time since the event featuring in the memory occurred, t(85)=1.02, p>0.10. For the second most prominent memory, the two groups did not differ in their IES intrusion scores (PTSD x̄=24.0, depressed x̄=23.4) or in their IES avoidance scores (PTSD x̄=22.3, depressed x̄=23.7), in both cases t(40)<1. 3.2. Content of memories Within the depressed sample, we identified 29 patients whose intrusive memories concerned events that would likely be described as traumatic according to DSM-IV, i.e. they involved actual or threatened death or serious injury, or a threat to the physical integrity of self or others. These were essentially similar to the memories described by the PTSD sample. The remaining 16 memories involved nontraumatic events. More detailed content analysis on the memories of the depressed and PTSD samples combined indicated that 93% of the events featuring in patients' intrusive memories were able to be categorized under four headings: death, illness or injury to family members; illness or injury to the patient; assault on the patient (including physical and sexual assaults either in childhood or adulthood); and interpersonal problems (including breakdown of intimate relationships and serious disputes with friends, family, or neighbors). The number of patients reporting events in these four categories for their first intrusive memory is given in Table 1. The depressed group were more likely to report memories consisting of death, illness or injury to family members, or interpersonal problems, whereas the PTSD group were more likely to report memories consisting of personal illness or injury, or personal assault, χ2(3)=22.5, p<0.001. Table 1 also shows that the distribution of memories between the groups is similar, if cases in which a patient describes more than one type of event are included. Table 1. Content of most prominent intrusive memory in patients with depression and PTSD (figures in parentheses include additional content of second most prominent memory where appropriate) Type of event Depressed (N=43) PTSD (N=39) Family death, illness or injury 17(20) 10(14) Illness or injury to the patient 3(6) 14(16) Assault on the patient 8(10) 14(16) Interpersonal problems 15(20) 1(6) Table options 3.3. Emotions and memory qualities The emotions most frequently mentioned spontaneously by patients as being associated with either their first or second most prominent memories were anger (59 mentions), sadness (43 mentions), fear (35 mentions), helplessness (25 mentions) and guilt (16 mentions). All other emotions were mentioned no more than 10 times. For each of the five most frequent emotions, patients were assigned a score of one if they mentioned an emotion in respect of either or both memories and zero if they failed to mention the emotion. The PTSD group mentioned helplessness significantly more often than the depressed group (60% vs. 33%, t(85)=1.96, p<0.05). There were no other differences between the groups, largest t(85)=1.50, p>0.10. (Subsidiary analyses comparing the PTSD group with the subset of depressed patients reporting nontraumatic memories yielded similar results). The two groups were next compared in respect of the reported qualities of their intrusive memories. The memories were overwhelmingly described as vivid (88% of the time), with 12% described as including some detail. None were described as unclear. In the sample as a whole, 12 intrusive memories lasted for seconds, 56 for minutes, 30 for up to an hour, 20 for several hours and 10 constantly preoccupied patients. There were no group differences in how long the memories lasted on average, t(83)<1. Thirty of the memories occurred once a week or less and 99 occurred several times a week or more. The mean frequency was significantly higher in the PTSD group, t(84)=2.14, p<0.05. The mean level of distress was 7.9/10, with no group differences, t(85)<1. Seventy-four percent of the PTSD group and 62% of the depressed group reported physical sensations accompanying at least one of the memories, t(85)=1.15, p>0.10. Seventy-four percent of the PTSD group and 71% of the depressed group reported a sensation of reliving associated with at least one of the memories, t(85)<1. Forty-two percent of the PTSD group and 20% of the depressed group reported an out-of-body experience associated with one of their memories, and this difference was significant, t(83)=2.29, p<0.05. (Subsidiary analyses comparing the PTSD group with the subset of depressed patients reporting nontraumatic memories yielded similar results, with the exception that there were no longer any group differences in memory frequency). 3.4. Emotions and memory qualities analyzed by memory content The next analysis involved testing whether emotions and memory qualities varied according to the four types of memory content previously identified. In order to achieve sufficient numbers, the patient groups were combined, and only data from the first intrusive memory were used. Each of the emotions was used as a dependent variable in separate one-way ANOVAs, with memory type as a factor with four levels. A significant main effect was followed up by post hoc (least significant difference) tests on the individual means. Table 2 shows that memory content had strong effects on which emotions were spontaneously mentioned by patients. Sadness was mainly associated with other-focused events such as family deaths or illnesses and interpersonal problems, whereas fear was mainly associated with self-focused events such as personal injury or assault. Guilt was particularly associated with family death or illness and helplessness with personal illness or injury. In contrast, anger was common across all event types. Table 2. Emotions associated with different memory content Type of memory Sadness Anger Guilt Fear Helplessness Family death, illness or injury 0.56b 0.37 0.33a 0.07a 0.22b Illness or injury to the patient 0.24ac 0.59 0.12b 0.35b 0.41a Assault on the patient 0.09c 0.59 0.05b 0.41b 0.09b Interpersonal problems 0.44ab 0.50 0.00b 0.06a 0.00b F(3,78) 4.97∗ 1.01 4.55∗ 4.46∗ 4.06∗ ∗ p<0.01. In each column, means followed by different letters differ significantly (p<0.05). Table options Memory content was next used as a factor in separate one-way ANOVAs looking at the various dimensions of memory quality. The overall effect of reliving was significant, F(3,78)=3.12, p<0.05. Post hoc tests showed that reliving was more highly associated with personal assault and than with the other three memory types (p<0.05). There were no other main effects associated with any of the memory qualities, largest F(3,78)=2.34, p>0.05. 3.5. Exploratory analysis: associations between memory characteristics and emotions In view of the high level of similarity in the emotions and memory characteristics reported by the two groups, their data were pooled for the final analysis. The five most common emotions reported and the memory characteristics of duration, distress, reliving, physical sensations and out-of-body experience were entered into a factor analysis in order to detect underlying patterns of association (vividness was not included owing to its lack of variability). Because the main aim was an empirical summary of the dataset we selected the principal components method of extraction, and this was followed by Varimax rotation. The analysis produced five factors with eigenvalues greater than 1, accounting for 17, 14, 13, 12 and 10% of the variance, respectively. The data are shown in Table 3. Taking item loadings of 0.40 and above as significant, factor 1 was characterized by high levels of distress, physical sensations and long duration of the intrusive memory. Factor 2 reflected high levels of sadness and low levels of fear and physical sensations. Factor 3 was characterized by high levels of helplessness and out-of-body experiences. Factor 4 reflected high levels of fear and reliving. Finally, factor 5 was characterized by low levels of anger and high levels of guilt. Table 3. Factor solution for emotions and memory characteristics Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Sadness 0.03 0.84 0.04 −0.09 0.10 Anger −0.05 0.03 −0.17 0.29 −0.75 Guilt 0.00 0.11 −0.21 0.21 0.72 Fear 0.08 −0.59 0.11 −0.61 0.05 Helplessness −0.14 0.25 0.61 0.04 −0.21 Reliving −0.03 0.13 −0.11 −0.84 0.01 Distress 0.77 0.27 0.24 −0.10 −0.09 Duration 0.63 −0.03 −0.39 0.05 0.13 Physical sensations 0.66 −0.42 0.02 0.06 0.06 Out-of-body experience 0.13 −0.20 0.74 0.04 0.15