اجتناب شناختی از خاطرات مزاحم: چشم انداز بهتر و ارتباط با افسردگی

Abstract Although recent research demonstrates that intrusive memories represent an overlapping cognitive feature of depression and post-traumatic stress disorder (PTSD), there is still a general paucity of research investigating the prevalence and maintenance of intrusive memories in depression. The current study investigated the association between a range of cognitive avoidant mechanisms that characterize PTSD samples (i.e., suppression, rumination, emotional detachment, and an observer vantage perspective) and intrusive memories of negative autobiographical events in relation to dysphoria. Hypotheses were based on the proposition that employment of these cognitive mechanisms would hinder the emotional processing of the negative event, thus contributing to the maintenance of intrusions. Results supported an association between negative intrusive memories, dysphoria, and avoidant mechanisms. Significant differences were also found between field and observer memories and measures of emotional detachment and rumination. Implications relating to intrusive memory maintenance and treatment approaches are discussed.

Introduction Intrusive memories A growing body of evidence has demonstrated that intrusive memories of negative past events are reported in depression samples (Brewin, Hunter, Carroll, & Tata, 1996; Carlier, Voerman, & Gersons, 2000; Kuyken & Brewin, 1994). Kuyken and Brewin (1994) initial study investigating the presence of intrusive memories in depressed females who experienced childhood abuse revealed that levels of intrusions, as measured by the Impact of Event Scale (IES; Horowitz, Wilner, & Alvarez, 1979), were comparable to reported IES scores of post-traumatic stress disorder (PTSD; APA, 1994) patients. Further, levels of intrusions and avoidance on the IES positively correlated with severity of reported abuse and depression severity (Kuyken & Brewin, 1994). Brewin et al. (1996) extended these initial findings by investigating intrusive memories in a sample of both female and male depressed patients. The authors investigated intrusive memories of a broader range of experiences and events (i.e., not solely confined to a history of abuse). Intrusive memories reported by this sample were classified into four main categories; illness/death, relationship/family problems, abuse/assault, and work/financial problems. Common emotional responses corresponding to these memories included guilt, sadness, helplessness, anger, shame, and anxiety. In support of Kuyken and Brewin (1994), the results indicated that depression severity was positively correlated with the number of intrusive memories associated with past abuse. Additionally, a relationship was observed between the duration of the current depressive episode and the total number of memories reported (Brewin et al., 1996). Evidence also supports a relationship between intrusive memories and the maladaptive cognitive styles that characterize depression. For example, depressed individuals with more intrusive memories of abuse reported lower self-esteem, a more negative attribution style and greater avoidance (Kuyken & Brewin, 1999). Extending these findings beyond cross-sectional studies, Brewin, Reynolds, and Tata (1999) found that in a sample of clinically depressed patients, avoidance of intrusive memories was predictive of depression symptoms at 6-month follow-up, after controlling for baseline levels of depression (Brewin et al., 1999). Additionally, Brewin, Watson, McCarthy, Hyman, and Dayson (1998) demonstrated an association between intrusive memories and depression in a longitudinal study of matched samples of depressed cancer patients. They found that the presence and avoidance of intrusive memories at baseline predicted anxiety levels at 6-month follow-up. These longitudinal findings indicate that intrusive memories play an important role in the course of depression, and, importantly, demonstrate that intrusive memories are more than an epiphenomenon of depression. Investigating the similarities of intrusive memories in PTSD and depression, Reynolds and Brewin (1999) found significant qualitative overlap in the intrusive memories of depression and PTSD samples, prompting their conclusion that ‘the presence of intrusive memories does not distinguish PTSD from major depression as clearly as might be inferred from diagnostic checklists’ (p. 212). Taken together, these studies confirm that intrusive memories are a key cognitive characteristic of depression, and are implicated in the persistence of the disorder. Furthermore, intrusive memories represent a shared feature of PTSD and depression. Due to this overlap, it is important to first outline the relationship between cognitive avoidance and PTSD symptoms in order to present a rationale for the operations of a similar process in depression. Cognitive avoidance mechanisms in PTSD and depression Cognitive conceptualizations of PTSD (e.g., Ehlers & Clark, 2000; Foa, Steketee, & Rothbaum, 1989) posit that successful emotional processing of the trauma and trauma intrusions is prevented when the individual employs avoidant cognitive strategies. Cognitive avoidance can take a range of forms, including intentional attempts at thought/memory suppression, efforts to dissociate or detach oneself from the affective qualities of the trauma experience, and engaging in rumination (Ehlers & Clark, 2000). To date, empirical research documenting the use of these avoidant coping strategies by depressed individuals in response to their intrusive autobiographical memories is minimal. Little is known of the cognitive mechanisms that are employed by depressed samples to manage intrusive memories, and in particular, of the degree to which avoidant strategies are associated with the maintenance of intrusive memories in depression. However, a number of lines of research do converge on the proposal that avoidant strategies may play a key role in mediating the management of intrusive memories in depression. First, thought suppression is a commonly employed strategy used by depressed individuals in an effort to avoid the pervasive negative intrusions that typify depressive thinking ( Beevers, Wenzlaff, Hayes, & Scott, 1999; Wenzlaff, 1993). Additionally, Reynolds and Brewin (1998) found that although suppression of intrusive cognitions was commonly reported as a response strategy by depressed individuals, it was rated as an ineffective coping response. In the same way that suppression of trauma information has been hypothesized to contribute to re-experiencing symptoms by preventing information from being processed conceptually, suppression of intrusive memories by depressed individuals may result in similar processing failures. Second, in the context of depression, rumination has been defined as repetitive but passive thinking about possible precipitating factors, current symptoms, and the consequences of these depressive symptoms (Nolen-Hoeksema, 1991). Prospective studies have convincingly linked rumination to depression onset (Nolen-Hoeksema, Morrow, & Fredrickson, 1993) and maintenance (Nolen-Hoeksema, 2000). Although rumination appears to involve the processing of symptom-relevant information, ruminating in response to intrusive memories may not translate into successful emotional processing of the memory. In relation to PTSD, rumination involves thinking about the causes and consequences of the trauma, thus avoiding direct reliving of the traumatic event which is proposed to interfere with the consolidation of the trauma memory ( Ehlers & Clark, 2000). Thus, rumination in depression may be construed as a cognitive avoidance mechanism in that it prevents the processing of emotionally relevant information from being integrated into conceptual memory. Third, dissociative responses, such as responding to intrusive memories in a detached manner and reports of emotional numbness, may also be a mechanism adopted by depressed individuals in response to negative intrusion. In PTSD, dissociation is theorized to impede the accessibility of trauma memories and their associated affect, thereby disallowing the processing of traumatic material that is necessary for the integration of trauma memories, and thus recovery from PTSD (Foa & Hearst-Ikeda, 1996, as cited in McIsaac & Eich, 2004). A similar, although not necessarily parallel process at the level of encoding may also operate in depression thus preventing integration of the affective qualities of the intrusive memory. In the current investigation we conceptualized ‘dissociation’ to refer to emotional disengagement without assuming reference to unconscious processes. Thus, while these lines of research provide evidence that depression is associated with avoidant cognitive strategies such as thought suppression and rumination, the use of these strategies in the management of intrusive memories has been minimally investigated. Starr and Moulds (2006) studied the role of avoidant cognitive strategies (i.e., thought suppression, rumination, dissociation) and negative interpretations of intrusions by drawing on the model of PTSD maintenance of Ehlers and Steil (1995). The authors reported significant associations between depression scores and indices of rumination and dissociation in response to negative intrusive memories. This study provides evidence of the endorsement of avoidant mechanisms in response to negative intrusive memories and provides support for the validity of drawing on PTSD models to help inform and guide parallel research in depression. In the current investigation, we were interested in both situational and prototypical responses to negative intrusive memories, thus we expanded upon the Starr and Moulds (2006) study by including additional measures of rumination and suppression to assess characteristic avoidant responding. Vantage perspective as a cognitive avoidance mechanism A potentially important dimension of intrusive memories in depression that may be linked to cognitive avoidance is the vantage perspective from which the memory is recalled. Memories recalled in the first-person are experienced from the individual's original or ‘field’ perspective, whereas memories recalled in the third-person are experienced from an ‘observer’ perspective. Field perspective memories contain more information on affective, physical, and physiological states, while observer memories contain more descriptive and less affect-laden information (Nigro & Neisser, 1983). Additionally, it has been found that field memories generally arise in relation to recent events, whereas observer memories are more commonly reported in relation to older events (Nigro & Neisser, 1983; Robinson & Swanson, 1993). It has been suggested that adopting a third-person perspective in the context of intrusive trauma memories may serve as a cognitive avoidance mechanism similar to suppression techniques (McIsaac & Eich, 2004). In effect, it may function as a means of removing oneself from reliving the specific event by becoming a ‘detached spectator’, and both research and clinical observations indicate that this is common in victims of disasters and assaults (Cardena & Spiegel, 1993; Foa & Rothbaum, 1998, as cited in McIsaac & Eich, 2004). In addition, it has been speculated that adopting the third-person vantage point may hinder the emotional processing of the trauma event. ‘Emotional processing’ is thought to require integration of both the cognitive and affective components of the intrusive memory (Foa & Kozak, 1986). The adoption of the observer perspective could prevent the integration of the affective components of the memory by inhibiting the individual from attending to these elements in favor of emphasizing the concrete and objective details of the original experience (McIsaac & Eich, 2004). Since emotional activation, and subsequent integration of this information is argued to be critical in modifying the frequency of intrusive memories, adopting an observer perspective may therefore interfere with exposure-based therapies (McIsaac & Eich, 2004), thus contributing to the maintenance of intrusions. In a sample of PTSD patients, McIsaac and Eich's (2002) empirical findings demonstrated that naturally recalled observer memories contained less affective information and were in fact experienced as less anxiety-eliciting than field memories. Eighty-nine percent of those who recalled their trauma from the observer perspective also reported that they did so in an effort to avoid reliving the experience. These results provide initial clinical evidence of the potential role of recall vantage perspective in the regulation of affective reactions in trauma. On the basis of preliminary evidence that the intrusive memories of PTSD and depression samples share common features (Reynolds & Brewin, 1999) and management strategies (Starr & Moulds, 2006), it may be that an observer perspective of intrusive memories in depression is linked to: (i) reduced emotional impact, and (ii) additional cognitive avoidance strategies. To date, the vantage perspective of intrusive memories has not been investigated in the context of depression. Aims of the current investigation Addressing the above-mentioned absences in the existing literature, the primary aims of the present study were to firstly establish the nature of the relationship between negative intrusive memories and dysphoria, and more specifically to: (ii) investigate the association between dysphoria and cognitive avoidance mechanisms, (iii) assess the rates of naturally occurring field and observer vantage perspectives of intrusive memories, (iv) ascertain whether adoption of a first- or third-person vantage perspective has differential effects on the reported experience of intrusive memories, (v) establish whether recall vantage perspective is associated with well-established cognitive avoidance strategies, and (vi) assess whether any associations between recall vantage perspective and cognitive avoidance are unique to dysphoria. We sought to investigate these objectives in a non-clinical sample. It was first hypothesized that intrusion-related distress would correlate with dysphoria (as indexed by BDI-II scores) irrespective of intrusion frequency. It was additionally hypothesized that participants would utilize cognitive avoidance mechanisms defined by measures of suppression and rumination in response to intrusive memories that elicited distress. As dissociative phenomena such as reports of numbness and detachment in response to intrusions are also conceptualized as indices of cognitive avoidance, it was hypothesized that those reporting negative intrusions and distress would endorse items of dissociation or emotional disengagement. As recall vantage perspective has not been assessed in relation to intrusive memories in the context of depression, hypotheses were guided by McIsaac and Eich's (2004) findings in relation to trauma memories. It was hypothesized that intrusive memories based on recent events would be more likely to occur in the first-person perspective and conversely, that intrusive memories based on older events would be more likely to occur in the third-person perspective. It was hypothesized that intrusive memories experienced in the first-person would be associated with greater ratings of distress, while memories recalled from a third-person perspective would demonstrate a relationship to measures of cognitive avoidance, in particular measures of emotional disengagement. Finally, it was expected that any observed associations between an observer vantage perspective and cognitive avoidance would be more evident in the dysphoric sub-sample (BDI-II⩾12).

Results Sample characteristics Preliminary screening of the data revealed multivariate outliers for one participant's scores across BDI-II and RRS scores. Further inspection revealed a pattern of elevated endorsement across multiple measures including intrusion frequency therefore data from this respondent was excluded from all subsequent analyses. Additional outliers appeared to be randomly dispersed across variables and since removal of these items had little effect on the relevant analyses or significance levels they were retained. This resulted in a sample of 102 females and 45 males with a mean age of 19.70 (SD=4.18). Regarding scores on the BDI-II, 53% percent of the samples scores fell within the minimal depressive range, 29% within the mild-to-moderate range, 13% within the moderate-to-severe range, and 5% within the severe range. Thirteen percent of the sample met criteria for a current major depressive episode using the SCID diagnostic criteria (First et al., 1996). The mean frequency of negative intrusions was 3.70 (SD=5.09). Memory content of intrusions was coded into four categories based on Brewin et al. (1996); interpersonal (58%), death/illness involving other (10%), illness/injury to self (12%), personal assault/abuse (4%), and other (16%). This final category consisted of atypical memories such as one participant's account of ‘seeing a ghost’ and variations of personal failure. Intrusion-related distress and dysphoria Dysphoria correlated with intrusion distress (r=0.25, p<0.01), IES Total (r=0.34), IES Intrusion (r=0.31), and IES Avoidance (r=0.24) subscales (all p<0.001). To ensure that intrusion frequency was not accounting for these associations, partial correlations controlling for frequency were conducted. All correlations remained significant (p<0.01). Cognitive avoidance mechanisms and dysphoria To assess the relationship between cognitive avoidance mechanisms and dysphoria, multiple Pearson correlations were conducted. Results are reported in Table 1. Table 1. Means and standard deviations of avoidance mechanisms and correlations with dysphoria controlling for intrusion frequency (n=147) M (SD) BDI-II Partial r Rumination 12.66 (3.92) 0.31⁎⁎ 0.28⁎⁎ ‘Worry it could happen again’ 3.45 (2.02) 0.30⁎⁎ 0.29⁎⁎ ‘Dwell on it’ 4.01 (1.73) 0.28⁎⁎ 0.27⁎⁎ ‘Think what I could do differently’ 4.79 (2.05) 0.06 0.03 Ruminative response scale (RRS) 49.45 (12.81) 0.57⁎⁎ 0.55⁎⁎ Suppression 11.01 (4.61) 0.16 0.18⁎ ‘Think about something else’ 3.55 (1.76) 0.09 0.09 ‘Push it out of my mind’ 3.09 (2.03) 0.23⁎ 0.23⁎ ‘Watch TV’ 3.57 (1.96) 0.06 0.04 WBSI 47.13 (12.17) 0.40⁎⁎ 0.39⁎⁎ WBSI intrusion 29.18 (7.73) 0.39⁎⁎ 0.37⁎⁎ WBSI suppression 17.94 (5.56) 0.35⁎⁎ 0.35⁎⁎ Dissociation 76.25 (45.61) 0.21⁎ 0.21⁎ ‘Detached’ 47.75 (27.34) 0.16 0.15 ‘Numb’ 28.50 (28.26) 0.19⁎ 0.17⁎ Note: All p's Bonferroni adjusted. ⁎ p<0.05. ⁎⁎ p<0.01. Table options As expected, dysphoria demonstrated a significant association with rumination items with the exception of the individual item “Think what I could do differently’ (r=0.06, p>0.05). Levels of dysphoria also correlated with suppression with the exception of two individual items “Think about something else” (r=0.09, p>0.05) and “Watch T.V.”(r=0.06, p>0.05). Dissociative responses demonstrated a weak relationship with levels of dysphoria. Dysphoria did correlate with ratings of numbness (r=0.19, p<0.01), but not with ratings of detachment once the significance level was adjusted for multiple comparisons (r=0.16, p=0.02). Partial correlations controlling for intrusion frequency are also reported in Table 1. All correlations remained significant with the exception of the Suppression total score (p=0.06). Intrusive memory characteristics and recall vantage perspective Across all participants, 56% of the intrusive memories were reportedly experienced/recalled in the first person, 32% were recalled in the third person perspective, and the remaining 12% reported a blended perspective encompassing both recall modalities (inclusive of scores −2 to 2 on the vantage perspective scale). Interpersonal events were the most frequently reported content of both the field (60%) and observer (54%) memories. Commonly reported examples within this category were arguments with family/friends and the dissolution of romantic relationships. The most commonly reported sensory modality in field memories was a combination of ‘feeling’ and visual elements (65%) whereas purely visual elements predominated in observer memories (58%). As expected, participants who recalled observer memories also reported experiencing them as if they were viewing a film of the original event to a greater degree than participants who recalled field memories (65.20, SD=33.32 vs. 35.00, SD=32.78; t(128)=5.04, p<0.001). Memories of older events have been identified as being more likely to be recalled in the third-person than in the first-person (Nigro & Neisser, 1983). In the current study, 22% of participants reported memories based on events within the previous 1-week time period while 16% reported memories based on events that occurred 10 years or more prior to the interview. The oldest memory reported was 14 years prior to the interview. Age of memory was coded in weeks since the original event and assessed for differences in temporal frame across vantage perspective. First-person memories were based on events that occurred on average 168.19 weeks prior to the interview while observer memories were based on events that occurred on average 139.85 weeks prior. An independent samples t-test revealed no significant differences between recall vantage perspectives based on age of the recalled event; t(128)=0.83, p>0.05. Recall vantage perspective as a cognitive avoidant mechanism To assess the relationship between vantage perspective and cognitive avoidance strategies, vantage perspective was recoded into a dichotomous variable and independent samples t-tests were conducted. As hypothesized, analysis revealed significant differences between ratings of cognitive avoidance across vantage perspectives. Bonferroni adjusted comparisons revealed that memories recalled in the third-person perspective opposed to the first-person perspective were associated with greater mean ratings of numbness and detachment. Significant differences also emerged for rumination as indexed by the RRS. Results approached significance for the RIQ rumination item “Think about something else”, p=0.02. Results of the independent sample t-tests for the field and observer perspective are reported in Table 2. Contrary to prediction, field memories were not reportedly experienced with greater mean levels of distress compared to observer memories 55.67 (26.91) vs. 56.25 (30.22), respectively; t(128)=0.11, p>0.05. Table 2. Results of independent samples t-tests across field and observer perspectives Field (n=82) Observer (n=48) t(128) Mean (SD) Mean (SD) IES intrusion 15.37 (7.40) 17.85 (8.68) 1.75 IES avoidance 15.04 (8.03) 15.75 (7.34) 0.45 Rumination 12.20 (3.83) 12.08 (3.94) 0.17 RRS 47.13 (11.72) 52.64 (14.02) 2.40** Suppression 10.42 (4.30) 11.47 (5.21) 1.24 WBSI 46.48 (11.63) 46.08 (13.17) 0.18 Dissociation 66.34 (42.76) 91.66 (48.39) 3.10** ‘Detached’ 42.56 (26.47) 55.20 (27.75) 2.58** ‘Numb’ 26.22 (23.78) 36.45 (30.96) 2.48** Note: IES intrusion=impact of events scale, intrusion subscale; IES intrusion=impact of events scale, avoidance subscale; RRS=ruminative response scale; WBSI=white bear suppression inventory. **p<0.01. Table options To explore the relationships between vantage perspective and dysphoria, the sample was divided into high (BDI-II⩾12) and low (BDI-II⩽4) dysphoric participants, and multiple point-biserial correlations were conducted. Results are reported in Table 3. Significant correlations were observed between an observer recall vantage perspective and the measures of cognitive avoidance in the high dysphoric group only. Specifically, memory recall in the observer perspective was associated with dissociative responses such as reports of numbness and feeling detached, and rumination as measured by the RRS. Notably, there were no associations between vantage point and avoidance in the low dysphoric group. Table 3. Point-biserial correlations across vantage perspective in high and low dysphoric participants Low BDI-II (n=32) High BDI-II (n=46) Observer rpb Observer rpb IES intrusion −0.08 0.30⁎ IES avoidance −0.09 0.12 Rumination −0.18 0.03 RRS −0.11 0.35⁎⁎ Suppression −0.01 0.16 WBSI −0.12 0.07 Dissociation −0.07 0.41⁎⁎ ‘Detached’ 0.01 0.35⁎⁎ ‘Numb’ −0.13 0.31⁎⁎ Note: IES intrusion=impact of events scale, intrusion subscale; IES intrusion=impact of events scale, avoidance subscale; RRS=ruminative response scale; WBSI=white bear suppression inventory. ⁎ p<0.05. ⁎⁎ p<0.01.