استفاده از رفتارهای ایمنی برای مدیریت خاطرات مزاحم در افسردگی

Abstract Cognitive models of clinical disorders conceptualise cognitive and behavioural safety-seeking behaviours as central to symptom persistence because they prevent disconfirmation of key maintaining beliefs. Despite growing evidence of the role of negative beliefs about intrusive memories in depression, it remains unclear why such beliefs persist. Accordingly, we examined whether safety behaviours in response to unhelpful beliefs about intrusive memories might play a role in their maintenance. Eighteen high dysphoric (i.e., BDI-II⩾12) individuals who reported an intrusive negative autobiographical memory in the past week completed a battery of measures about their memory, associated negative beliefs and safety behaviours adopted in response to their beliefs. The most commonly endorsed beliefs reflected themes of wanting to control memories (e.g., ‘I should be able to rid my mind of this memory’) and self-deprecation about experiencing them (e.g., ‘Because I can’t control this memory, I am a weak person’). The beliefs prompted a range of safety behaviours, with cognitive distraction being the most common. The findings demonstrate that safety behaviours are common in response to maladaptive beliefs about intrusive memories. Treatment developments in this area are needed, and should incorporate strategies to challenge beliefs about memories, reduce the use of safety behaviours, and promote processing of intrusive memories.

Introduction A growing literature has established that intrusive memories of negative autobiographical events commonly feature in the cognitive profile of major depression (e.g., Brewin, Hunter, Carroll, & Tata (1996a) and Brewin, Hunter, Carroll, & Tata (1996b); Kuyken & Brewin, 1994). For example, as many as 86% (Kuyken & Brewin, 1994) and 87% (Brewin, Hunter, Carroll, & Tata (1996a) and Brewin, Hunter, Carroll, & Tata (1996b)) of depressed samples experience intrusive memories. Not only are such memories experienced in depression, in addition, their presence and intrusiveness is positively associated with cognitive correlates of depression such as poor self-esteem and a negative attribution style (Kuyken & Brewin, 1999). Longitudinal studies have also highlighted the predictive role of intrusive memories in the course of depression (Brewin, Reynolds, & Tata, 1999). As a core symptom of PTSD, the characteristics and management of intrusive memories have primarily been studied in traumatised samples (e.g., Ehlers et al., 2002; Hackmann, Ehlers, Speckens, & Clark, 2004). However, recent studies have borrowed theoretical models and experimental methodologies from the PTSD literature to delineate the characteristics (Patel et al., 2007; Williams & Moulds, 2007a) and cognitive processes (Starr & Moulds, 2006; Williams & Moulds, 2007b) associated with intrusive memories in depression. For example, Williams and Moulds (2007a) found that intrusive memories reported by a dysphoric sample contained high levels of sensory experience and were marked by a sense of ‘nowness’. Sensory features predicted depression over and above intrusion frequency—in accord with findings in PTSD (Michael, Ehlers, Halligan, & Clark, 2005). Similarly, intrusive memories are more likely to be recalled from a third person, detached ‘observer’ perspective in high dysphoric individuals (Williams & Moulds, 2007b), a recall vantage point that is common in PTSD (Kenny & Bryant, 2007; McIsaac & Eich, 2004). In both disorders, compared with a first person ‘field’ perspective, an observer vantage point is associated with avoidant cognitive strategies (McIsaac & Eich, 2004; Williams & Moulds, 2007b). Another parallel is that dysfunctional negative beliefs about both the content (e.g., ‘I am to blame for what happened’) and experience (e.g., ‘Remembering what happened over and over must mean that I’m going mad’) of intrusive memories is linked to intrusion maintenance in both conditions. Ehlers and Steil (1995) postulated that individuals with PTSD who endorse negative meanings of intrusions are more likely to be distressed by their occurrence, and as a result, be motivated to engage in avoidance strategies such as rumination, dissociation and suppression. However, these strategies promote the persistence of PTSD by preventing changes in the meaning of the trauma or the intrusions, and increasing intrusion frequency (Ehlers & Steil, 1995). This model is well supported by retrospective (e.g., Clohessy & Ehlers, 1999) and prospective (Dunmore, Clark, & Ehlers, 2001) studies. Drawing on this work of Ehlers and colleagues, two studies have found parallel associations for intrusive memories in dysphoric samples (Starr & Moulds, 2006; Williams & Moulds, in press). That is, negative beliefs about intrusive memories were significantly correlated with cognitive avoidance and depression severity, and remained so when intrusion frequency and severity of memory content were covaried. Negative beliefs were the strongest predictor of depression, explaining variance over and above intrusion frequency. Together, these findings build an emergent picture of commonalities between intrusive memories in depression and PTSD. Furthermore, they pose a number of clinically important questions; for example, why don’t negative beliefs about intrusive memories remit? A useful concept from the clinical literature that may help to answer this question is that of safety behaviours. Individuals with anxiety disorders engage in a range of cognitive (e.g., thought suppression) and behavioural (e.g., avoidance) strategies that, although intended to ameliorate anxiety, paradoxically maintain it. According to Salkovskis (1989) and Salkovskis (1991), safety behaviours maintain anxiety via two key pathways: (i) preventing disconfirmation of dysfunctional cognitions and beliefs and (ii) increasing the likelihood of occurrence of the feared outcome that the safety behaviours are in fact intended to avoid. For example, in PTSD, the belief ‘I’ll go crazy if I think about the accident’ will likely promote the safety behaviour of suppression of trauma memories. However, suppression prevents learning that one in fact can think about the memory and although feel distressed, not ‘go crazy’. In this way, suppression contributes to PTSD persistence by both preventing disconfirmation of this faulty belief and increasing the frequency of intrusive memories. Safety behaviours are central to cognitive conceptualisations of anxiety disorders (e.g., Ehlers & Clark, 2000), an emphasis that is supported by evidence for their role in disorder maintenance (e.g., Salkovskis, Clark, Hackmann, Wells, & Gelder, 1999; Wells et al., 1995). Harvey (2002) extended this line of investigation beyond anxiety and found that safety behaviours are common in insomnia, and are linked to the persistence of sleep difficulties. The presence of negative beliefs about intrusive memories in depression and accruing evidence that such beliefs play an important role in the persistence of these memories raises the possibility that safety behaviours contribute to their maintenance. Initial evidence of the use of avoidance strategies such as rumination and thought suppression in dysphoric samples (Starr & Moulds, 2006; Williams & Moulds, 2007b) offers preliminary support for this proposal. However, we note that the methodology of these studies involved administration of self-report measures that indexed the specific cognitive variables of interest. Thus, participants were not given the opportunity to provide open-ended responses that may have potentially indicated a range of additional cognitive and behavioural responses to intrusive memories that could function as safety behaviours. Accordingly, our aim was to conduct a broad investigation of the use of safety behaviours in depression. We predicted that high dysphoric individuals would report cognitive and behavioural safety behaviours that would: (i) prevent dysfunctional beliefs about intrusive memories from being challenged and (ii) result in increased intrusive memory frequency.

Results Participants (N=18) ranged in age from 17 to 60 years (M=30.72, SD=13.60). 67% (N=12) were female. BDI-II scores ranged from 12 to 43, with a mean of 24.89 (SD=9.65). Participants had a mean BAI score of 18.11 (SD=10.41), ranging from 3 to 38. In terms of content, intrusive memories were classified according to the categories outlined by Brewin, Hunter, Carroll, & Tata (1996a) and Brewin, Hunter, Carroll, & Tata (1996b) as follows: interpersonal (14/18, 78%), illness/injury (2/18, 11%), and other (e.g., receiving Higher School Certificate results, 2/18, 11%). IES scores indicated significant levels of intrusion and avoidance of intrusive memories: mean Intrusion and Avoidance subscale scores were 18.39 (SD=6.72) and 19.89 (SD=6.70), respectively. Table 1 presents the number of participants who endorsed each belief, the mean belief rating assigned to each item and the number of participants who reported the use of a safety behaviour in response to the belief. Overall, the SBIMQ items were strongly endorsed across the sample, demonstrating that the items were representative of a dysphoric sample's appraisals of intrusive memories. The most commonly endorsed beliefs (i.e., by 80% or more of the sample) reflected themes of wanting to control memories (e.g., ‘I must regain control of this memory’, ‘I should be able to rid my mind of this memory’, ‘I should not be having this memory’) and self-deprecation about experiencing them (e.g., ‘Because I can’t control this memory, I am a weak person’, ‘Having this memory means I have a psychological problem’, ‘Having this memory means I am inadequate’). Not surprisingly, the items with the highest belief ratings (i.e., mean ratings >50/100) overlapped significantly with the most endorsed items: i.e., ‘I should not be having this memory’ (M=65/100), ‘I should be able to rid my mind of this memory’ (M=56.11/100), ‘I must regain control of this memory’ (M=54.44/100) and ‘Having this memory means something is wrong with me’ (M=52.78/100). Table 1. Summary of participant responses on SBIMQ SBIMQ item No. and % who endorsed belief Mean strength of belief (across total sample) (/100) No. and % who endorsed belief and reported a safety behaviour 1. I must regain control of this memory. 16/18 (89%) 54.44 9/16 (56%) 2. I should be able to rid my mind of this memory. 17/18 (94%) 56.11 9/17 (53%) 3. Because I’ve had this intrusive memory, what I’m doing will be ruined. 14/18 (78%) 41.11 10/14 (71%) 4. Because I can’t control this memory, I am a weak person. 17/18 (94%) 42.78 6/17 (35%) 5. Having this intrusive memory means that I could lose control of my mind. 13/18 (72%) 33.33 6/13 (46%) 6. Having this intrusive memory means I’m out of control 13/18 (72%) 32.22 4/13 (31%) 7. If I don’t control this intrusive memory, something is bound to happen. 11/18 (61%) 32.22 0/11 (0%) 8. I should not be having this memory. 16/18 (89%) 65.00 6/16 (38%) 9. If I don’t control this memory, I’ll be punished. 7/18 (39%) 19.44 2/7 (29%) 10. Having this memory means something is wrong with me. 15/18 (83%) 52.78 2/15 (13%) 11. Having this memory means someday I will go out of my mind. 13/18 (72%) 37.22 2/13 (15%) 12. Having this memory means I am inadequate. 16/18 (89%) 45.56 3/16 (19%) 13. Having this memory means I have a psychological problem. 15/18 (83%) 36.67 2/15 (13%) 14. Having this memory means I cannot cope. 15/18 (83%) 43.33 5/15 (33%) 15. Having this memory means I will not achieve future goals that are important. 13/18 (72%) 43.89 0/13 (0%) Table options All but one belief was endorsed by at least 60% of the sample. The least endorsed belief (N=7) was ‘If I don’t control this memory, I’ll be punished’. This item also had the lowest mean belief rating (19.44). This poor endorsement likely reflects its limited relevance to the concerns of depressed samples—although no doubt this III item captures a central concern of OCD samples, the clinical group for which the measure was developed. Similarly, the belief ‘If I don’t control this intrusive memory, something is bound to happen’ also received relatively poor endorsement (N=11); again, plausibly a consequence of its relevance to OCD. Overall, the rates of endorsement and belief ratings accord with previous findings of the prevalence of negative beliefs about intrusive memories in depression ( Starr & Moulds, 2006; Williams & Moulds, in press). Next, we calculated the proportion of participants who described the use of a safety behaviour in response to the experience of a negative belief. Seventy-one percent of participants who endorsed the belief ‘Because I’ve had this intrusive memory, what I’m doing will be ruined’ engaged in a safety behaviour in response. Consistent with the endorsement and strength of belief data, for a significant proportion of the sample (56% and 53%, respectively), the beliefs ‘I must regain control of this memory’ and ‘I should be able to rid my mind of this memory’ prompted the use of a safety behaviour. A range of types of safety behaviours were elicited. Some participants reported more than one safety behaviour in response to their memory (e.g., ‘try to think about something else or do physical activity’). The small numbers prevent meaningful statistical comparison of the prevalence of different safety behaviour types by beliefs. However, we coded safety behaviours into four categories. The categories are as follows: (i) suppression of the memory and/or associated emotion, (ii) cognitive distraction, (iii) behavioural distraction and (iv) use of alcohol/drugs (see Table 2 for exemplar items). Cognitive distraction was the most common type of safety behaviour, with 35 of the responses to intrusive memories so classified. The numbers of safety behaviours in the remaining categories were as follows: suppression=20, behavioural distraction=11 and alcohol/drugs=3. Table 2. Examples of safety behaviours by category Category Examples Suppression ‘Push it out’ ‘Completely delete the memory from mind’ ‘Tell myself to block it’ Cognitive distraction ‘Try to think about something else’ ‘Listen to music and focus on that’ ‘Focus on something else like a positive image’ Behavioural distraction ‘Social distraction’ ‘Keep busy with anything I can to avoid the memory’ ‘Exercise’ Use of alcohol/drugs ‘Have a cigarette’ ‘Wine’ Table options Finally, on the basis that avoidance represents a safety behaviour, we explored the association between participants’ scores on the Avoidance subscale of the IES and the number of safety behaviours reported. A positive but non-significant association was observed (r=.25, p=.31).