حافظه شرح حال بدنبال رفتاردرمانی شناختی برای سوگ و داغداری پیچیده

Abstract A deficit in the ability to retrieve specific autobiographical memories has been linked to a number of negative consequences, including poor problem solving skills, reduced ability to imagine the future, and the onset of symptoms following trauma. This study investigated the impact of cognitive behaviour therapy (CBT) on memory retrieval specificity in patients with Complicated Grief (CG). Twenty individuals with CG who were seeking treatment were administered an autobiographical memory task (AMT) before and after completing a 10-week CBT program. Pre-treatment retrieval specificity did not predict treatment outcome. However, there was a significant correlation between symptom reduction and increased specific retrieval to positive cues following treatment. These results suggest that overgeneral retrieval in CG can be modified by CBT, and may point to one means by which CBT can alleviate CG symptoms.

1. Introduction Complicated Grief (CG; alternatively know as Prolonged Grief Disorder) is a chronic and disabling bereavement related condition that impacts between 10% and 15% of bereaved individuals (Prigerson, Vanderwerker, & Maciejewski, 2008). The essential feature of CG is an intense and prolonged yearning for the deceased continuing 6 months beyond the loss. Other symptoms include difficulty accepting the loss, emotional numbing, bitterness, a loss of meaning and sense of purpose, a difficulty re-engaging in life, confusion about one’s role in life or a diminished sense of self, and avoidance of reminders (Prigerson et al., 2009 and Prigerson et al., 2008). Although related to depression and PTSD, the condition is associated with unique negative consequences for the individual (Boelen and Prigerson, 2007 and Prigerson et al., 2008). There is increasing evidence that the treatment of choice for CG is cognitive behaviour therapy (CBT). Three trials have now indicated that CBT that incorporates psycho-education, exposure, and cognitive restructuring is an efficacious intervention for CG (Boelen et al., 2007, Shear et al., 2005 and Wagner et al., 2006). Despite the promise shown by CBT, a significant proportion of patients do not benefit from this therapy. Accordingly, there is an important need to understand the factors associated with alleviation of CG with CBT. People with CG display impaired retrieval of specific memories (Golden et al., 2007 and Maccallum and Bryant, 2010). This finding is consistent with a sizable body of evidence that individuals with depression and post-traumatic stress disorder (PTSD) have difficulty recalling specific autobiographical memories (for review see Moore and Zoellner, 2007 and Williams et al., 2007). The finding that CG is associated with overgeneral retrieval is significant because this form of retrieval has been linked to impaired social problem solving (Evans et al., 1992 and Goddard et al., 1996), difficulties imagining the future in a specific way (Williams et al., 1996) and poorer response to treatment (Brittlebank, Scott, Williams, & Ferrier, 1993). Findings that memory specificity does not typically improve when depression remits (Brittlebank et al., 1993, Peeters et al., 2002 and Raes et al., 2006), and that overgeneral memory may represent a risk factor for developing symptomatology following stressful experiences (Bryant et al., 2007 and Mackinger et al., 2000) have lead to the proposition that overgeneral retrieval may represent a trait that increases one’s vulnerability to emotional disturbance (Williams, 1996). However, there is also growing evidence showing that retrieval specificity is modifiable (e.g., Barnard et al., 2006, Raes et al., 2009, Watkins and Teasdale, 2001, Watkins et al., 2000 and Williams et al., 2006). A number of studies have investigated whether overgeneral memory is amenable to change through psychological therapy. Williams, Teasdale, Segal, and Soulsby (2000) found that mindfulness-based cognitive therapy significantly reduced categorical retrieval in a sample of participants with remitted depression. Serrano, Latorre, Gatz, and Montanes (2004) utilized a therapy which encouraged the retrieval of specific positive memories across life stages in an elderly depressed sample. They reported greater symptom resolution and increased specificity to positive cues in the experimental group (see also McBride, Atkinson, Quilty, & Bagby, 2006). Sutherland and Bryant (2007) found that improvement in PTSD symptoms following cognitive behaviour therapy was significantly associated with improved retrieval of specific memories in response to positive cues. Raes et al. (2009) found that depressed patients instructed in recalling specific memories showed concurrent decreases in rumination and improved outcomes. There is overwhelming evidence that overgeneral retrieval is associated with a range of detrimental factors, including rumination, impaired problem solving, and deficits in imagining a future. These potential problems might be particularly relevant for bereaved individuals because of their need to retrieve positive memories and manage future challenges without the deceased. Accordingly, it is useful to know if successful treatment of CG is linked to more specific retrieval of personal memories. The current study investigated the relationship between overgeneral retrieval and treatment outcome in CG following CBT. It was predicted that symptom improvement would be associated with increased retrieval specificity.

. Results 3.1. Participant characteristics Table 1 presents participant characteristics for age, relationship to the deceased, years since death, and pre- and post-treatment CGA, BDI-II and CAPS-2 scores. As expected, symptoms of CG decreased significantly from pre (M = 32.45; SD = 4.17) to post-treatment (M = 21.85; SD = 8.22), t(19) = 5.67, p < 0.001. At the post-treatment assessment, 11 of the 20 participants no longer met diagnostic criteria for CG. There were also significant decreases in scores on the BDI-II from pre- (M = 32.10; SD = 11.05) to post-treatment (M = 22.76; SD = 11.72), t(16) = 2.39, p < 0.03, and the CAPS-2 (M = 51.80; SD = 19.72) to post-treatment (M = 30.65; SD = 20.57), t(19) = 5.95, p < 0.001. Table 1. Participant characteristics at pre- and post-treatment assessments. Pre N = 20 Post-treatment N = 20 Age 52.30 (11.69) Relationship of the deceased Partner 30% Child 30% Parent 40% Years since death 3.36 (2.48) CGA Mean 32.45 (4.17) 21.85 (8.22) Range 25–40 9–35 BDI Mean 32.1 (11.05) 22.76 (11.72)a Range 17–52 1–44 CAPS Mean 51.80 (19.72) 30.65 (20.57) Range 26–95 0–87 LNS 10.10 (1.44) 10.75 (1.91) NART 34.85 (7.20) Note: Standard deviations appear in parentheses. a BDI-II scores were unavailable for 3 participants at post-treatment. Table options 3.2. Memory patterns over time Table 2 presents the mean number of specific, categoric and extended memories at pre- (Time 1) and post-treatment (Time 2). A 2 (Time) × 2 (Cue Valence) ANOVA on specific memories indicated a main effect for Time (F(1, 19) = 11.91, p < 0.004); Overall, participants recalled more specific memories at post-treatment than pre-treatment assessment. Separate 2 (Time) × 2 (Cue Valence) ANOVAs of categoric and extended memories indicated no significant overall effects. A 3 (Memory Content) × 2 (Cue Valence) × 2 (Time) repeated measures ANOVA revealed a main effect for Memory Content, (F(2,38) = 30. 70 p < 0.001) and a significant 2 way interaction for Cue Valence × Content, (F(2,38) = 3.75., p < 0.04) (see Table 3). Overall, participants recalled more Other than Loss or Person memories; however this pattern varied for positive and negative cues. Whereas the difference between Loss and Other memories was not significant for negative cues, it was significant for positive cues [t(19) = −8.51 p < 0.001]. That is, in response to negative cues participants tended to recall a greater proportion of Loss related memories than in response to positive cues. There were no significant differences in memory content from pre- to post-treatment. Table 2. Mean memory specificity before and after treatment. Time 1 Time 2 Specific Positive 2.65 (1.34) 3.25 (1.33) Negative 2.75 (1.33) 3.90 (1.17) Categoric Positive 1.45 (1.09) 1.05 (1.05) Negative 1.45 (1.15) 0.85 (1.81) Extended Positive 0.40 (0.50) 0.65 (0.81) Negative 0.30 (0.57) 0.20 (0.41) Note: Standard deviations appear in parentheses. Time 1 = Pre-treatment assessment, Time 2 = Post-treatment assessment. Table options Table 3. Mean proportion of loss, person and other memories before and after treatment. Time 1 Time 2 Loss Positive 0.10 (0.14) 0.12 (0.16) Negative 0.39 (0.29) 0.29 (0.25) Person Positive 0.14 (0.21) 0.19 (0.27) Negative 0.04 (0.08) 0.05 (0.12) Other Positive 0.75 (0.25) 0.69 (0.31) Negative 0.56 (0.30) 0.79 (1.02) Note: Standard deviations appear in parentheses. Time 1 = Pre-treatment assessment, Time 2 = Post-treatment assessment. Table options 3.3. Memory specificity and treatment outcome Hierarchical multiple regression analyses were conducted to examine the degree to which treatment outcome (post-treatment CGA score) was predicted by pre-treatment specific or categoric memory retrieval. Due to the small sample size, only a limited number of predictors were included in these analyses. On Step 1, pre-treatment CG scores were entered and on Step 2, pre-treatment specificity (or categoric) scores were entered. All equations were non-significant (Specificity: R2 = 0.06, adjusted R2 = −0.05, p < 0.59, Categoric R2 = 0.07, adjusted R2 = −0.03, p < 0.51). Neither Time 1 severity of CG symptoms nor memory specificity (or categoricness) were significant predictors of outcome in this sample. 3.4. Memory specificity and symptom change To examine the relationship between change in memory specificity and change in symptom levels standardized residual gain scores were calculated. Residual gain scores are a well validated measure of change (Mintz et al., 1979 and Steketee and Chambless, 1992). These scores are the standardized residuals of the regression of pre-treatment measures on post-treatment measures, and thus represent the change in symptoms (memories) that is not expected on the basis of pre-treatment scores alone. Overall, there was a significant negative correlation between change in CGA and memory specificity to positive cues (r = −0.61, p < 0.005). Reductions in CG symptoms were correlated with increases in specificity to positive cues. The correlation between change in CG and negative cues was non-significant (r = 0.03). 2 To examine the extent to which this relationship was impacted by concurrent changes in depression, post-traumatic stress, and working memory (as indexed by the LNS) residual gain scores were calculated for each of these variables. Partial correlations between CGA and positive and negative cue specificity residual gain scores controlling for change in these variables were calculated (adjusted alpha p < .012). The correlation between change in positive memory specificity and change in CG was marginally reduced (r = −0.57, p < .05). Partial Correlation coefficients for specific negative retrieval (r = −0.01) remained non-significant.