اثرات شناختی گسلش و حواس پرتی فکر در ناراحتی عاطفی و باور افکار خود ارجاعی منفی

Abstract Previous research has shown that rapid vocal repetition of a one-word version of negative self-referential thought reduces the stimulus functions (e.g., emotional discomfort and believability) associated with that thought. The present study compares the effects of that defusion strategy with thought distraction and distraction-based experimental control tasks on a negative self-referential thought. Non-clinical undergraduates were randomly assigned to one of three protocols. The cognitive defusion condition reduced the emotional discomfort and believability of negative self-referential thoughts significantly greater than comparison conditions. Favorable results were also found for the defusion technique with participants with elevated depressive symptoms.

Introduction In recent years, acceptance- and mindfulness-based interventions (see Hayes, Follette, & Linehan, 2004) have been widely discussed and studied in the field of cognitive behavior therapy (CBT). As a group, these therapies generally focus more on the functional effects of private events (e.g., thoughts, feelings, physiological sensations, memories) than on content and frequency of these events in understanding and treating psychopathology (e.g., Fisher and Wells, 2005, Linehan, 1993 and Segal et al., 2004). An example of these interventions is Acceptance and Commitment Therapy (ACT) (Hayes, Strosahl, & Wilson, 1999). ACT includes various techniques to change the function of private events in order to promote psychological health (e.g., Hayes, Luoma, Bond, Masuda, & Lillis, 2006). One set of techniques used particularly for this purpose are cognitive defusion strategies (see Luoma & Hayes, 2009). Cognitive defusion is roughly conceptualized as altering the literal meaning and behavior-regulatory function of private events without necessarily altering the form, frequency, or situational sensitivity of these events ( Blackledge, 2007). In ACT, defusion techniques are often employed in contexts where clients are excessively entangled or fused with their private events, such as negative self-referential thought (e.g., “I am depressed”; “I” = “depression”). Control-based strategies, such as avoidance, thought suppression, and distraction are conventional coping methods for unwanted private events (Hayes, Wilson, Gifford, Follette, & Strosahl, 1996). Research has shown that these strategies, which directly target the form and frequency of unwanted private events, may be ineffective and potentially paradoxical (e.g., Eifert and Heffner, 2003 and Gutiérrez et al., 2004). Among these, the iatrogenic effect is especially clear in the case of suppression methods (e.g., Feldner et al., 2003, Levitt et al., 2004 and Marcks, and Woods, 2007). For example, in a study by Marcks and Woods (2005) participants instructed to suppress their personally relevant intrusive thoughts were found to report difficulty in doing so and increased distress after suppression attempt. For this reason, ACT aims at strengthening the process of cognitive defusion for difficult private events, instead of employing control-based coping strategies. Several focused studies have investigated the effects of specific cognitive defusion techniques on negative self-referential thoughts (e.g., Healy et al., 2008, Masuda et al., 2004 and Masuda et al., 2009). One study investigated the effects of a commonly used cognitive defusion technique, “I am having a thought that…” (Healy et al., 2008). The study suggests that negative self-referential statements (e.g., “my life is pointless”) when presented in the defused format (e.g., “I am having a thought that my life is pointless”) can decrease the emotional discomfort associated with that statement and increase willingness to be exposed to these statements. Another study examined the effects of rapid vocal repetition of a one-word version of a negative self-referential thought—another commonly used defusion technique (Masuda et al., 2004). This technique is derived from the observation that when a word is rapidly repeated out loud, the context required for the word to have its literal meaning is altered, and the literal meaning of the word disappears (Titchener, 1910). In the study, the defusion protocol consisted of a defusion rationale, training, and 30-s rapid vocal repetition of a one-word version of the negative self-referential thought. The defusion condition was compared to a distraction-based experimental control task (reading an article on Japan) and thought-control task. Primary dependent variables were reductions in emotional discomfort and believability of the self-referential negative thoughts. Results revealed that the defusion condition decreased the stimulus functions (i.e., emotional discomfort and believability) associated with these thoughts more so than comparison conditions across all participants. A subsequent group parametric study examined whether the duration of one-word thought repetition systematically altered the impact of this defusion strategy ( Masuda et al., 2009). The reduction of emotional discomfort was found to bottom out after 3–10 s of rapid repetition, whereas the maximum reduction of believability occurred after 20–30 s of repetition. These findings additionally suggest that the actual experiential exercise of rapid thought repetition is crucial for altering the stimulus function of negative self-referential thoughts, and that emotional discomfort and believability may be distinctive functional aspects of cognitive events. While interesting and encouraging, previous studies did not clearly reveal the relative effects of rapid thought repetition. Although the original defusion study (Masuda et al., 2004) reported the superiority of the defusion condition to comparison conditions, multiple treatment interference likely occurred because each participant in the study received multiple interventions. Additionally, the thought-control condition in the study was employed for controlling gross demand characteristics. The subsequent parametric study (Masuda et al., 2009) did not compare the defusion protocol with an active comparison condition, either. Additionally, the effects of this cognitive defusion technique have not been examined in clinical or sub-clinical samples. To date, there are several focused experimental studies reporting the positive effects of an acceptance-based emotion regulation strategy in clinical samples with emotional disorders (Campbell-Sills et al., 2006 and Liverant et al., 2008). These results are relevant to the present study in part because the acceptance-based strategy shares aims with defusion strategies (i.e., changes in the stimulus function of aversive private events). Given the lack of empirical evidence, it is worthwhile to investigate the effects of defusion using a subgroup from a non-clinical sample, such as non-clinical college undergraduates who report elevated psychological symptoms. As a response to these emerging questions, the present study investigates the relative impact of the cognitive defusion technique (i.e., rapid thought repetition) on a self-referential negative thought, as compared to a thought distraction strategy. The present study had several methodological and conceptual advantages over previous defusion studies (Masuda et al., 2004 and Masuda et al., 2009). The study employed a group design format in order to reduce extraneous variables, such as multiple treatment interference, and it more clearly identified the active comparison condition. The thought distraction strategy, which is roughly defined as an effort of selectively attending to an emotionally less distressing event or situation (Gross, 1998), was selected as the active comparison condition for several reasons. First, a distraction technique is a defined control strategy (e.g., McCaul & Malott, 1984) that appears theoretically distinct from cognitive defusion because of its primary focus on reducing the frequency of an unwanted private event by shifting attention away from it. Second, distraction can be an appropriate active comparison condition because it has been found effective in some contexts, especially in the situations of mildly and moderately aversive events (e.g., Gutiérrez et al., 2004). Finally, the distraction-based experimental control condition (e.g., reading an article about Japan) was also added to the study as a control group in order to control non-specific factors. Based on previous research findings (e.g., Gutiérrez et al., 2004, Masuda et al., 2004 and Masuda et al., 2009), it was hypothesized that the cognitive defusion condition would reduce the emotional discomfort and believability of negative self-referential thoughts greater than the comparison conditions. It was further predicted that positive effects of cognitive defusion also would be seen among participants who reported “elevated depressive symptoms.”

Results 3.1. Characteristics of sample The mean score for the 132 participants on the BDI-II, which was used to assess participants' general psychological functioning at pre, was 9.09 (SD = 7.64). The score fell in the average range of a non-clinical population. Of the 132 participants who were randomly assigned to the cognitive defusion (n = 41), thought distraction (n = 48), or control condition (n = 43), 15 participants (defusion = 5, thought distraction = 6, & control = 4) were excluded due to extraneous methodological factors. Specifically, 13 of these participants were excluded because of difficulty in identifying a negative self-referential thought or inability to follow instructions. Two participants (one from each active condition) discontinued their participation prior to the completion of assessment because they expressed elevated levels of frustration and became argumentative toward the experimenter. Furthermore, an additional 14 participants (defusion = 5, thought distraction = 4, & distraction-based experimental control = 5) were excluded at the time of data analyses because of the inclusion criteria (i.e., 50 or greater in both emotional discomfort and believability at pre-intervention). Results were equivalent when the data of the 14 participants were included in the following data analyses. As a result, 103 participants with relatively high degrees of emotional discomfort and believability were left for data analyses; cognitive defusion (n = 31), thought distraction (n = 38), and distraction-based experimental control condition (n = 34). Of those, 76% were females (n = 78). The age of the participants ranged from 17 to 60 years (M = 21.17, SD = 7.72). The final sample was ethnically diverse with 38% (n = 39) identifying as “European American,” 30% (n = 31) identifying as “African American,” 17% (n = 17) identifying as “Asian/Pacific Islander,” 11% (n = 11) identifying as “Hispanic American,” and 4% (n = 4) identifying as “other” or “bicultural.” 3.2. Pre-intervention group differences & the effects of potential confounding factors Chi-square testes and ANOVAs revealed that the groups did not differ in the compositions of gender and ethnicity background. BDI-II score, the age of participants, and pre-intervention emotional discomfort and believability (p > .10), did not differ significantly by group. Furthermore, results of two 4 (i.e., experimenters) by 2 (i.e., time) repeated measures ANOVAs revealed no main effect of experimenter or interaction of time and experimenter in the emotional discomfort or believability of negative self-referential thoughts, Fs < 1.17, ps > .32. 3.3. Effects on self-referential negative thoughts The means, standard deviations, and effect sizes of emotional discomfort and believability scores for all conditions are presented in Table 1. The results for the emotional discomfort and believability of the negative self-referential thoughts were analyzed separately, using two 3 (condition: cognitive defusion, thought distraction, & control condition) by 2 (time: pre- and post-intervention) repeated measure ANOVAs. Table 1. Average scores, standard deviations, and effect sizes of emotional discomfort and believability of negative self-referential thoughts by condition and time. Conditions Emotional discomfort Believability Pre Post Pre–post within d Pre Post Pre–post within d 1. Cognitive defusion (n = 31) 77.65 (13.25) 38.81 (24.76) 1.70 73.16 (14.49) 39.10 (27.69) 1.60 2. Thought distraction (n = 38) 78.03 (11.04) 52.17 (24.89) 1.39 80.11 (14.44) 59.11 (26.34) 1.47 3. Control (n = 34) 75.50 (9.89) 63.29 (18.70) 1.36 79.41 (16.33) 70.62 (24.70) 1.60 Between condition Cohen's d Condition 1 vs. condition 2 −.54 −.74 Condition 1 vs. condition 3 −1.13 −1.20 Condition 2 vs. condition 3 −.51 −.45 Table options 3.3.1. Emotional discomfort Results showed a main effect for time, F(1, 100) = 129.16, p < .001, and a two way interaction between condition and time, F(2, 100) = 5.15, p < .01 (see Fig. 1). The interaction was decomposed both by looking at the effects of time across each condition and by looking at the effects of condition at each time point. Emotional discomfort at post-intervention was found to be significantly lower than pre-intervention emotional discomfort across all conditions (Fs > 15, ps < .001). Pairwise comparisons revealed that at post, the defusion group reported significantly lower levels of emotional discomfort than the thought distraction group (p < .05) and distraction-based experimental control group (p < .001). The thought distraction group reported significantly lower emotional discomfort than the experimental control group (p < .05). Means of emotional discomfort and believability of negative self-referential ... Fig. 1. Means of emotional discomfort and believability of negative self-referential thoughts at pre and post by condition. Figure options 3.3.2. Believability Almost identical results were found in the believability of the negative self-referential thoughts with one exception. In the believability of negative self-referential thought, no group difference between thought distraction and experimental control groups was found at post-intervention (p > .05). 3.4. Effects on individuals with elevated depressive symptoms The mean score of BDI-II in 103 participants was 9.9 (SD = 8.18). As a result, using the BDI-cut off score of 10, 42 participants were selected as participants with elevated depressive symptoms; cognitive defusion (n = 14), thought distraction (n = 10), and distraction experimental control condition (n = 18). An ANOVA revealed that BDI-II score did not significantly differ by group (p > .19). 3.4.1. Emotional discomfort The means, standard deviations, and effect sizes of emotional discomfort and believability scores for all conditions among those with elevated depressive symptoms are presented in Table 2. Given a significant group difference of emotional discomfort at pre-intervention, an analysis of covariance (ANCOVA) on post-intervention emotional discomfort, covarying the pre-treatment levels of emotional discomfort was performed. Results revealed the main effect of condition, F(2, 38) = 4.73, p < .05. Pairwise comparisons revealed that the cognitive defusion group had significantly lower emotional discomfort than the experimental control condition (p < .01). Pairwise comparisons did not reveal other significant group differences. Table 2. Average scores, standard deviations, and effect sizes of emotional discomfort and believability of negative self-referential thoughts and Beck Depression Inventory-II by condition and time among participants with elevated depressive symptom. Conditions Emotional discomfort Believability BDI-II Pre Post Pre–post within d Pre Post Pre–post within d 1. Cognitive defusion (n = 14) 83.71 (11.10) 43.36 (28.87) 1.90 77.21 (13.60) 42.21 (27.11) 1.30 20.74 (8.42) 2. Thought distraction (n = 10) 82.90 (10.16) 52.60 (32.85) 1.16 80.80 (17.42) 52.60 (33.18) .61 17.70 (8.51) 3. Control (n = 18) 76.00 (6.88) 64.00 (15.44) .87 81.33 (16.25) 67.94 (24.79) .67 15.38 (4.80) Between condition Cohen's d Cond. 1 vs. cond. 2 −.30 −.65 Cond. 1 vs. cond. 3 −.93 −.99 Cond. 2 vs. cond. 3 −.47 −.22 Note: BDI-II = Beck Depression Inventory-II; Cond. = condition. Table options 3.4.2. Believability A 3 (condition: cognitive defusion, thought distraction, & control condition) by 2 (time: pre- and post-intervention) repeated measure ANOVA revealed the main effect for time, F(1, 100) = 129.16, p < .001. Pairwise comparisons revealed a significant reduction of believability at post-treatment as compared to pre-treatment (p < .001). The time by condition interaction effect was not found to be significant, F(2, 39) = 2.57, p = .089. However, at post-intervention, a medium effect size (d = −.65) was found in the comparison between cognitive defusion and thought distraction conditions, and a large effect size (d = −.99) was found in the comparison between cognitive defusion and experimental control conditions. Both of these results favored the cognitive defusion condition. 3.5. Exploratory analyses on the role of depressive symptoms Because similar patterns were found between the overall participant group and the subgroup with elevated depressive symptoms, the roles of depressive symptom (i.e., BDI-II scores) on negative self-referential thoughts and the effects of intervention were further investigated. A correlational analysis, using the overall sample of 103 participants, revealed that the depressive symptom (BDI scores) was not a predictor of emotional discomfort (r = .06, p = .53) or believability (r = .08, p = .44) at post-intervention. Subsequently, a hierarchical regression analysis was conducted separately for each post-intervention outcome to investigate depressive symptom as a moderator of the relations between intervention and that outcome variable. In the analysis, the variable of interest at pre-intervention was entered into the first step, followed by depressive symptom in the second step. Subsequently, intervention condition (i.e., categorized as 1 = defusion, 2 = thought distraction, & 3 = control condition) was entered in the third step. Finally, depressive symptom and the interaction term were entered in the third step. Results revealed that depressive symptom was not found to be a moderator of the relations between intervention and post-intervention emotional discomfort (β = −.02, t = −.08, p = .93) or between intervention and post-intervention believability (β = −.16, t = −.70, p = .49).