چگونه سرکوب افکار بر باورها در مورد کنترل ناپذیری از نگرانی تاثیر می کند؟

According to Wells’ metacognitive model of generalised anxiety disorder (GAD; [Wells, A. (1995). Meta-cognition and worry: A cognitive model of generalised anxiety disorder. Behavioural and Cognitive Psychotherapy, 23, 301–320]), GAD patients attempt to suppress intrusions that trigger worry. Wells postulates that these attempts are rarely effective and may increase the frequency of worry triggers. These apparent failures are interpreted as evidence for loss of mental control, thereby exacerbating beliefs about worry uncontrollability. The current study tested these predictions. Sixty-two high worriers completed a naturalistic experiment comprising two sessions separated by 1 week. In Session 1, participants recorded their beliefs about worry in general, including its uncontrollability. They then selected a current worry and recorded how often it came to mind over the following week. The Suppression group (N=32) suppressed their chosen worry during the week. The Mention group (N=30) simply monitored its occurrence. In Session 2, Session 1 measures were repeated. Contrary to prediction, the Suppression group reported a significant increase in worry controllability in general. No shift was demonstrated by the Mention group. In addition, relative to the Mention group, the Suppression group reported more success at suppressing their chosen worries, spent less time thinking about them, and found them more controllable and less distressing. Findings are discussed within the context of Wells’ model.

An investigation by Wegner, Schneider, Carter, and White (1987) has had a remarkable impact upon subsequent conceptualisations of psychological problems. Their study comprised two 5-min experimental periods. For the first period, participants were randomly assigned to one of two conditions, suppression and expression. In the suppression condition, participants verbalised their thoughts whilst trying not to think about a white bear. In the expression condition, participants tried to think of a white bear. In the second period, instructions were swapped between groups. Two important findings emerged from this work. First, participants were not able to suppress white bear thoughts entirely. Second, comparison of performance under expression instructions revealed that participants who had already suppressed in the first period recorded significantly more white bear thoughts relative to those who expressed in the first period. This was termed a “rebound effect” since suppression led to a subsequent surge in thoughts. These observations suggested that thought suppression can have paradoxical effects as a self-control strategy (Wegner et al., 1987). Difficulties in interpreting the Wegner et al. (1987) findings are well documented. In particular, employment of a crossover design and use of an expression, rather than a mention, control condition have been questioned (Clark, Ball, & Pape, 1991; Lavy & van den Hout, 1990). Even so, this work stimulated experimental research into thought suppression. Many authors have assessed effects of suppressing neutral thoughts. Some noted rebound effects (e.g. Clark, Winton, & Thynn, 1993), others did not (e.g. Merckelbach, Muris, Van den Hout, & De Jong, 1991). In addition, some found that suppression led to “immediate enhancement” whereby, whilst suppressing, participants experience more target thoughts than controls (e.g. Lavy & van den Hout, 1990). Other studies did not demonstrate this effect (e.g. Clark et al., 1991). Immediate enhancement and rebound effects have clear relevance for understanding psychological disorders characterised by persistent unwanted thoughts. Accordingly, studies have examined suppression of worry-related (e.g. Becker, Rinck, Roth, & Margraf, 1998), anxious (e.g. Koster, Rassin, Crombez, & Naring, 2003), trauma-related (e.g. Shipherd & Beck, 2005), and obsessional thoughts (e.g. Purdon, Rowa, & Antony, 2005). Most studies have been conducted in non-clinical populations, again with rather inconsistent results (for reviews, see Abramowitz, Tolin, & Street, 2001; Purdon (1999) and Purdon (2004)). A recent meta-analysis, incorporating studies of neutral and clinically relevant thoughts, confirmed a small-to-medium rebound effect, although no immediate enhancement (Abramowitz et al., 2001). Overall, these data may, to some extent, justify the inclusion of thought suppression in cognitive conceptualisations of psychological disorders, including obsessive–compulsive disorder (OCD; Salkovskis, 1989), post-traumatic stress disorder (PTSD; Ehlers & Clark, 2000), and generalised anxiety disorder (GAD; Wells, 1995). Wells’ metacognitive model of GAD Metacognition refers to cognitive factors that appraise, control, and monitor thinking. Wells (2005) argues that styles of thinking and real or experienced uncontrollability and intrusiveness of thoughts will only be understood by recourse to metacognitive levels of explanation. This view is embraced in his metacognitive model of GAD (Wells, 1995), which attempts to explain the difficult-to-control, excessive, distressing, and generalised nature of worry that is the central feature of this diagnosis. Wells’ model of GAD distinguishes two types of worry, termed Type 1 and Type 2. Type 1 worry concerns external daily events such as finances, and non-cognitive internal events such as bodily sensations. Type 2 worry is worry about the nature and occurrence of thoughts themselves; in essence, worry about worry. Wells’ model places especial importance upon the role of Type 2 worry in GAD. Upon encountering an intrusive thought, often in the form of a “What if…?” question, Wells argues that a person with GAD activates positive beliefs about using worry as a coping strategy (e.g. worrying helps solve problems). This instigates a Type 1 worry sequence in which various negative outcomes are contemplated and possible ways of coping are conceived. During this sequence, negative beliefs about worry are triggered. These beliefs fall into two broad domains concerning (1) uncontrollability of worry and (2) dangers of worry for mental, physical, or social well-being. Once negative worry beliefs become activated, the person negatively appraises the occurrence of worry; Type 2 worry or worry about worry. When established, Wells argues that additional factors escalate and maintain worry about worry. One factor is thought control. It is the purported role of thought control that provides the focus for the current study. According to Wells, by having conflicting beliefs about worry, individuals with GAD are left in a state of cognitive dissonance. Patients attempt to overcome this conflict by trying not to think thoughts that might trigger worry. In other words, they try to suppress these thoughts. Pointing to findings from the experimental literature (Clark et al., 1991; Wegner et al., 1987), Wells argues that this is problematic because thought suppression attempts are rarely entirely effective and may, in fact, increase frequency of worry triggers. This ineffectiveness is then interpreted as evidence for loss of mental control, thereby strengthening beliefs about worry uncontrollability. Accordingly, worry about worry is perpetuated. Rationale for the current study The present study aimed to investigate the impact of thought suppression attempts upon beliefs about worry. Whilst some studies have investigated how suppression influences worry frequency (e.g. Becker et al., 1998; Mathews & Milroy, 1994), no research has examined the effect of suppression on worry appraisal. Given Wells’ prediction, it is clear how the extant literature can be extended. We were keen to maximise ecological validity. The vast majority of studies examining thought suppression have been conducted within a laboratory setting and over a matter of minutes. In contrast, the present study was completed within the participants’ natural environment and over 1 week. This approach seems more relevant to clinical experience. Finally, in piloting, we endeavoured to be consistent with Wells’ model by asking participants to identify intrusions that triggered their Type 1 worry sequences. However, many found it challenging to make the distinction between intrusive triggers and the worry sequence itself. Therefore, the protocol was amended so that participants were requested (i) to select a current worry, and if assigned to the suppression group, (ii) to try to suppress it as soon as it came to mind. Accordingly, in a more indirect way, participants still suppressed the initial thought relating to their chosen worry (i.e. the intrusion). Another benefit of this approach was that it provided greater consistency in the instructions. The primary hypothesis was that suppression attempts would lead to the strengthening of beliefs about worry uncontrollability. In keeping with Wells’ model, the study also monitored other metabeliefs cited as being important in GAD.

Participant characteristics From the original cohort of 72 high worriers, 62 participants successfully completed the study. Six failed to attend Session 1 and four withdrew between Sessions 1 and 2 (three from the Mention (M) group, one from the Suppression (S) group). The two groups (N: M=30; S=32) did not differ significantly on age [z=−0.09, p>0.9; mean (SD): M=21.7 (3.0); S=22.9 (6.1)], gender [χ2(1)=0.15, p>0.6; females: M=28; S=29], PSWQ [z=−0.35, p>0.7; mean (SD): M=65.7 (5.6); S=66.1 (6.0)], GAD-Q-IV [z=−0.64, p>0.5; mean (SD): M=8.1 (3.0); S=7.5 (3.2)], or BDI-II [z=−0.37, p>0.7; mean (SD): M=14.3 (7.6); S=16.1 (11.3)]. Analysis did not reveal a significant association between group and GAD diagnosis [χ2(1)=1.35, p>0.2; N diagnosed with GAD: M=22; S=19], as determined using a cut-off score of 5.7 on the GAD-Q-IV ( Newman et al., 2002). Appraisals of overall worrying VAS data and repeated measures ANOVA results are presented in Table 1. Given the main hypothesis, the key analysis concerned beliefs about uncontrollability of worry in general. A significant interaction effect between group and session was found. This is depicted in Fig. 1. Investigation of simple effects established that in contrast to the Mention group [F<1], the Suppression group rated their worry as significantly more controllable in Session 2 compared with Session 1 [F(1,31)=6.0, p<0.05]. At both Sessions 1 and 2, ratings of uncontrollability did not differentiate the two groups [Session 1: F<1; Session 2: F(1,90)=3.6, p=0.061