نگران نباشید و مراقب خرس سفید باشید: سرکوب افکار در اضطراب

The ability to suppress unwanted thoughts was investigated in patients with Generalized Anxiety Disorder (GAD; n = 29), Speech Phobics (n = 25), and nonanxious controls (n = 28). All participants spent 5 minutes thinking aloud about anything that came to mind while trying not to think of white bears. In another task, they thought aloud for 5 minutes while trying not to think of their main worry. Intrusions of unwanted thoughts were signaled by button presses and recorded on tape. In accordance with the disorder’s definition and complaints of the GAD patients, they showed more intrusions of their main worry than of white bears. The opposite was true for other participants. Compared to a baseline measure, all participant groups were unable to reduce duration of main worry thoughts when trying to suppress them. Worry is a familiar companion to many of us, and constitutes a symptom associated with many anxiety disorders, such as phobias, Panic Disorder, and Generalized Anxiety Disorder Barlow 1988, Rachman 1982 and Wells & Morrison 1994. The phenomenon of worry has aroused considerable interest; therefore, it seems surprising that it has not been subjected to empirical study very often. One reason for this lack of research was suggested by Eysenck (1992), who pointed out that there is no generally accepted definition of the term worry. However, the different definitions proposed Borkovec, Robinson, Pruzinsky, & DePree 1983, MacLeod, Williams, & Bekerian 1991 and Mathews 1990 agree on several characteristics of worries: (a) Worries are a cognitive phenomenon experienced as aversive, people worry about future events and potential catastrophes, and worries are very hard to control. Furthermore, worries differ from similar cognitive phenomena, such as obsessions (see Turner, Beidel, & Stanley, 1992, for a review). For instance, worries are typically related to realistic experiences of everyday life, whereas the contents of obsessions frequently include topics such as contamination, religion, sex, or aggression. (b) Worries are more frequently triggered and generally occur in the form of thoughts, whereas obsessions often occur as images and impulses. (c) Finally, although both are experienced as uncontrollable, worries are not as strongly resisted as obsessions. The last characteristic of worries, their uncontrollability, is particularly important. The anxiety disorder most closely associated with the phenomenon of worrying is Generalized Anxiety Disorder (GAD), and its new definition according to Diagnostic and Statistical Manual of Mental Disorders (DSM-IV; American Psychiatric Association, 1994) stresses that GAD patients’ worrying is uncontrollable as well as excessive. Indeed, self-reports of GAD patients indicate that they feel they have very little control over their worrying (Craske, Rapee, Jackel, & Barlow, 1989), that they worry for a longer time than nonanxious controls (60% of each day compared to 18%; Craske et al., 1989), and they worry about more topics (Vasey & Borkovec, 1992). Regarding the most common topics, GAD patients report more worries about daily hassles and their health than nonanxious controls, whereas worries about family, personal relations, and finances seem to be common in both groups (e.g., Craske et al. 1989 and Shadick & Roemer 1991). So far, these conclusions are based on self-reports only. Therefore, the present study was conducted to provide an experimental test of GAD patients’ complaints about the uncontrollability of their worries. In particular, we wanted to find out whether GAD patients’ mental control is indeed impaired in comparison to control participants, and whether the impairment is limited to their worries or affects unwanted thoughts in general. We compared GAD patients to nonanxious controls as well as to patients suffering from Speech Phobia to determine whether the expected impairment is indeed specific to GAD or related to anxiety disorders in general. In order to design an experimental test of mental control, we adopted a paradigm introduced by Wegner (1989). Wegner’s conception of mental control comprises two processes: concentration on thoughts that one wants to think of and suppression of unwanted thoughts (see Wegner, 1994, for an extensive theoretical discussion). In the case of GAD patients, the second process might be impaired: Their complaint of “uncontrollable worries” indicates that the worries cannot be successfully suppressed. Wegner developed an ingeniously simple method to study the suppression of thoughts (the “white bear paradigm”). He asks his participants to spend 5 minutes thinking (aloud or silently) about anything they want, but NOT about white bears. Whenever the thought or image of a white bear comes to mind against the instruction, participants indicate this by saying so or by pressing a button. As one might expect, it is very hard NOT to think of something. Accordingly, an “enhancement” effect was observed repeatedly. The attempt to suppress thoughts of target objects, such as white bears, caused a considerable number of intrusions (i.e., unwanted thoughts of white bears) during the 5-minute period (Lavey & van den Hout, 1990; Muris, Merckelbach, van den Hout, & Jong, 1992). An enhancement of negative intrusive thoughts caused by suppression during a 4-day period outside the laboratory was reported by Trinder and Salkovskis (1994). More often than enhancement, a “rebound” effect was observed, that is, participants thought about the suppressed target objects more often during a subsequent period of unrestricted thinking or deliberate thinking of the target objects Clark, Ball, & Pape 1991, Clark, Winton, & Thynn 1993, Wegner et al. 1987, Wegner, Schneider, Knutson, & McMahon 1991 and Wenzlaff, Wegner, & Klein 1991. The exact conditions that determine whether enhancement or rebound effects are observed are still unclear (see Kelly & Kahn, 194; Mathews & Milroy 1994 and Roemer & Borkovec 1994), however, the paradigm is obviously well suited for the investigation of mental control and the variables affecting it. Several authors have pointed out the relevance of mental control for clinical disorders such as Obsessive-Compulsive Disorder, depression, and anxiety disorders. Surprisingly though, to our knowledge there have been no published studies that used the white bear paradigm to investigate mental control in clinical disorders.1Mathews and Milroy (1994) investigated the rebound effect following a period of worry, suppression of worry, or nonworrying thought. However, they compared a nonclinical sample of excessive worriers with nonworrying control participants. Other nonclinical participant groups include high-trait worriers (East & Watts, 1994), people attempting smoking cessation (Salkovskis & Reynolds, 1994) and nonclinical participants reporting distressing negative intrusions (Salkovskis & Campbell, 1994). In other studies, mood effects on thought suppression were studied using depressed (Wenzlaff, Wegner, & Roper, 1988) or dysphoric students Conway, Howell, & Giannopoulos 1991 and Howell & Conway 1992 as participants, or inducing happy versus sad moods Howell & Conway 1992 and Reynolds & Salkovskis 1992. These studies found mood-congruent effects (e.g., participants in a sad or dysphoric mood exhibited more negative intrusions and participants in a happy mood more positive intrusions). The white bear paradigm seems especially well-suited for the investigation of mental control in GAD, since it offers an opportunity to put their complaint of uncontrollable worries to an empirical test. Thus, we adopted the paradigm in the experiment reported here. GAD patients, Speech Phobics, and control participants first completed a 5-minute period of unrestricted loud thinking to familiarize them with the think-aloud task. Next, they spent 5 minutes thinking aloud, with the instruction not to think of white bears. Afterwards, they thought aloud for 5 minutes again, this time with the instruction not to think of their main worry. The order of the white bear task and the main worry task was reversed for half of the participants in each group. Intrusions, that is, thoughts of white bears in the bear task and thoughts of the main worry in the worry task, were recorded by having participants press the button of a computer mouse whenever an intrusion occurred. In addition to recording the intrusions by button presses, participants’ verbalizations were recorded on tape. After finishing the think-aloud task, participants rated the anxiety, tension, and excitement they had felt during the experiment as well as the intensity of their wish to leave the situation.