سهم همجوشی فکر-عمل و سرکوب اندیشه در توسعه رسوخ وسواس مانند در افراد نرمال

Both thought–action fusion (TAF: i.e., a cognitive bias implying an inflated sense of responsibility for one's own thoughts) and thought suppression have been claimed to contribute to the development of obsession-like intrusions. Therefore, it seems plausible that conjunction of these phenomena results in highly intense intrusions. However, possible interactions between TAF and thought suppression have not yet been investigated experimentally. In the current study, healthy volunteers were exposed to a TAF-like intrusion. They were, then, randomly assigned to a suppression (n=21) or non-suppression condition (n=19). Next, visual analogue scales (VASs) were completed measuring anxiety, feelings of responsibility and guilt, urge to neutralise and so on. Contrary to expectation, several VAS scores were lower for participants in the suppression group than for those in the non-suppression group. Hence, it is concluded that thought suppression may, at least in the short term, alleviate discomfort caused by TAF-like intrusions.

In their study on intrusive thoughts, Rachman and De Silva (1978) (see also Salkovskis & Harrison, 1984) noted that clinical obsessions as seen in obsessive–compulsive disorder (OCD; American Psychiatric Association, 1994) do not differ in content from everyday intrusions experienced by the majority of the non-clinical population. However, clinical obsessions are experienced as more intense, frequent, discomforting, and resistance eliciting than “normal” intrusions. Two psychological theories have tried to explain the occurrence of these variations in the absence of a difference in content. First, the cognitive theory of obsession implies that the interpretation of intrusive thoughts determines the distressing effects of these intrusions (Rachman, 1993, Rachman, 1997, Rachman, 1998 and Salkovskis, 1985). Furthermore, the interpretation of intrusions is, in turn, affected by cognitive biases. For example, the impulse of wanting to harm someone (i.e., an intrusive thought) will be perceived by most people as normal or somewhat strange, but not very alarming. Imagine, however, a person who believes that wishing to do something is the moral equivalent of doing that same thing. This bias results in a catastrophic misinterpretation of the intrusion (“I want to harm ⋯, and that is just as bad as actually harming that person”). The flawed interpretation of the intrusion will then result in a constellation of symptoms that are reminiscent of those seen in OCD, for example, increased anxiety, discomfort, feelings of responsibility and guilt, and the urge to engage in neutralising behaviours (cf. compulsions). The bias discussed in this example is known as the morality bias (i.e., thinking that thoughts and overt actions are morally equivalent). There are several other biases that may result in a misinterpretation of intrusions (see Emmelkamp & Aardema, 1999). Since most of these biases involve an inflated sense of responsibility for one's own thoughts, they are referred to as responsibility biases. According to Rachman, Thordarson, Shafran, and Woody (1995), two types of responsibility bias are especially relevant to our understanding of obsessional symptoms. The first is the above mentioned morality bias. The second is referred to as the likelihood/probability bias and implies that thinking of a particular event (e.g., a loved one being involved in a car crash) increases the probability that this event will actually happen. Since both biases share the notion that thoughts and actions are spuriously related, the morality and probability biases together are referred to as “thought–action fusion” (TAF; Shafran, Thordarson, & Rachman, 1996). A second theory addressing the transformation of normal intrusions into clinical obsessions focuses on the coping strategy that people rely on in the case of unwanted intrusions. There is good reason to believe that under some conditions, cognitive avoidance strategies such as thought suppression may be counterproductive. For example, in their often-cited studies, Wegner, Schneider, Carter, and White (1987) showed that suppression attempts result in more rather than less intrusions. Interestingly, the increased frequency of intrusive thoughts due to suppression attempts is often found to occur not only during these attempts, but also later on (i.e., when suppression instructions are no longer applicable). The latter phenomenon is known as the rebound effect of thought suppression. The increased frequency of intrusions both during and after suppression is referred to as the “white bear effect”, because Wegner et al. used thoughts of white bears as targets. While the content of this target is, of course, neutral, its heightened frequency is, at least to some extent, reminiscent of obsessional thinking. Furthermore, such a heightened frequency may elicit discomfort. Accordingly, Wegner (1989) claims that the paradoxical effect of thought suppression may result in full blown obsessions: “An obsession can grow from nothing but the desire to suppress a thought” (p. 167). There is some evidence to suggest that both TAF (e.g., Rassin, Merckelbach, Muris, & Spaan, 1999) and thought suppression (see Purdon, 1999) are implicated in obsessive–compulsive symptomatology. Some authors have even speculated that TAF and thought suppression may interact in the development of obsessional problems (see Rassin, Muris, Schmidt, & Merckelbach, 2000). In Rachman's (1998) words: “an inflated increase in the significance attached to an unwanted intrusive thought, such as an obsession, will lead to more vigorous and intense attempts to suppress such thoughts” (p. 393), and: “Given that patients can misinterpret the frequency with which they experience the obsession as evidence for the importance of the obsession (⋯), paradoxical increases in frequency that arise from attempts at suppression, may actually strengthen the catastrophic misinterpretation themselves. A vicious cycle is established” (p. 394). Little or no empirical research has been done on these possible interactions between TAF and thought suppression. From earlier studies, it is known that TAF can be experimentally manipulated and results in an increased frequency of intrusions and discomfort (see Rachman et al., 1996 and Rassin et al., 1999). The present study sought to investigate the possible interactions between TAF and thought suppression in the development of obsession-like intrusions. It was hypothesised that participants who are exposed to a combination of TAF and thought suppression will experience (even) more intrusions and discomfort than participants who are exposed to TAF in the absence of thought suppression instructions. To investigate the effects of TAF and thought suppression, an experimental design derived from Rachman et al. (1996) was used. Participants were given a piece of paper with the sentence (“I hope that ⋯ will soon be in a car accident”) and had to fill in the name of a person close to them. Although this procedure was originally used by Rachman et al. (1996) with participants who had been previously selected because of their TAF-tendency, results from a recent study by Van den Hout (personal communication, December 1999) suggest that the manipulation is strong enough to produce obsession-like stress (i.e., anxiety, feelings of responsibility and guilt, and the urge to neutralise) in non-selected participants. This supports the idea that TAF is not exclusively linked to obsessional thinking, but may be present in a majority of the normal population (see also Koehler, 1991). Shafran (1999) argues: “Although at first reading TAF may seem strange or unusual, consider for a moment whether you would be willing to write the sentence: `I hope ⋯ [loved one] dies in a car accident tomorrow'. Many people — the majority perhaps — would find it distasteful and would consider that such actions might serve to tempt fate. Again, the point is that normal processes are contributing to the maintenance of OCD” (p. 591). In short, it appears that rudiments of TAF are present in most people and can be activated experimentally. After participants completed the sentence, half of them were instructed to suppress any thoughts about the accident, while the other half received non-suppression instructions. Subsequently, participants were asked a number of questions pertaining to anxiety, feelings of responsibility and so on.

Most of the items were answered twice, and were analysed by means of 2 (groups)×2 (times: T1 versus T2) analyses of variance (ANOVARs) with repeated measurement on the last factor. Mean scores on these items are shown in Table 1. Analyses revealed that all scores, but one, dropped significantly from Tl to T2 (all p<0.01 for the main effect of time). The one exception was the estimated likelihood of the imaginary accident happening. For this variable, a main effect of group was present (F[1]=5.2, p=0.03). That is to say that suppression participants estimated the chance of the imagined accident happening in real life smaller than did non-suppression participants. Significant group effects were (on top of the time effect) also present with respect to the time consumed by accident-related intrusions (F[1]=4.8, p=0.03), and perceived moral wrongness (F[1]=9.6, p<0.01). In addition, there were non-significant trends for suppression participants to score lower than non-suppression participants on the variables responsibility (F[1]=3.4, p=0.07), and guilt (F[1]=3.2, p=0.08). Group×time interactions were observed for the effort to suppress accident-related thoughts (F[1]=13.8, p<0.01), and for the time consumed by accident-related intrusions (F[1]=12.9, p<0.01). Table 1. Mean scores (and standard deviations) of participants in the suppression (n=21) and non-suppression group (n=19) on the various items at T1 (i.e., after the first 5 min) and at T2 (i.e., after neutralisation and the second 5 min) a Suppression Non-suppression T1 T2 T1 T2 Effort to suppress 48.0 (25.0) 17.2 (22.6) 27.7 (16.6) 22.1 (17.7) Frequency of intrusions 6.0 (3.6) 3.4 (4.0) 5.6 (2.8) 2.2 (1.4) Time spent thinking about the accident 29.7 (15.3) 26.3 (20.8) 52.5 (22.1) 25.3 (16.2) Anxiety 28.5 (21.3) 14.2 (12.9) 28.7 (19.3) 16.4 (15.1) Moral wrongness 31.5 (21.7) 19.7 (15.6) 53.2 (25.2) 36.7 (23.7) Responsibility 41.0 (28.1) 23.7 (20.8) 57.5 (28.6) 37.1 (30.4) Guilt 45.5 (27.3) 24.6 (20.4) 59.3 (30.4) 41.5 (34.4) Urge to neutralise 31.8 (23.8) 26.6 (23.8) 43.5 (28.1) 27.4 (29.3) Estimated likelihood (%) 2.6 (4.7) 2.3 (3.6) 8.8 (12.3) 8.6 (1.9) a Variables were measured by means of VASs (ranging from 0 to 100), except for frequency of intrusions (which was an open-ended item), and estimated likelihood of the accident (which was expressed as a percentage). Baseline anxiety scores were 18.1 (SD=14.2) and 17.4 (SD=9.9) for suppression and non-suppression participants, respectively. Results from a 2 (groups)×2 (times: baseline versus T1) ANOVAR indicated that anxiety increased (F[1]=15.3, p<0.01), while there was no group or group×time interaction effect (both F<1.0). Table options The item about using a specific cognitive strategy to reduce stress was endorsed by nine suppression participants and ten non-suppression participants at T1 (χ2<1.0), while at T2, the item was endorsed by six suppression and four non-suppression participants (χ2<1.0). Thus, although there were no group differences, there was a time effect, given that at T1, approximately 50% of the participants reported to rely on a stress-relieving strategy (χ2<1.0), while at T2, only 25% of the participants reported to do so (χ2=10.0, p<0.01). Some of the cognitive strategies most often reported by participants were denying that they actually hoped for an accident, rehearsing the idea that they completed the sentence merely because of the experiment and that, therefore, the chance of the accident really happening was small, and thinking that the accident, if any, would not be a severe one. As to neutralisation, 15 suppression participants and 14 non-suppression participants carried out a neutralising activity (χ2<1.0). Neutralisations were scratching out the name or whole sentence (n=13; one participant replaced the name by “Sadam Hussayn”), altering the sentence (n=9), writing a neutralising sentence (n=6), and tearing up the paper (n=1). Mean scores on the VAS addressing the differences between participants' original and current mental representation of the imaginary accident were 37.9 (SD=18.4) and 29.3 (SD=23.2) for suppression and non-suppression participants, respectively (t[38]=1.3, p=0.20), indicating that the imaginary accident had developed into a less severe, paler version, towards the end of the experiment in both the suppression and non-suppression group. Finally, mean scores on the TAF-scale and its subscales were computed. Table 2 presents mean scores of suppression and non-suppression participants, as well as the scores of a sample of 285 undergraduate students (227 women) who completed the TAF-scale on a different occasion (mean age in this sample was 18.9 years, SD=1.5). Interestingly, analyses of variance (ANOVAs) revealed that the scores of participants in the experiment, especially those in the non-suppression group, were higher than those of respondents who had completed the TAF-scale without prior exposure to an experimental procedure.1 The F-values for the total score, morality, and probability-others scales were 10.3, 8.3, and 6.6, respectively, and all reached significance (p<0.01). With respect to the probability-self scale, no group differences were observed (F[2,320]=1.8, p=0.16). Table 2. Mean TAF-scale scores (and standard deviations) for suppression (n=21) and non-suppression participants (n=19) and controls (n=285) a Suppression Non-suppression Control TAF-scale 26.4a (12.3) 30.2a (10.4) 20.3b (10.5) Morality 18.7a (10.5) 20.7a (6.6) 14.11b (8.0) Probability-self 4.4a (2.9) 5.4a (3.0) 4.2a (2.8) Probability-others 3.3ab (3.9) 4.1a (3.5) 2.1b (2.5) a Scores not sharing the same subscript differed in the Tukey honestly significant difference comparison at p<0.05.