Thought-action fusion: a review

https://doi.org/10.1016/j.jbtep.2004.04.002Get rights and content

Abstract

The cognitive bias of “Thought Action Fusion” (TAF) has received significant research attention in the past decade. The review addresses the assessment of TAF, its place in cognitive theories of obsessional difficulties, and the evidence demonstrating that TAF is relevant to disorders beyond Obsessive Compulsive Disorder (OCD). Data on the components of TAF, its extension to positive outcomes and its role in the aetiology, maintenance and treatment of OCD are reviewed. It is concluded that the moral form of TAF is less robust than the likelihood form and that scales may be best used as a starting point in identifying beliefs and conducting experimental investigations. It is also suggested that the scales be amended to include harm avoidance, which would also increase their clinical utility.

Introduction

In the course of studying the unwanted intrusive thoughts that feature in Obsessive Compulsive Disorder (OCD), it emerged that some patients believe that their unpleasant, unacceptable thoughts can influence events in the world. Two forms of this phenomenon, “Thought-Action Fusion” (TAF) have been identified. “Likelihood TAF” is the belief that having an unwanted, unacceptable intrusive thought increases the likelihood that a specific adverse event will occur. The event can be related to oneself (e.g., “if I think about falling ill, it makes it more likely that I will become ill”). This type of TAF is referred to as “Likelihood-Self”. If the adverse event involves someone else (e.g., “if I think about someone else falling ill, it makes it more likely that they will become ill”), it is described as “Likelihood-Other”. In addition to Likelihood TAF, there is a “Moral TAF”, the belief that having an unacceptable intrusive thought is almost the moral equivalent of carrying out that particular act. For example, “if I think about swearing in Church, this is almost as bad as actually swearing in Church”.

This review addresses a number of questions. First, how has TAF been assessed? Second, where does TAF fit into current cognitive theories of obsessional difficulties? Third, is there a specific relationship between TAF and obsessional complaints or is it a phenomenon that is common to other difficulties such as generalized anxiety disorder and depression? Fourth, what are the components of TAF? Fifth, is TAF confined to negative unwanted thoughts, or does the belief extend to positive, even desired outcomes? Finally, what is the evidence, if any, that TAF is important in the aetiology and maintenance of OCD; and what are the implications for treatment? The review leads to a proposed revision of the construct of TAF based on the empirical work over the past decade.

Section snippets

Origins

Before addressing the questions posed above, it is helpful to review the origins of the construct with TAF. Clinicians have long described patients with obsessional problems who appear to fuse thoughts and actions. Bleuler (1934/1916) used the term “omnipotence of thought” to describe the belief that “patients also fear that they might destroy their beloved ones through a thought” (p. 561). In the modern approach to OCD, it was observed that “It is not uncommon for the person to feel that the

Question 1: how has TAF been assessed?

Broadly, phenomena of this type should be assessable by psychometric methods and also, preferably, by experimental methods. TAF meets both of these criteria. The findings from the studies summarized in Table 1, Table 2 show that TAF has most commonly been measured using a self-report questionnaire. It is not a unitary phenomenon, but consists of two factors in clinical samples—Likelihood TAF and Moral TAF which have accounted for up to 71% of the variance in psychometric studies (Shafran et

Question 2: where does TAF fit within current cognitive models?

Salkovskis (1985) regarded the phenomenon later described as TAF as a specific example of the excessive responsibility for harm which he suggested lies at the heart of obsessional problems. Moral TAF (“Having a thought about an action is like performing the action”; p. 579) was viewed as a “dysfunctional assumption” that interacts with intrusive thoughts and as a form of exaggerated perceived responsibility for harm. Likelihood TAF also contains a strong element of responsibility. Later,

Question 3: what is the relationship between TAF and obsessional complaints?

The first empirical study of TAF was a study of perceived responsibility (Rachman et al., 1995) but subsequent studies have investigated the phenomenon of TAF in its own right. The studies conducted into TAF are summarized in Table 1, Table 2, Table 3.

The majority of the empirical studies find a moderate association between TAF and obsessional complaints which are measured with scales such as the Maudsley Obsessional Compulsive Inventory (MOCI; Hodgson & Rachman, 1977), the Padua Inventory (

Specificity

In the original study, there were significant differences between Likelihood-Other for students and people with obsessional symptoms (see Table 2; Shafran et al., 1996) but not for Moral TAF and Likelihood-Self. This appears to be a reliable finding as it has been replicated with a sample of patients with OCD and student controls (n=285) (Rassin et al., 2001b) and with OCD patients (n=20) and adult non-patient volunteers (n=25) (Abramowitz et al., 2003). These findings are consistent with the

Question 4: what are the components of TAF?

The original study describing the psychometric properties of the TAF Scale (Shafran et al., 1996) was largely replicated by Rassin and his colleagues (Rassin et al., 2001b). In the original study, a three-factor solution dividing TAF into Likelihood-Other and Likelihood-Self was optimal for students but for people with obsessional problems, the two factors of “Likelihood” and “Moral” TAF was optimal. A two factor solution with “Likelihood” and “Moral” components was also supported in the

Question 5: is TAF confined to negative unwanted thoughts, or does the belief extend to positive, even desired outcomes?

In the study by Amir and colleagues, the students with the high levels of obsessional symptoms scored more highly in their estimation of the likelihood of positive events happening as a result of their thoughts about the event. More needs to be known about this and its relationship to the other components of TAF but clinically, it is most unusual for OCD patients to believe that their positive thoughts will increase the likelihood of specific, positive outcomes. As noted above, creating

Question 6: evidence that TAF plays a role in the aetiology and maintenance of OCD and treatment implications

Experimental investigations have been conducted into the possible role that TAF may play in the aetiology and maintenance of OCD. In an ingenious study, 19 high school students were wired up to electrical equipment and told that the equipment would monitor their thoughts (Rassin, Merckelbach, Muris, & Spaan, 1999). They were told that having the thought “apple” would result in a mild electric shock being administered to another person. Participants were also informed that they could prevent

Implications of TAF for treatment

An indirect means of assessing the role of TAF in the maintenance of OCD is to correct TAF and determine the impact on the disorder. This has not yet been done but two studies have investigated TAF and treatment issues. In the first study, Rassin et al. (2001a) found that TAF decreased with successful treatment of OCD, even without specifically addressing it. This indicates that TAF may only need to be addressed in cases which prove to be resistant to standard interventions. In the second study

Summary and conclusions

This review aimed to address six questions regarding the phenomenon of TAF. The first concerned how TAF has been assessed. The data show that TAF has been measured using a self-report instrument which has robust psychometric properties and that TAF can reliably be provoked under experimental conditions. The Likelihood subscales have predictive validity. The second question regarding the place of TAF within current cognitive theories suggests that it may be the case that TAF is one example of

Acknowledgements

RS is funded by a Wellcome Trust Research Career Development Fellowship (63209).

References (46)

  • S Rachman

    A cognitive theory of obsessionsElaborations

    Behaviour Research and Therapy

    (1998)
  • S Rachman et al.

    How to remain neutralAn experimental analysis of neutralization

    Behaviour Research and Therapy

    (1996)
  • S Rachman et al.

    Perceived responsibilityStructure and significance

    Behaviour Research and Therapy

    (1995)
  • E Rassin

    The contribution of thought-action fusion and thought suppression in the development of obsession-like intrusions in normal participants

    Behaviour Research and Therapy

    (2001)
  • E Rassin et al.

    The correlation between thought-action fusion and religiosity in a normal sample

    Behaviour Research and Therapy

    (2003)
  • E Rassin et al.

    Thought-action fusion and thought suppression in obsessive-compulsive disorder

    Behaviour Research and Therapy

    (2001)
  • E Rassin et al.

    The thought-action fusion scaleFurther evidence for its reliability and validity

    Behaviour Research and Therapy

    (2001)
  • E Rassin et al.

    Thought-action fusion as a causal factor in the development of intrusions

    Behaviour Research and Therapy

    (1999)
  • E Rassin et al.

    Relationships between thought-action fusion, thought suppression and obsessive-compulsive symptomsA structural equation modeling approach

    Behaviour Research and Therapy

    (2000)
  • P.M Salkovskis

    Obsessional-compulsive problemsA cognitive-behavioural analysis

    Behaviour Research and Therapy

    (1985)
  • R Shafran et al.

    Thought-action fusion in obsessive compulsive disorder

    Journal of Anxiety Disorders

    (1996)
  • F Tallis

    Obsessions, responsibility, and guiltTwo case reports suggesting a common and specific aetiology

    Behaviour Research and Therapy

    (1994)
  • M van den Hout et al.

    Instructed neutralization, spontaneous neutralization and prevented neutralization after an obsession-like thought

    Journal of Behavior Therapy and Experimental Psychiatry

    (2002)
  • Cited by (139)

    • Towards a computational psychiatry of juvenile obsessive-compulsive disorder

      2020, Neuroscience and Biobehavioral Reviews
      Citation Excerpt :

      This observed dissociation of beliefs and actions docks onto a thought and action fusion seen as a common clinical symptom. Thought and action fusion (TAF) entails the belief that thinking something is equivalent to carrying out that action and that a thought of event is more likely to happen in the future (Shafran and Rachman, 2004; Shafran et al., 1996). The observed action-confidence misappropriation might be a mechanism related to TAF.

    View all citing articles on Scopus
    View full text