Bouton, M. E., Mineka, S., & Barlow, D. H. (2001). A modern learning theory perspective on the etiology of panic disorder. Psychological Review, 108(1), 4-32.     [pages 17-18 of handout]

The first sentence of this paper refers to Pavlov (1927), Watson and Rayner (1920), Mowrer (1947) and Solomon Kamin and Wynne (1953), as people very interested in the relevance of their experimental work on conditioning and learning to human neurosis.

The third sentence notes that enthusiasm for this relationship waned in the 1970s in response to critics such as Rachman (1977).

In the second paragraph they announce that “The major goal of the present article is to spell out the relevance of some contemporary work on classical conditioning, and learning theory more broadly, to understanding the etiology and maintenance of anxiety disorders with an emphasis on one of the more common anxiety disorders: panic disorder (PD).”


Panic disorder

Criteria include: unexpected panic attacks without any obvious cues or triggers, plus anxiety about having another attack or about the implications of the attack (e.g. having a heart attack, losing control, going crazy).

Many people with PD develop agoraphobia, the fear of being in places in which a panic attack might cause major difficulties. Typically agoraphobia involves fears of clusters of situations, most broadly fear of being outside alone, and also commonly fears of being on public transport, or being in crowds, or being in shops. Agoraphobia, according to these authors, is the most severe of all phobias. However not all PD suffers have it, and among those who do, there is a continuum from mild to severe avoidance of feared situations.


Current theories of panic disorder

Cognitive theories

Associated with Clark's (e.g. 1986) notion of “catastrophic misinterpretations” (see page 11 of the handout).

Bouton et al. suggest that problems for the purely cognitive theory include the incidence of nocturnal panic attacks during periods of non-dreaming sleep, panic attacks during the day with no reported catastrophic misinterpretations, and the phenomena of “non-fearful panic”, in which people have physical symptoms alone, without reported subjective anxiety (e.g. Kushner & Beitman, 1990 — see page 14 of the handout)


Anxiety sensitivity theory

E.g. Reiss (1991) and McNally (1996) {neither listed on the handout but see Beck et al, 1999, Forsyth et al., 1999, Liddell and Gosse, 1998; Norton et el. 1999, Schmidt et al. 1997, Van der Does et al., 2000; Watt, Stewart and Cox, 1998}

This theory assumes that some individuals have an enduring trait like tendency to be excessively frightened by the symptoms of anxiety, because they believe that anxiety and its symptoms will have long-term deleterious consequences. This differs from the cognitive theory in that the cognitive theory has people worried about immediate impending disaster, whereas anxiety-sensitivity theory has people worried about long-term side effects.


Conditioning theories

They are on the side of conditioning theories, saying that these have “a long and distinguished tradition of helping to understand the etiology of anxiety disorders”, quoting Eysenck (1979) and Wolpe and Rowan (1988) which are both on the handout, among others.

When stimuli, events, or situations are paired with a panic attack, they become conditioned stimuli capable of triggering more panic and anxiety when they are encountered again.

Bouton et al. say that an important additional idea in conditioning theory is the “fear of fear” process (see Walker, 1984, page 114) and they also mentioning the conditioning of internal physiological effects ("interoceptive conditioning") which relates to the “3-systems model” (see page 13 of the handout).


Some pertinent Clinical Evidence (p.15 of Bouton et al, 2001)

In detailed surveys, it transpires than panic attacks to not come out of the blue, even though this claim is sometimes made in retrospective recall. 80% of panics in one study were classed as predictable, and the majority of patients report that panics arise from a preceding period of milder anxiety. Bouton et al. support the position that “an underlying apprehension is a precursor to panic”, and that expectancies of panic attacks “can be thought of as developing from a conditioned fear of panic attacks”.

Vulnerabilities for the Development of PD (page 18 of Bouton et al, 2001)

Nonspecific Biological (Genetic) Factors

E.g. the heritability of “trait anxiety” or “neuroticism”.

Prior experience with control and mastery, vs prior experience with unpredictability

They use the word “diathesis” to refer to various sources of vulnerability for the development of panic disorder — e.g.

    “Thus, early experience with uncontrollable stress may create a nonspecific diathesis for later life events perceived as unpredictable and/or uncontrollable.” (p.20)
Vicarious and Instrumental Learning

Bouton et al. review studies suggesting that, for instance, adults are more likely to miss work as a result of illness if they have been reinforced (e.g. special toys or food) when they were ill as children. Generally, learning experiences which encourage “sick role behaviour” are correlated with vulnerability to PD, and patients with PD are also more likely than other anxious groups to report that they had observed their parents suffering from panic symptoms.


Lecturer's comments

This is a very long and exhaustive account of some current conditioning theories of one particular anxiety disorder. There are many alternative theories, but this paper is a useful reference for showing that, rather contrary to my own expectations, conditioning theories still count as one of the options for interpreting some of the phenomena relating to human anxiety disorders.