Psychophysiology of Anxiety Disorders
G. Wiedemann and A. Muhlberger
The use of psychophysiological methods in the assessment of a condition and the evaluation of treatment can be separated from research on mechanisms in the aetiology of fear (see also Hugdahl, 1989). In the following section, we will focus on condition assessment and treatment evaluation. For an extended discussion of experimental research on the mechanisms of fear, we refer you to the corresponding section in this book or to the Handbook of Psychophysiology (Bradley, 2000; Öhman, Hamm and Hugdahl, 2000).
The investigation of anxiety in psychophysiological research has a long history and is based on the dualistic view of emotions as consisting of both a cognitive and a somatic component (James, 1884). Obvious changes in different physiological systems, such as pounding of the heart, sweaty palms, and trembling hands upon confrontation with fear-eliciting stimuli, indicate the importance of physiological assessments, especially in anxiety disorders. The physiological reactions may tell us not only about the intensity of fear, but also about the quality of fear responses. In current psychophysiology research, a three-system concept of fear and emotion (Lang, 1968; Foa and Kozak, 1993) is widely accepted. The phobic reaction is seen as consisting of three loosely coupled components: psychophysiological responses, cognitive reports, and avoidance behaviour (Lang, 1971). As the three systems do not always react to the same extent or in the same temporal course (e.g. Rachman, 1977), the assessment of fear responses in all three component levels is recommended (e.g. Hugdahl, 1989). Many of the issues arising regarding one anxiety disorder may be equally applicable to other anxiety disorders. Research on panic disorder, for instance, may be difficult because of diagnostic co-morbidity, particularly with depression. On the other hand, me issue of the nature of the stimuli used to elicit physiological responses is illustrated by disorders such as post-traumatic stress disorder.
This paper is subdivided according to the Diagnostic and Statistical Manual (DSM-IV, American Psychiatric Association, 1994) diagnoses of the different anxiety disorders: (1) phobias; (2) panic disorder (PD); (3) generalized anxiety disorder (GAD); (4) posttraumatic stress disorder (PTSD); and (5) obsessive-compulsive disorder (OCD).
Phobias are subdivided into three subcategories: specific phobia; social phobia and agoraphobia.
No evidence of abnormal reactions of subjects (Ss) with specific phobia was found if no phobic context was present (Klorman, Weissberg and Wiesenfeld, 1977; Lang, Cuthbert and Bradley, 1998b). However, there is clear evidence for differences during exposure to phobic stimuli. The most established paradigm is to compare phobic Ss with control Ss in their reactions to phobic stimuli (e.g. spider pictures) in contrast to neutral, pleasant, or unpleasant stimuli. Early investigations focused on responses of the autonomic nervous system (ANS, e.g. heart rate (HR), skin conductance (SC) and respiration). More recently, focus was drawn to protective reflexes (e.g. startle response (SR), measured by corrugator electromyography (EMG) upon exposure to electric shocks or loud noise) and direct central nervous system (CNS) parameters, e.g. EEG/EP (evoked potentials). Spontaneous EMG measures were most frequently used to measure frontalis or forearm extensor muscle activity. The results were inconsistent (see e.g. review by Sartory and Lader (1981). Hamm et al., (1997) found no significant group-specific differences in corrugator EMG in animal and mutilation phobic Ss during perception of phobic or neutral pictures. This seems to reflect heterogeneity of EMG reactions among subjects. This would be confirmed by subjective evaluations. Phobic Ss quite commonly experience mixed feelings of muscles becoming weak or tense (Hugdahl and Öst, 1985).
Many studies dealing with specific phobias investigated selected fearful, but non-clinical Ss. Since these Ss were comparable to diagnosed phobic Ss in many measures, selected investigations including analogous samples will be included in this section. Fearful samples without a clinical diagnosis are described as fearful Ss whereas diagnosed samples are described as phobic Ss.
According to the DSM IV, phobias can be subdivided in five types: animal type; natural environment type; blood-injection-injury type; situational type and other type. Most research was done on the animal type, the blood/injection/injury type and the situational type (e.g. flight phobia). The following sections will deal with these three types only.
The autonomic system activity profile is dominated by a strong activation of the sympathetic branch. SC level (SCL) is considered a good candidate for measuring sympathetic activation, because the electrodermal system is thought to be controlled exclusively by the sympathetic part of the autonomic nervous system (Bradley, 2000). Nevertheless, there are some drawbacks. Firstly, sweat gland activity relies on a cholinergic mechanism rather than an adrenergic mechanism like most other sympathetic actions. Secondly, it is assumed that there are different types of sweating (e.g. Dawson, Schell and Filion, 2000). It is possible, for example, that sweating of the palms, which is most commonly measured, is an effect of parasympathetic regulation (Guyton and Hall, 1996). Nevertheless, SC reaction (SCR) is reliably modulated by emotional arousal in perception, anticipation and imagination (Bradley, 2000) and