Next | Previous | Table of contents
1. Types of Interventions
Clinicians may treat anxiety disorders from a range of perspectives (e.g., insight-oriented therapy and hypnosis). However, the treatments with well-recognized empirical support include two main approaches: (1) pharmacotherapy (drug therapy) and (2) cognitive-behavioural therapy (CBT), a form of psychotherapy. Although a few studies have compared these approaches to other treatments (e.g., analytic psychotherapy), these alternative approaches have generally not been especially effective compared to CBT and medication.
Pharmacological approaches, including antidepressants (e.g., monoamine oxidase inhibitors) and antianxiety medications (e.g., benzodiazepines), have been shown to be helpful in treating each of the anxiety disorders, except specific phobias. Cognitive- behavioural therapy (CBT) appears to be an effective psychological treatment for each anxiety disorder, although few properly controlled studies have been conducted to evaluate its effectiveness with PTSD and appropriately diagnosed specific phobias. CBT strategies shown to be helpful include cognitive restructuring (i.e., changing anxious thoughts, interpretations, and predictions into more rational and less anxious thoughts), exposure to feared objects and situations, and relaxation training.
Specific techniques have been developed and tested for particular anxiety disorders. For individuals with panic disorder, systematic exposure to feared sensations using exercises such as hyperventilation and spinning (i.e., interoceptive exposure) as well as breathing retraining (i.e., learning to breathe in a slow and relaxed manner) appear to be helpful. For specific phobias involving blood and injections, applied tension has been shown to be a helpful method of increasing blood pressure in order to prevent the fainting that is often associated with such phobias.
A variety of newer treatments have begun to be tested in uncontrolled research studies. Among medications, the SSRI (selective serotonin reuptake inhibitor) antidepressants (e.g., paroxetine) have been the subject of much research activity. New cognitive and behavioural strategies are currently being tested as well. These include a technique called eye movement desensitization and reprocessing, as well as computer-administered CBT. It remains to be shown whether these newer treatments will be as effective or even more effective than established treatments.
2. Pharmacological vs. Psychological Treatments
Despite a large literature supporting the use of medications and CBT, there remains debate among clinicians and researchers regarding the relative efficacy of both approaches, as well as the relative importance of biological and psychological processes in the etiology of anxiety disorders.
As reviewed by Antony, Brown, and Barlow (1992), biological theorists tend to minimize the importance of psychological variables, citing research demonstrating differences between those with and without anxiety disorders on a variety of biological variables (e.g., brain imaging, hormone levels, biological challenges, neurochemical levels, and genetics). Cognitive and behavioural theorists explain these findings as biological manifestations of a primarily cognitive or behavioural phenomenon, and often cite studies demonstrating that these biological findings can be influenced by psychological variables. In addition, cognitive and behavioural theorists point to data showing fear-relevant biases of attention and memory and anxious beliefs regarding fear-relevant objects and situations among individuals with anxiety disorders. For biological theorists, these findings tend to be explained as cognitive manifestations of a primarily biological process.
Antony et al. (1992) state that biological and psychological findings can generally be explained from either perspective and there is little reason to accept only one of these views. A more parsimonious approach is one that integrates biological and psychological findings. Recent models of anxiety disorders (e.g., Antony and Barlow, 1996; Barlow, 1988) have attempted to account for the ways in which both biological and psychological factors influence the development of anxiety disorders. Nevertheless, investigators from biological and psychological perspectives rarely collaborate on treatment studies and rarely read one another's work, except to criticize it. Biological and psychological treatment studies tend to be conducted at different sites, use different measures, and get published in different journals. Over time, more investigators and clinicians have been willing to consider multidimensional approaches to understanding and treating anxiety disorders, but there is still much work to be done to educate practitioners and researchers about the nature of anxiety and its disorders.
3. Treatment Considerations
a. Possible Side Effects of Medications
Concern has been expressed over possible side effects of some of the medications used to treat anxiety disorders, particularly the benzodiazepines. Common side effects associated with these medications, which may decrease over the course of treatment, include sedation, fatigue, ataxia, slurred speech, and amnesia (Noyes et al., 1988).
The side effects associated with discontinuation of benzodiazepines have also been explored. Pecknold, Swinson, Kuch, and Lewis, (1988) reported that 35% of patients treated with alprazolam experienced a withdrawal syndrome during discontinuation of the drug. Withdrawal symptoms included confusion, disorientation in time, place or person, heightened sensory perception, dysosmia (abnormality in taste or smell), paresthesias (sensation of numbing or tingling), muscle twitch, blurred vision, diarrhea, decreased appetite, weight loss, and muscle cramps. These were not incapacitating or life-threatening in any of the cases.
Discontinuation of benzodiazepines may also be associated with relapse and recurrence of symptoms. In a 1991 study, Noyes, Garvey, Cook, and Suelzer found that between 63% to 84% of patients who were taking alprazolam or diazepam for panic disorder experienced a relapse. Variations in the rate of relapse were related to use of different outcome criteria. Because of the difficulty that most patients with panic disorder have discontinuing treatment with benzodiazepines, attention has turned to the development of specific programs to help patients stop taking anxiolytics (Klein, Colin, Stolk, and Lenox, 1994).
Research suggests that the beneficial effects of some forms of medications may be dependent on continuing use of the drug. For example, although clomipramine is effective in treating obsessive-compulsive disorder, one study (Thorén, Åsberg, Cronholm, Jörnestedt, and Träskman, 1980) found that the therapeutic effects of five weeks of treatment with clomipramine were not maintained after discontinuation. Further research in this area is warranted.
b. Appropriateness of Treatment
There is some evidence that many individuals with anxiety disorders are not offered appropriate treatments. Health care professionals outside of specialized anxiety disorders clinics are less likely to use empirically validated treatments. For example, Swinson et al. (1992) found that although the majority of patients with panic disorder and social phobia had tried psychotropic medications (89% and 75%, respectively), most had never received the treatments with the most empirical support. Among patients with PD, only 15% had received imipramine, 13% had received alprazolam, and 11% had received cognitive- behavioural therapy. Only 4% of patients with social phobia had received monoamine oxidase inhibitors and 4% had received cognitive-behavioural therapy. Swinson et al. (1992) suggest that education regarding early recognition and intervention is needed for general and emergency department physicians. Educational activities should also be targetted to other health and mental health care professionals, including psychologists, psychiatrists, occupational therapists, social workers, psychiatric nurses, et cetera. Seminars on the different types of anxiety disorders and corresponding treatments, and standardized methods of assessment could be effective methods for disseminating information within the health and mental health fields.
A survey of psychiatrists and psychiatric residents (McCarley, Steinberg, Spears, and Essock-Vitale, 1987) showed that medical professionals are more likely to favour biological approaches over behavioural approaches in their training and practice. In addition, psychoanalytic and psychodynamic approaches were favoured over CBT despite the lack of empirical support for these treatments. A recent survey of family practice physicians (Hecker, Fink, and Fritzler, 1993) was somewhat more optimistic in its findings. Participants tended to rate CBT as a more acceptable treatment for PD than client-centered therapy, with pharmacological approaches falling somewhere between the two psychological approaches in terms of acceptability.
Results were more promising in specialized anxiety disorders clinics. Swinson, Cox, Kerr, Kuch, and Fergus (1992) sought to investigate the availability and appropriateness of services for anxiety disorders in Canada. They sent a questionnaire to 240 hospitals across the country, of which 117 responded. Of the responding hospitals, only 18 had an anxiety disorders clinic. Clinics saw an average of 208 patients per year. The most common diagnoses seen were PDA (25.4%), GAD (22.1%), PD (15.2%), social phobia (13.5%), agoraphobia without panic (9.9%), OCD (8.9%), and PTSD (8.8%). Despite the high prevalence of specific phobias in the general population, these individuals tended not to present to anxiety disorders clinics. Most of the clinics reported having an orientation consistent with CBT, pharmacotherapy, or combinations of these approaches.
4. Research Findings: Effective Treatments for the Anxiety Disorders
In reviewing the findings below, it should be kept in mind that symptom reduction is increasingly being viewed as only one of the factors to be considered when comparing the effectiveness of various treatment approaches. Other aspects to be considered include the acceptability of the treatment to the patient and family members, the impact on the patient's quality of life, including functional impairment, the time needed to show results, the cost of treatment, and the ease with which the treatment can be taught to professionals and made available to the public.
a. Panic Disorder with (PDA) and without Agoraphobia (PD)
Among the anxiety disorders, PD and PDA are the most extensively researched. Numerous outcome studies have demonstrated the effectiveness of pharmacological, psychological, and combined treatments. In addition, compared to the other anxiety disorders, studies of PD and PDA have typically been more sophisticated in their designs and measures, including assessments of panic frequency, generalized anxiety, depression, agoraphobic avoidance and other domains of functioning.
In pharmacological trials, a variety of medications have been demonstrated to be more effective than placebo for treating PD and PDA. These include a range of antidepressants as well as antianxiety medications (e.g., benzodiazepines). Although alprazolam (e.g., Ballenger et al., 1988) and imipramine (e.g., Mavissakalian and Perel, 1995) have been the most frequently researched medications, other drugs that have been shown to be effective include clonazepam (Beauclair, Fontaine, Annable, Holobow, and Chouinard, 1994), adinazolam (e.g., Carter et al., 1995), clomipramine (e.g., Johnston, Troyer, and Whitsett, 1988), lorazepam (e.g., Schweizer et al., 1990), fluvoxamine (de Beurs, van Balkom, Lange, Koele, and van Dyck, 1995), and several other medications. Furthermore, there appear to be few differences in effectiveness among these medications (e.g., Cross National Collaborative Panic Study, Second Phase Investigators, 1992). However, benzodiazepines (e.g., alprazolam) tend to be associated with earlier improvements (Cross National Collaborative Panic Study, Second Phase Investigators, 1992) and greater rates of relapse following discontinuation (Rickels, Schweizer, Weiss, and Zavodnick, 1993), relative to antidepressants such as imipramine. Interestingly, CBT has been shown to help patients discontinue their medications without relapse (Otto et al., 1993).
Predictors of outcome for pharmacological treatment for PD and PDA have been identified by Basoglu, Marks, Kiliç, Brewin, and Swinson, (1994). In their study, patients who attributed their improvement to medication (regardless of whether they were taking alprazolam or placebo) were more likely to experience withdrawal symptoms and loss of gains than were patients who attributed their improvements to their own efforts. Pre-treatment predictors of poor outcome at post-treatment included first time psychotropic medication use, more severe agoraphobia, and longer duration of illness. Predictors of poorer long-term outcome included more severe agoraphobic symptoms, older age, past history of depression, and longer duration of illness.
Numerous studies have demonstrated the effectiveness of CBT relative to no treatment and to alternative psychological treatments such as supportive psychotherapy (See Clum, Clum, and Surls, 1993 for a review). Although there appears to be few consistent differences across studies in the effectiveness of various cognitive-behavioural strategies (e.g., cognitive restructuring, exposure, applied relaxation), some studies have shown that relaxation is somewhat less effective than other techniques (e.g., Marks et al., 1993).
Studies have also explored the role of family support in cognitive-behavioural treatment of PDA. One study reported that patients whose spouses were included in treatment sessions showed significantly more improvement on measures of agoraphobia than patients treated without spouses (Barlow, O'Brien, and Last, 1984). A two-year follow-up study (Cerny, Barlow, Craske, and Himadi, 1987) revealed that patients treated with their spouses continued to make gains, whereas those who were treated alone did not show continued improvement, especially in the first year. The spouse-treated group also showed less functional impairment during the first year.
Several studies have shown CBT to be more effective than pharmacological approaches for PD and PDA, particularly over the long-term (e.g., Clark et al., 1994; Marks et al., 1993). Other studies (e.g., de Beurs et al., 1995) have found few differences between drug treatments and psychological interventions. With few exceptions (e.g., Telch, Agras, Taylor, Roth, and Gellen, 1985), combined medication and CBT appear to be no more effective than either alone, although there is some evidence that patients who believe that their improvements are due to the medications are at a greater risk for relapse than individuals who attribute their improvement to psychological interventions (Basoglu et al., 1994).
A variety of studies have shown that patients with PD and PDA can benefit from self-help treatments and interventions with reduced therapist contact, such as self-help books (Gould, Clum, and Shapiro, 1993) and treatment by telephone (Swinson, Fergus, Cox, and Wickwire, 1995).
b. Obsessive-Compulsive Disorder (OCD)
Overall, there is evidence that OCD can be effectively treated with a variety of medications, especially serotonin specific agonists such as clomipramine (Clomipramine Collaborative Study Group, 1991), sertraline (Chouinard et al., 1990), fluvoxamine (Perse, Greist, Jefferson, Rosenfeld, and Dar, 1987), and fluoxetine (Tollefson et al., 1994).
Controlled studies of CBT have consistently shown that behavioural techniques are effective for OCD (e.g., Fals-Stewart, Marks, and Shafer, 1993; Foa and Goldstein, 1978). Strategies typically used in CBT to decrease anxiety include exposure to feared situations and thoughts, as well as prevention of rituals. In addition, some studies have incorporated cognitive restructuring into the treatment of OCD.
In comparative outcome studies, no consistent differences in effectiveness have been found for CBT, medications, and their combination (e.g., Cottraux et al., 1990; Marks et al., 1988). This conclusion has been confirmed in meta-analytic reviews of treatments for OCD (e.g., Cox, Swinson, Morrison, and Lee, 1993; van Balkom, van Oppen, Vermeulen, and van Dyck, 1994).
c. Social Phobia
Although few controlled pharmacological trials exist for social phobia, preliminary evidence suggests that a variety of drugs may be effective, including clonazepam (Davidson et al., 1993), phenelzine (Liebowitz et al., 1992), sertraline (Katzelnick et al., 1995), and brofaromine (Fahlen, Nilsson, Borg, Humble, and Pauli, 1995).
In addition, there are numerous studies supporting the use of CBT for individuals with social phobia. Strategies that have been found to be effective for social phobia include exposure to feared situations, role playing, cognitive therapy, and social skills training. Overall, CBT appears to be more effective than supportive psychotherapy (e.g., Heimberg, Dodge, Kennedy, and Zollo, 1990), and no treatment at all (Newman, Hofmann, Trabert, Roth, and Taylor (1994).
Studies comparing various cognitive and behavioural approaches have led to inconsistent findings. A recent meta-analysis comparing cognitive-behavioral packages and pure exposure-based treatments for social phobia (Feske and Chambless, 1995) found no differences overall. In other words, adding cognitive strategies to exposure-based therapies did not seem to affect outcome. Length of treatment was similarly unrelated to outcome, although a larger number of exposure sessions was associated with better results.
Preliminary evidence suggests that CBT is at least as effective as pharmacological approaches and probably more effective than medications in the long-term (Gelernter et al., 1991; Heimberg et al., 1994). Much more research is required before such a conclusion can be drawn with confidence. In addition, there are no published studies examining the efficacy of combined psychological and pharmacological treatments for social phobia.
d. Generalized Anxiety Disorder (GAD)
There appears to be effective pharmacological and psychological treatments for GAD. Although most medication studies are based on old criteria for GAD (which have since been substantially revised), there is evidence that a range of pharmacological interventions may be helpful for generalized anxiety, including buspirone, imipramine, and a variety of benzodiazepines such as diazepam, alprazolam, and lorazepam, among others. With a few exceptions, most studies have found these medications to be more effective than placebo, but few differences have been reported among medications (e.g., Hoehn-Saric, McLeod, and Zimmerli, 1988; Rickels et al., 1982; Rickels et al., 1993; Sacchetti, Zerbini, Banfi, and Tansella, 1994). One difference among medications that has been shown in some studies is the finding that benzodiazepines often work more quickly than other medications (e.g., Rickels et al, 1993).
In general, studies using CBT to treat individuals with GAD have been based on a larger number of measures and more recent diagnostic criteria, compared to medication studies. A variety of approaches appear to be helpful for patients with GAD, including cognitive strategies (e.g., cognitive restructuring) and behavioural strategies (e.g., relaxation training). CBT appears to be more effective than alternative treatments including diazepam, analytic psychotherapy, non-directive psychotherapy and placebo (e.g., Durham et al., 1994; Power, Simpson, Swanson, and Wallace, 1990). Although some studies have shown differences in the effectiveness of different cognitive and behavioural treatments, most studies have found few differences in the effectiveness of these approaches (e.g., Barlow, Rapee, and Brown, 1992; Borkovec et al., 1987; Borkovec and Costello, 1993).
e. Specific Phobia
No controlled studies have been conducted with medications for individuals meeting diagnostic criteria for specific phobias.
Numerous studies have demonstrated that behaviour therapy is an effective method of decreasing fears of a variety of objects and situations including heights (Bourque and Ladouceur, 1980), dental treatment (Jerremalm, Jansson, and Öst, 1986), enclosed places (Öst, Johansson, and Jerremalm, 1982), animals (Öst, 1989), and blood (Öst, Lindahl, Sterner, and Jerremalm (1984). In general, the principal component of any effective treatment for specific phobias is exposure to the feared situation. Typically, exposure is conducted in vivo (i.e., actual confrontation of the feared situation), although exposure in imagination is also used at times. Some studies have compared exposure to applied relaxation (an approach that integrates relaxation training and exposure to feared situations) and both appear to be effective. In addition, applied tension (i.e., tensing muscles in order to increase blood pressure) has been shown to be an effective treatment for blood phobias, which are often associated with fainting upon exposure to blood.
Almost all of the studies evaluating exposure-based treatments for specific fears have involved volunteers with heightened fears who may or may not have met full criteria for specific phobias (i.e., including significant distress or impairment). In fact, in only four studies were participants carefully diagnosed using recent diagnostic criteria. In these studies (based on individuals with animal or blood phobias), behaviour therapy was shown to be very effective for almost all individuals who were treated (e.g., Öst, Fellenius, and Sterner, 1991; Öst, Salkovskis, and Hellström, 1991).
Preliminary studies with specific phobias suggest that self-instruction methods (e.g., self-help books) may be less effective, compared to studies of PD and PDA.
f. Posttraumatic Stress Disorder (PTSD)
Very little is known about the effectiveness of treatments for PTSD. With respect to medications, few controlled studies have been conducted, and the few studies that have been published have yielded conflicting results. Some studies have shown antidepressants (e.g., amitriptyline, desipramine, phenelzine) to be more effective than placebo (e.g., Kosten, Frank, Dan, McDougle, and Giller, 1991); others have failed to show differences (e.g., Davidson et al., 1990). Medication studies completed to date suffer from a variety of limitations including small sample sizes and a duration of treatment that may be too brief to be helpful.
Studies of CBT have yielded more promising findings (e.g., Foa, Hearst-Ikeda, and Perry, 1995; Keane, Fairbank, Caddell, and Zimering, 1989), although there are still too few studies to conclude that CBT is an effective treatment. Furthermore, no one cognitive-behavioural strategy appears to be more effective than other strategies.