SOCIAL PHOBIA

Social phobia is an exaggerated fear of negative evaluations, ranging from specific to generalized types. Some consider avoidant personality disorder to be the most severe form of social phobia. Social phobia is a chronic, disabling condition whose prevalence is being increasingly recognized.

Double-blind, placebo-controlled studies confirmed that the most effective medication for the treatment of generalized social phobia is the MAOI phenelzine. Therapeutic dosages of phenelzine range from 45 to 90 mg a day, with response rates ranging from 50 to 70 percent. Approximately 5 to 6 weeks are needed to assess the efficacy of phenelzine. Responders should continue treatment for at least a year. Dietary restriction and numerous troublesome adverse effects put MAOIs at a significant disadvantage compared with less-efficacious but more acceptable medication choices. For instance, the SSRIs have recently emerged as viable alternatives to phenelzine. Although not subjected to the same rigorous evaluations, the results of controlled trials (fluvoxamine, paroxetine, sertraline), a controlled discontinuation study (paroxetine), and open-label case reports (citalopram [Celexa]) with SSRIs suggest that their ease of administration and tolerability more than make up for their slightly lower efficacy.

Most clinicians consider SSRIs the first-line treatment choice for patients with generalized social phobia. Since patients with social phobia do not exhibit the supersensitivity syndrome described in panic disorder, SSRI administration can be initiated at usual antidepressant starting dosages (e.g., fluoxetine, 20 mg; paroxetine, 20 mg) and titrated on the basis of clinical response. The benzodiazepines alprazolam and clonazepam also seem efficacious in both generalized and specific social phobias. Their favorable adverse-effect profile and quick onset of action are countered by lower response rates and earlier relapse upon discontinuation than with phenelzine. As in panic disorder, tolerance does not develop to the therapeutic effects of benzodiazepines. Starting and therapeutic dosages for benzodiazepines are the same as those for panic disorder (Table 15.7-1).

Specific social phobias such as fear of public speaking respond moderately well to b-adrenergic receptor antagonists, although the data are mostly anecdotal. Propranolol (20 to 60 mg per dose) counters several of the physiological symptoms of excessive autonomic arousal. Propranolol's short half-life makes it necessary to repeat the effective dose every 2 to 4 hours. If propranolol works but the social phobic situations are unpredictable during the day, a b-adrenergic receptor antagonists with a longer half-life can be tried. Atenolol (Tenormin), 50 to 100 mg before sleep, should control the symptoms for 24 hours. The adverse effects of b-blockers may include sedation, fatigue, dry mouth, gastric cramping, and occasionally confusion and memory problems. Long-term use can lead to depression. Because of the adverse effects and unpredictability of response, the patient is strongly encouraged to test these drugs before the actual event or performance. The utility of b-adrenergic receptor antagonists in generalized social phobia is much more limited. Preliminary evidence supports the use of buspirone as well. Modest response has been noted in patients with social phobia at dosage above 45 mg a day.

Hyperhidrosis (excessive sweating) may be controlled by topical application of aluminum chloride in alcohol (Drysol) while a generalized form of the symptom usually responds to clonidine or terazosin (Hytrin). Ondansetron (Zofran) is the preferred medication for patients with social phobia if the most

disabling symptom is nausea or the fear of vomiting. Nausea may also respond to cisapride (Propulsid), but caution should be exercised when prescribed concomitantly with drugs that inhibit cytochrome P450 (CYP) A4 (CYP 3A4). Paruresis a frequent symptom of social phobia, manifesting as the inability to void in public restrooms, usually requires in vivo exposure and relaxation training. Pharmacological approaches have been largely disappointing but a trial of furosemide and bethanechol may be warranted during behavioral treatment.

The RIMA brofaromine (150 mg a day) was found efficacious in a controlled trial. The finding that most patients maintained their gains at a 9-month follow-up assessment suggests that brofaromine could play an important role in the management of social phobia if it becomes available in the United States.

SPECIFIC PHOBIAS

Specific phobias are quite common and only require treatment if they interfere significantly with the functioning. Treatment is usually behavioral exposure. Medications are used occasionally to alleviate the anticipatory anxiety associated with beginning exposure treatment. Low-dose benzodiazepines and b-adrenergic receptor antagonists can be used for this purpose on an as-needed basis.

GENERALIZED ANXIETY DISORDER

Patients with generalized anxiety disorder suffer from excessive and uncontrollable anxiety and worry for at least 6 months and experience a series of somatic symptoms such as restlessness, irritability, insomnia, and muscle tension. The illness is chronic, with periodic exacerbations and relative quiescence. The relative sparsity of biological data and pharmacotherapy research is due to a number of factors. First, because of their multiple somatic complaints, patients with generalized anxiety disorder are usually seen by generalists and medical specialists other than psychiatrists; generalized anxiety disorder is more likely to be diagnosed as a comorbid condition in psychiatric practices. Second, pharmacotherapy is considered less effective in generalized anxiety disorder than in some other anxiety disorders. Third, the diagnostic features are not clear-cut, and comorbid conditions make the diagnosis difficult.

The efficacy of benzodiazepines in the pharmacological treatment of generalized anxiety disorder gave rise to theories implicating the benzodiazepine–g-aminobutyric acid (GABA) receptor system in the pathophysiology of generalized anxiety disorder, but evidence exists for the involvement of the serotonergic and noradrenergic systems as well. Benzodiazepines remain the traditional medication choice for patients with generalized anxiety disorder. Data do not support the advantage of any one benzodiazepine over others, and no correlation has been established between clinical response and dosage or plasma concentration. A daily equivalent of 15 to 25 mg of diazepam usually suffices to relieve most symptoms in up to 70 percent of generalized anxiety disorder patients. Both somatic and psychic anxiety symptoms respond within the first week of treatment. Tolerance to the sedative effects of benzodiazepines develops quickly, but the antianxiety effect of a given dosage is well maintained over

time in generalized anxiety disorder. However, the relapse rate upon discontinuation of benzodiazepines is high.

Buspirone, the only currently available azapirone, is a potential alternative to benzodiazepine treatment in generalized anxiety disorder. Response rates between 60 and 80 percent have been reported at dosages ranging from 30 to 60 mg a day in three divided doses. While response rates seem comparable, more patients drop out of buspirone trials than benzodiazepine trials. The relative merits of buspirone and benzodiazepines are further detailed under panic disorder. One notable exception is that generalized anxiety disorder patients exposed to benzodiazepines may still be responsive to buspirone unlike panic patients.

Antidepressants are also effective in generalized anxiety disorder patients. Dosages and response patterns are similar to those observed in panic disorder (Table 15.7-1). Increased initial physiological symptoms and anxiety may be related to adverse effects such as dry mouth, constipation, sedation, and positional hypotension rather than to the hypersensitivity syndrome described in panic disorder.

Isolated (but prominent) symptoms such as palpitation, tremor, and sweating may respond to b-blockers within 1 week of treatment, but the full generalized anxiety disorder picture usually requires the use of benzodiazepines, antidepressants, or buspirone. Controlled studies are unavailable, but clinical experience suggests the benefits of combination treatments for generalized anxiety disorder. For instance, the combination of benzodiazepines or b-adrenergic receptor antagonists with antidepressants could yield a rapid response, and when the antidepressant becomes effective, the benzodiazepine or b-adrenergic receptor antagonists can be tapered off. Several new drugs, currently unavailable in the United States, have also been found promising in recent clinical trials. The only double-blind, placebo-controlled study with the partial benzodiazepine agonist abecarnil showed significant improvement on many measures in patients with generalized anxiety disorder. Tropisetrone, a 5-HT3 antagonist, has good antianxiety efficacy with few, mild adverse effects, and serazepine, a 5-HT2 antagonist, also worked well. Gepirone and ipsapirone, two compounds related to buspirone, have demonstrated efficacy comparable to that of buspirone.

Maintenance pharmacotherapy of patients with generalized anxiety disorder should follow the principles given above for long-term treatment of anxiety disorders. Given the fluctuating course of this disorder, periodic discontinuation may be attempted.

PANIC DISORDER WITH OR WITHOUT AGORAPHOBIA

The symptomatic triad of panic disorder consists of panic attacks, usually complicated by anticipatory anxiety and phobic avoidance, or agoraphobia. For a long time the pharmacological approach to panic disorder assumed that blocking panic attacks with medications lead to gradual improvement in both anticipatory anxiety and phobic avoidance. Therefore, panic blockade was the main focus of pharmacotherapy. Response rates to panic blockade alone range from 50 to 80 percent, depending on the definition of response. Since some of these panic-free patients remain avoidant and apprehensive,

clinical practice must consider the full spectrum of panic disorder, including anticipatory anxiety, phobia, and comorbid conditions such as mood disorders.

The natural course of panic disorder varies. A significant proportion of these patients have very favorable outcomes, others suffer from long-term disability, and some have an alternating course. To address these problems, the traditional, relatively straightforward, antipanic drug regimens are gradually replaced or modified in clinical practice by complex decision trees, treatment algorithms, drug combinations, and augmentation strategies. The recommendation for refractory cases frequently includes addition of specific psychotherapeutic techniques as well.

At the time of this writing only five drugs have been approved by the Food and Drug Administration (FDA) to treat panic disorder, but the number of efficacious, off-label antipanic agents are in the dozens. They include high- and low-potency benzodiazepines, selective serotonin reuptake inhibitors (SSRIs), tricyclic and tetracyclic, MAOIs, reversible MAOIs (RIMAs), and novel antidepressants. In explaining the pharmacological options, the clinician should tell the patient about the risks and benefits of these medications. Specifically, the patients should know that most of these medications are equally efficacious and that the choice is usually made on the basis of side effects.

Current consensus is to start treating a patient with uncomplicated panic disorder with low dosages of an SSRI (Table 15.7-1). Even at these low starting dosages many panic patients may experience initial agitation and more-frequent panic attacks. This so-called supersensitivity syndrome is usually limited to the first week or two of treatment. Further dosage reduction, switching to a different compound in the same family, or addition of a high-potency benzodiazepine, such as clonazepam (Klonopin) or alprazolam (Xanax), usually gets the patient through this relatively short period. Dosages can subsequently be increased over several weeks until they reach therapeutic range. For full panic blockade, most patients require dosages equivalent to those used in depression. A therapeutic trial with these agents should last for at least 5 weeks. Panic blockade often leads to improvement in both anticipatory anxiety and phobic avoidance. Preliminary evidence suggests that improvement in phobic avoidance may require higher dosages than panic blockade.

Table 15.7-1 Recommended Dosages for Antipanic Drugs (Daily Unless Indicated Otherwise)

Approximately 60 percent of patients respond to this approach. Partial responders can benefit from the addition of a high-potency benzodiazepine or buspirone. Partial responders with tachycardia may try a combination of an SSRI and b-adrenergic receptor antagonists. b-adrenergic receptor antagonists may also alleviate cardiac discomfort associated with mitral valve prolapse. The dosage of b-adrenergic receptor antagonists such as propranolol (Inderal) or atenolol (Tenormin) should be titrated until the heart rate is reduced by 5 to 10 beats per minute. Contraindications to using b-adrenergic receptor antagonists include bradycardia, heart block and asthma. Drug-drug interactions between some SSRIs and some b-adrenergic receptor antagonists should be kept in mind.

Nonresponders should be tapered off the SSRI and offered one of the following options: a second SSRI, venlafaxine (Effexor), a high-potency benzodiazepine, a tricyclic or tetracyclic, or an MAOI. Persistent adverse effects such as weight gain, hypomania, or sexual dysfunction, may also necessitate a switch from an SSRI in otherwise fully responding panic patients, although first an attempt should be made to address these adverse effects.

Originally, one of the most important advantages of SSRI treatment was considered the absence of withdrawal reaction upon discontinuation. However, withdrawal reactions ranging from mild, transient anxiety and insomnia to severe headache, nausea, dizziness, and “electric jolts” lasting for several months have since been reported in up to 86 percent of patients who abruptly discontinue taking SSRIs. Therefore, gradual taper over several weeks is strongly recommended. Anecdotal reports suggest that the addition of benzodiazepines, or trazodone for sleep may alleviate SSRI withdrawal.

MAOIs MAOIs such as phenelzine (Nardil) or tranylcypromine (Parnate) are the most likely to be efficacious in panic disorder, but they present the dilemma of dietary restrictions and the danger of hypertensive reaction, weight gain, orthostatic hypotension, insomnia, and a series of anticholinergic adverse effects. While somewhat less effective, a RIMA, such as moclobemide (Aurorix), is an alternative. After 2 weeks of washout (6 weeks for fluoxetine), the MAOI can be given and the dosage gradually raised to usual antidepressant levels. Comorbid major depressive in disorder in panic disorder patients may respond best to MAOIs. Because of severe withdrawal reactions such as disinhibition, irritability, agitation, insomnia, myoclonic jerks, and occasionally, delirium, thought disorder, cognitive impairment, and mania, gradual taper is strongly recommended before MAOIs are discontinued. The dietary restriction should continue for at least 2 weeks beyond discontinuation.

High-Potency Benzodiazepine High-potency benzodiazepines should be considered if the adverse effects of all other alternatives are unacceptable to the patient, or they may be first-line choices if the patient is unwilling or unable to wait out the 4- to 5-week delay in response associated with most antidepressants. The advantage of switching to a benzodiazepine also includes no wash-out time for the SSRI. In fact, a benzodiazepine can be added to the SSRI during taper. A typical switch would entail adding 0.5 mg of clonazepam twice a day to 40 mg of paroxetine (Paxil). A few days later the paroxetine dose can be lowered to 20 mg, and depending on its sedative effects, clonazepam dosage can be raised to 1 mg twice a day. Every 3 to 4 days the dosage of paroxetine can be halved while clonazepam dosage is adjusted as tolerated.

High-potency benzodiapines and perhaps low-potency benzodiazepines as well are powerful antipanic drugs. Their sedative and habituating potentials are easily managed in most patients and should not significantly diminish their clinical utility. A major advantage of benzodiazepines is their quick onset of action. Most responders benefit within 1 week. Panic patients with prior history of substance use disorder should not be treated with benzodiazepines. Older patients should take significantly lower doses of benzodiazepines because they are more sensitive to sedation and the potential cognitive side effects of benzodiazepines. A history of organic brain disease or significant character pathology may predispose patients to experience disinhibition while taking benzodiazepines.

The serious and clinically limiting problem of withdrawal symptoms from benzodiazepines prompted many studies attempting to alleviate the withdrawal reaction, also referred to as discontinuance syndrome. Studies with clonidine (Catapres), propranolol (Inderal), and carbamazepine (Tegretol) yielded inconsistent results, and substituting other medications such as antidepressants or azapirones that are easier to taper has not been studied sufficiently. Cognitive-behavioral therapy may best counter benzodiazepine withdrawal. The severity of withdrawal is related to the dosage and length of use and to the rate of taper. The rule of lowest efficacious dose for the shortest time needed for improvement should always be observed. Tapering should never exceed a 10 percent dosage reduction every 3 days. Several months of tapering after years of high-dose benzodiazepine use is frequently unavoidable.

Follow-up studies suggest that benzodiazepine-responsive panic patients maintain their gains over several years and do not develop tolerance to the drugs' antianxiety effects. In fact, maintenance dosages of benzodiazepines are frequently lower than dosages during acute treatment. Residual symptoms may require additional treatment. Since use of short-half-life benzodiazepines such as alprazolam may be associated with between-dose rebound anxiety, some clinicians prefer benzodiazepines with longer half-lives, which may also be easier to taper.

Tricyclics and Tetracyclics Switching to a cyclic antidepressant should begin with tapering the SSRI dosage down to the starting dosage. Some clinicians consider it safe to add the tricyclic at low starting dosages while others recommend a drug-free period lasting from 24 hours to 1 week before administering the second drug. The reason for caution is the potential development of the serotonin syndrome, a rare but severe reaction that may occur when two serotonergic drugs are combined. Also, plasma concentrations of tricyclic and tetracyclics may increase due to inhibition of their metabolism by the SSRI.

The first drug ever shown to possess antipanic efficacy is the tricyclic drug imipramine, and many clinicians still consider it the gold standard in panic disorder. Tricyclic and tetracyclic are generally well tolerated. Their clinical utility is clearly limited, however, by anticholinergic and cardiovascular adverse effects and toxicity in overdose. One of the most effective tricyclic drug for panic is clomipramine (Anafranil), which seems to be effective at lower than full antidepressant dosage. A linear relationship may exist between total plasma tricyclic drug concentrations and response rate. In general, agoraphobia may respond to higher dosages than panic attacks alone. Because of the variable nature of the withdrawal reaction upon discontinuation of tricyclic or tetracyclic drugs (somatic distress, sleep

disturbance, behavioral activation, cardiac arrhythmias), taper should be tailored to the patient's needs. Halving the original dosage will give an early indication of withdrawal severity.

Secondary Alternatives While controlled data are unavailable, clinicians increasingly turn to venlafaxine (Effexor) as an alternative to SSRIs and tricyclic drugs in the treatment of panic disorder. If such initial adverse effects as nausea and agitation are overcome, venlafaxine is a powerful antipanic agent at a dosage significantly below its antidepressant dosage. Since the safety of combining venlafaxine and an SSRI has not been systematically assessed, the SSRI should be tapered and discontinued before switching to venlafaxine. A potential limitation of venlafaxine is an unusually severe reaction to taper. Many successfully treated patients complain of nausea, headache, fatigue, dizziness, and gastrointestinal disturbance when trying to reduce the venlafaxine dosage. Anecdotal evidence suggests that the temporary addition of fluoxetine (10 mg a day) may reduce the withdrawal reaction and allow discontinuation of venlafaxine.

Switching to a second SSRI may be justified if the adverse effects of all alternative possibilities are unacceptable to the patient or there is a relative contraindication to using anything but an SSRI. Again, in the absence of controlled data, opinions differ on how to switch between two SSRIs. Since the potential for the serotonin syndrome is even higher than in the case of combining an SSRI and a tricyclic antidepressant, caution is warranted, and most clinicians recommend a drug-free period between two SSRIs.

Buspirone was promoted as a less-sedating alternative to benzodiazepines in the treatment of panic disorder. Buspirone has lower potential for abuse and dependence than benzodiazepines and produces relatively few adverse effects and no withdrawal syndrome. Buspirone does not alter cognitive or psychomotor function, does not interact with alcohol, and is not a muscle relaxant or an anticonvulsant. However, the efficacy of buspirone in panic disorder is disappointing, and with its further handicap of delayed onset of action and the need for multiple dosing, its use is limited to potentiating the efficacy of other antidepressants and countering the adverse sexual effects of SSRIs. Also, buspirone seems even less effective in patients previously exposed to benzodiazepines.

Bupropion, maprotiline (Ludiomil), and trazodone have not been found efficacious for panic disorder in controlled studies, while the anticonvulsants divalproex (Depakote) and gabapentin (Neurontin), the polyol second-messenger precursor inositol, nefazodone, and the calcium channel inhibitor verapamil (Calan, Isoptin) have shown promise as antipanic agents.

While the short-term efficacy of antipanic medications has been established, the question of how long to treat a panic patient who responds to treatment remains open. The results of follow-up studies are mixed. Several reports indicate that most panic patients relapse within 2 months to 2 years after the medication is discontinued. A recent review concludes that following medication discontinuation, only about 30 to 45 percent of the patients remain well, and even remitted patients may display a variety of symptoms. Others find that while occasional panic attacks are quite common, otherwise responding patients rarely revert back to significant phobic avoidance or serious vocational or social disability. Improvement may continue for years following a single course of medication treatment. This favorable

outcome may be explained by the heterogeneity of panic disorder, spontaneous learning experience of patients in clinical trials, and concomitant self-monitoring. Given the uncertainty about the optimal duration of treatment, the current recommendation is to continue full-dosage medication for panic-free patients for at least 1 year. Medication taper should be slow, with careful monitoring of symptoms. Distinction should be made among return symptoms, withdrawal, and rebound anxiety.

Since longer duration of illness at baseline predicts poor long-term outcome, all efforts should be made to identify and treat panic patients as early as possible. More-severe phobic avoidance and comorbid depression and social phobia at baseline also predict poor long-term outcome. Higher depression scores coincide with greater severity of avoidance and disability. The poorer overall outcome in panic disorder patients with comorbid recurrent depression is more likely due to the simultaneous presence of the two conditions. Comorbid depression usually improves in parallel with panic symptoms.

Atypical responses to medications have been reported in panic patients whose first panic attacks were precipitated by cocaine use. These patients respond preferentially to benzodiazepines and anticonvulsants, while tricyclic drugs seem to worsen their anxiety symptoms. This pattern of medication response suggests that cocaine-induced panic attacks may be related to a kindling-like phenomenon.

In a small series of patients with comorbid depression and panic disorder electroconvulsive therapy (ECT) seemed highly effective. None of these patients experienced panic attacks after their fourth ECT treatment until discharge.

15.8 ANXIETY DISORDERS: PSYCHOLOGICAL TREATMENTS

LAWRENCE A. WELKOWITZ, PH.D.

Cognitive-Behavioral Therapy

Psychosocial Therapy

Future Directions

Suggested Cross-References

Since the late 1980s there has been tremendous progress in the nonpharmacological treatment of anxiety disorders. Cognitive-behavioral therapies, which reflect a recent integration of the cognitive theories and methods associated with Aaron Beck, Albert Ellis, and David Clark, and the behavioral theories and methods of B.F. Skinner and Ivan Pavlov, currently have the greatest degree of empirical validation. Panic disorder, which has been said to be the most disabling of the anxiety disorders in terms of social and occupational functioning is a case in point in that recently developed treatments contain elements based on classical or Pavlovian conditioning, behavioral techniques of exposure, and cognitive restructuring of irrational beliefs and overvalued ideas. Other anxiety disorders for which cognitive-

behavioral approaches have been particularly effective include social phobia (both generalized and nongeneralized type), obsessive-compulsive disorder, specific phobia, social phobia, and generalized anxiety disorder. Further validation is needed for two recently developed treatments for posttraumatic stress disorder, one based on exposure methods, and the other based on classical conditioning (eye movement desensitization and reprocessing [EMD/R]). Psychosocial (e.g., supportive and psychodynamic) therapies have also been used in the treatment of anxiety disorders.

COGNITIVE-BEHAVIORAL THERAPY

Panic Disorder Cognitive-behavioral treatment of panic disorder or panic control therapy produces 80 to 90 percent panic-free status, at least within 6 months of treatment. Two-year follow-up indicates that more than 50 percent of those patients who originally responded to panic control therapy have occasional panic attacks, and more than a quarter will seek additional treatment. Nonetheless, these treatment responders do tend to have a significant decline in panic-related symptoms and most maintain many of their treatment gains. A large multi-center study comparing medication and panic control therapy treatment being carried out at the University of Pittsburgh, Yale University, State University of New York at Albany, and Columbia University should help to resolve many questions about long-term gains in panic control therapy and pharmacological treatment.

Panic control therapy is an amalgam of techniques drawn from a variety of cognitive and behavioral methods and based largely on Peter Lang's three-system model of cognitive psychophysiology, which focuses on the interactions of three systems affecting the human experience of panic and anxiety: physiological (e.g., palpitation, sweating, dizziness, nausea), cognitive (fears of losing control or going crazy), and behavioral (avoidance, pacing). Panic control therapy techniques are thus directed towards each of these systems: breathing control techniques are designed to control the physiological effects of hyperventilation and progressive muscle relaxation helps reduce the overall negative effects of muscle tension on anxiety. Cognitive restructuring techniques, which focus on catastrophic thinking errors (e.g., “I will have a heart attack and die”) as well as on misinterpretation of harmless bodily associations, are directed towards problems in the cognitive system. Education about the nature of anxiety and panic, which also counters the myths of panic attacks, (i.e., that they are associated with schizophrenia or heart disease) also serves as a form of stress inoculation in the cognitive domain. Exposure to feared situations and events, such as crowded public areas, helps to reduce avoidance symptoms of the behavioral domain. A novel technique called interoceptive exposure involves repeated exposure to the physical sensations associated with panic. This method, which is based on a classical or Pavlovian conditioning model of panic, consists of a series of activities, such as deliberately overbreathing (to produce physical effects of hyperventilation), breathing through a straw while holding one's nose (feelings of breathlessness), spinning in a chair (dizziness), running in place (increased heart rate), staring at one's hand (feelings of unreality), and so on. These activities are rehearsed until an habituation of the anxiety response has been achieved.

In clinical practice the degree to which various techniques are emphasized may depend on particular features of the panic disorder. For example, a patient with a high degree of agoraphobic avoidance may require additional exposure to fearful situations. The house-bound person with agoraphobia may

require therapist-assisted in vivo exposure therapy as opposed to less labor intensive and more instruction-oriented exposure therapy for the patient with minimal agoraphobia and panic disorder. These panic control therapy techniques may also be carried out successfully by therapists whose primary mode of treatment is pharmacological. Finally, further research is needed to identify effective components of treatment and to determine whether certain subtypes of patients may require certain components of panic control therapy over others. In clinical practice, for example, it appears that a substantial number of patients with panic disorder respond well to education and breathing control alone whereas others seem to require more extensive treatment.

José was a 27-year-old laboratory technician who began having full-blown panic attacks 8 months prior to seeking help at our research clinic. While he was unable to identify specific situations that elicited attacks, he was particularly concerned about the possibility of them occurring while he was engaged in laboratory procedures with patients. His attacks typically involved a sudden explosion of autonomic arousal and included palpitations, sweating, dizziness, feelings of unreality, and tingling in his arms and legs. He dreaded the idea that the attacks might reoccur. In the beginning of his cognitive-behavioral program, he found an educational handout that described the myths of panic attacks (e.g., that they will lead to heart attacks, losing control, or going crazy) particularly reassuring. He began practicing diaphragmatic breathing each evening, and after several weeks, became effective in challenging his negative way of thinking about the consequences of panic attacks. In the latter few weeks of his 12-week program, he practiced exposing himself to physical sensations of panic by doing a variety of interoceptive exercises at home, including hyperventilating for 1 or 2 minutes at a time (designed to help José acclimate to the physical sensations associated with overbreathing), and spinning in a chair repeatedly (designed to help acclimate him to symptoms of dizziness and feelings of unreality). At the conclusion of the treatment program José's panic attacks had disappeared, and at 6-month follow-up he had maintained his treatment gains by attending “booster sessions” with his therapist once every 2 months.

In contrast to cognitive-behavioral formulations, it has been reported that patients who develop panic symptoms in their late 20s have histories replete with relationship problems as well as other neurotic symptoms stemming from parent-child disruptions. These patterns, in turn, are said to lead to disturbances in empathy and understanding, which may ultimately elicit panic symptoms. Recently, detailed treatment protocols have been developed that are based on psychodynamic formulations of anxiety, consisting primarily of supportive-expressive psychodynamic psychotherapy. Preliminary outcome research indicates that these psychodynamically oriented protocols are useful in modifying anxiety, depression, and related interpersonal problems. Studies on hypnotherapy and hypnoanalysis, in which early childhood conflicts are identified and ultimately tied to the origins of panic and anxiety, provide some support for a more traditional analytic approach.

The importance of cultural issues in the treatment of anxiety disorders should not be overlooked. African-Americans with panic disorder appear to respond less with panic control therapy compared with whites. This may be related to the high prevalance of untreated hypertension among African-Americans, which may, in turn, exacerbate anxiety symptoms or possibly lead to higher comorbidity of emotional problems. Conversely, it has been observed that Asian Americans and Native-Americans, whose philosophical systems emphasize the control of mind over body and who have also utilized breathing control methods for centuries, are particularly receptive to the breathing and cognitive control aspects of panic control therapy. Interestingly, data from the Epidemiologic Catchment Area (ECA) study did not detect any differences in prevalence of panic disorder between African-American, Hispanic, and white groups.

Social Phobia As with panic disorder, considerable progress in the psychological treatment of social anxiety or social phobia is linked to the application of cognitive-behavioral methods. Unlike more traditional psychotherapies, cognitive-behavioral approaches do not address the origins of social anxiety, but instead focus on the use of coping strategies that can be implemented in current fearful situations. The most thoroughly studied form of cognitive-behavioral therapy for social phobia is a group therapy consisting of several discrete entities including (1) presentation of a three-system (cognitive-behavioral-physiological) model of social anxiety; (2) training in identification and restructuring of irrational beliefs regarding social performance; (3) in-session exposure to feared social situations via group role-playing scenarios; and (4) homework assignments directing patients to utilize cognitive and exposure techniques in vivo). Groups are particularly amenable to the treatment of social phobia in that they provide natural opportunities for patients to practice feared behaviors in a supportive and informative context.

Outcome research is somewhat limited but one study showed that cognitive-behavioral group therapy was nearly twice as effective as standard educational-supportive group psychotherapy. Responders to cognitive-behavioral group therapy were also shown to maintain treatment gains to a considerable extent at 5-year follow-up. Questions remain as to the effective treatment component in these therapies that blend cognitive and exposure-based methods. For example, it is unclear whether exposure to feared situations alone, without cognitive therapy, would be just as effective as the combined treatment. Also, it is not known whether group therapies other than educational or supportive group therapy, such as interpersonal group therapy, may be effective in treating social phobia.

In terms of the differential effects of cognitive-behavioral group therapy on patients of varying subtypes (generalized versus nongeneralized social phobia), it appears that patients with more discrete or nongeneralized fears, such as of public speaking, respond best to these type of treatments. Pharmacological research has also shown differential effects based on subtype, with beta-blockers being more effective for nongeneralized fears, and antidepressant agents such as the monoamine oxidase inhibitor (MAOI) phenelzine (Nardil), being more effective for generalized social phobia. For generalized as well as more severe forms of social phobia, a combination of psychological and pharmacological therapy may be in order. Alternatively, it may be that longer-term cognitive-behavioral therapy, with or without medication, is warranted.

Despite the lack of empirical validation via controlled studies on large samples, social phobia is commonly treated with other behavioral methods. Systematic desensitization, first described in 1958, is based on a classical conditioning model of fears. It involves a pairing of previously conditioned stimuli (e.g., hierarchical presentation of images of feared social situations) with a relaxed state (promoted by means of progressive muscle relaxation). This new pairing is said to create a new set of associations that lead the patient to no longer fear previously feared stimuli. Regardless of the specific behavioral technique utilized, the key mechanism of effective treatment is related to the development of self-efficacy, or a sense of confidence that a patient experiences when confronting challenging social events or situations.

Jane was a 31-year-old marketing executive at a rapidly expanding technology company. She contacted our research clinic for help with her fear of public speaking, which had become so severe in recent months that she was at risk for losing her job, which required a considerable amount of public presentation and leading training programs. Jane was assigned to a treatment program that entailed a combination of medication (b-adrenergic receptor antagonist the beta-blocker atenolol [Tenormin]) and a 12-week group-based cognitive-behavioral treatment program for nongeneralized social phobias, such as fear of public speaking. Before beginning the group program, Jane met with her behavior therapist who helped her develop a hierarchy of the public-speaking situations she feared. Difficulty levels ranged from speaking up at meetings involving five or fewer colleagues, which caused a low level of anxiety, to leading a training program for 100 or more employees and colleagues at annual corporate gatherings, which caused a very high level of anxiety. In her group treatment, Jane began by identifying and challenging negative thoughts associated with each situation (e.g., “They won't understand what I am saying” became “I have always explained myself clearly in the past”). The other group members then assisted Jane in designing and carrying out role-plays that allowed her to practice employing her new coping responses and helped her to develop new public speaking skills. By the end of treatment, Jane still experienced occasional anxiety when making a public presentation to large numbers people, but she no longer avoided such situations.

Finally, it is important to consider cultural factors when treating social phobia. In Japan, a condition called Taijin Kyofusho refers to a type of social anxiety in which the sufferers become preoccupied with the idea that they are causing others to be embarrassed or humiliated. Other Taijin fears include fear of emitting a body odor and fear of one's facial expression becoming too stiff. In utilizing cognitive therapy with Japanese and other Asian patients, it is important to consider general cultural characteristics, such as the importance of displaying hierarchical deference (i.e., respect for those in authority) as well as an emphasis on communal versus individual expression. The emphasis on rigid moral codes and religious ritual and custom in Arab culture is said to place individuals at higher risk of social phobia, reflecting an underlying fear of betraying valued ideas and formalities. It has been reported that 13 percent of all outpatient psychiatric visits (excluding for schizophrenia) in Saudi Arabia are for treatment of social phobia.

Posttraumatic Stress Disorder Treatment of posttraumatic stress disorder is based largely on the aversive effects of a traumatic event or series of traumatic events. Specifically, fears related to this disorder are developed via classical or Pavlovian conditioning in which the unconditioned stimulus (e.g., a horrible accident or sexual assault) that automatically produces a fear response or unconditioned response is paired with other stimuli (e.g., the sounds, smells, and sights associated with the traumatic event). These conditioned stimuli then come to elicit a fear response independent of the original unconditioned stimuli. According to O. Hobart Mowrer's two-factor theory these new classically conditioned fears are subsequently maintained via operant conditioning in that escape or avoidance behaviors are reinforced by their anxiety-relieving effects. Through a process of higher-order conditioning, increasing numbers and types of stimuli may become elicitors of an anxiety or a posttraumatic stress disorder response. For example, if particular smells present during the original trauma are later present in more innocuous situations, such as when walking down the street or visiting a friend or relative, the patient may eventually evidence a fear response.

It has been argued that Mowrer's two-factor theory only partially explains the cognitive, affective, and physiological response unique to posttraumatic stress disorder. Symptoms such as trauma-related nightmares and psychological numbing are believed to be best understood by cognitive network theories, which posit the development of a “fear network” of memories. These cognitive networks, which include cognitive, behavioral, and physiological information about trauma responses, are formed following a traumatic event and are activated by trauma-related stimuli. According to this model, posttraumatic stress disorder responses can change across time as the interpretation of original trauma events produces changes in its meaning. Thus, a delayed response in this disorder can be explained by a subsequent change in the personal meaning of the original trauma, such as when an originally asymptomatic woman who has been sexually assaulted later discovers that her assailant has been released from prison and has been stalking previous victims.

Regardless of the emphasis on cognitive or conditioning processes, empirically validated treatments of posttraumatic stress disorder employ similar components, including (1) cognitive monitoring and restructuring, (2) exposure in imagination to the original traumatic event, and (3) exposure in vivo to trauma-related stimuli. The overall goal is to reduce or eliminate the patient's anxiety reaction to trauma-related cognitive and environmental stimuli. Cognitive restructuring is based on the notion that fear networks or structures can be directly modified by identifying and modifying negative or nonproductive thinking patterns related to concerns about threat or danger. Cognitive therapy may proceed in traditional fashion, based largely on therapist-patient dialogue, or may utilize homework-based written exercises in which the patient records irrational beliefs and composes more adaptive, coping-oriented responses that are based on new interpretation of the meaning of these trauma-related events. At least one study has found this cognitive processing therapy to be more effective than a no-treatment control group in reducing the symptoms of posttraumatic stress disorder.

In vivo and imaginal exposure techniques, which involve graduated exposure either to real life or cognitive trauma-related stimuli, are clearly useful in reducing overall anxiety as well as trauma-related fears. Both imaginal and in vivo techniques may involve return to the original place where the traumatic event occurred, or may be limited to situations that produce anxiety in the present (e.g., approaching

certain people or places that evoke trauma-related memories). These methods have been shown to be effective for combat veterans suffering from the disorder and for women who have been sexually assaulted. More recently, a 4-week program has been developed to prevent the symptoms of this disorder for women who have been recently raped or assaulted. This program consists of (1) education about common reactions to trauma, (2) relaxation skills, (3) exposure in imagination to the original trauma, and (4) cognitive restructuring. A preliminary study comparing 10 women participants and 10 nonparticipants showed that 70 percent of the untreated women and only 10 percent of the treated women suffered symptoms of posttraumatic stress disorder at 2 months following the original trauma.

Jennifer was a 21-year-old college student who sought treatment in the anxiety clinic after being raped repeatedly at a fraternity party. She had difficulty immediately remembering details of the rape because she had ingested an unknown recreational drug that had been slipped into the punch bowl at the party. In the days following the rape, she slowly began to remember details of the traumatic event, often in the form of flashbacks and nightmares. She also experienced increased generalized anxiety, feelings of unreality, and an intense fear reaction to stimuli associated with the attacks, such as the sight of college men wearing sweat shirts with fraternity names and logos, as well as certain odors present during the party. Her treatment program was administered one-on-one by a psychologist who specialized in posttraumatic stress disorder. In the first few sessions, Jennifer was educated about the disorder and the nature of anxiety and was provided a rationale for an exposure-based treatment program. Ensuing sessions involved extensive discussion about the details of the traumatic event, as well as in-session exposures in-imagination to the original trauma. Exposure-based homework assignments complemented these in-session activities and were designed to treat Jennifer's new tendency to avoid trauma-related situations, including campus parties and other social gatherings. Slowly, Jennifer experienced decreased reactivity to trauma-related stimuli and began to enter situations she had previously avoided. After approximately twenty-five sessions Jennifer's symptoms were almost gone and she had nearly returned to normal in terms of participating in campus-related social events and parties. Despite her gains, Jennifer remains, in her words, “more careful” about certain high-risk social events (e.g., she will rarely enter a fraternity house), and finds that her trust in men is markedly diminished. A previous supporter of fraternities on campus, Jennifer is now an ardent advocate of removing fraternities from all colleges.

Unfortunately, little has been written about cross-cultural aspects of treatment of posttraumatic stress disorder. This is surprising in light of its particular relevance for minority groups in the United States, such as African-Americans from socially disadvantaged groups who witness a great deal of violence in their communities. For example, anxious black children, compared with their non-anxious peers, were found to experience a significantly greater degree of trauma. Observers of Caribbean-American immigrants have noted an association between acculturation and symptomatology of posttraumatic stress disorder. These researchers support the use of group supportive treatments of the disorder for

this ethnic group, but warn against the practice of mixing patients with those whose disorder is of differing causes (e.g., sexual assault versus acculturation).

Generalized Anxiety Disorder Recent data from the National Anxiety Disorders Screening Day (NADSD) revealed a surprisingly high 1-month prevalence (11.5 percent) of generalized anxiety disorder for those participants in an anxiety screening program who met criteria for one anxiety disorder only as compared with 5.4 percent for panic disorder, 3 percent for posttraumatic stress disorder, 3.4 percent for social phobia, and 0.7 percent for obsessive-compulsive disorder. Generalized anxiety disorder is particularly prevalent among those with multiple anxiety problems; For example, in the NADSD sample 10.2 percent met the criteria for panic disorder and generalized anxiety disorder, and 7.9 percent met criteria for panic, social phobia, and generalized anxiety disorder. The latter may therefore be considered a hidden problem in that other comorbid anxiety problems may supercede it in terms of focus of treatment.

Cognitive-behavioral formulations and treatment programs dominate this area, despite the lack of data on etiology or psychotherapy outcome. Debate continues as to whether generalized anxiety disorder should be considered a discrete anxiety disorder at all, with some arguing that high comorbidity with other anxiety and depressive disorders suggests that it is merely a basic anxiety trait that serves as a platform for the development of other related problems. Citing the inability of the Center for Epidemiologic Studies Depression Scale (CES-D) to distinguish between depression and generalized anxiety disorder, at least in a sample of mothers of handicapped children, symptoms were noted as being, in some circumstances, a natural component of more severe depression. Data from the ECA study revealed that generalized anxiety disorder was more common among people of lower socioeconomic status as well as in African-Americans, thus suggesting a sociocultural factor in etiology.

Psychological theories and treatment programs alike have tended to focus on the issue of chronic and excessive worrying among individuals who have generalized anxiety disorder. Unlike those suffering panic attacks who experience catastrophic-type cognition (e.g., “I'm going to have a heart attack and die”), those with generalized anxiety disorder exhibit less emotionally charged cognitions on an ongoing basis (e.g., “I'll never make it through the day,” or “I'm sure something will go wrong”). Excessive worriers actually experience less physiological arousal (e.g., decreased heart rate) in response to a fearful stimuli as compared with nonworriers, and it has been speculated that generalized anxiety disorder may actually serve as a type of defense against response to more fearful events.

Treatment programs to date contain one or both of the following components: (1) cognitive restructuring designed to identify worry-related thoughts and replace them with more positive coping responses and (2) relaxation training designed to reduce excessive physiological arousal. While controlled studies have shown that these methods are superior to either nondirective therapies or no-treatment controls, identification of the most effective treatment components remains unclear as does the superiority of these methods to other types of treatments, such as interpersonal-oriented therapies or other more alternative treatments, such as biofeedback or meditation.

A large-scale study comparing cognitive restructuring, relaxation training, a combined cognitive plus relaxation program, and a no-treatment group revealed improvements in all three active treatments

relative to the control condition. While responders were remarkably successful in maintaining gains at 2-year follow-up, no differences were found among the three active treatments. Additionally, study dropouts and nonresponders remained symptomatic at follow-up. In similar fashion, a study comparing relaxation training, a combined cognitive therapy with relaxation training, and a nondirective psychotherapy demonstrated the superiority of the cognitive-behavioral programs compared to a treatment consisting of Rogerian-style reflecting about worrisome thoughts. A longer-term follow-up indicated the slight superiority of the combined cognitive-behavioral treatment over the relaxation-alone condition.

Melinda's friends and family refer to her as a “worrywart.” She constantly worries about something, whether it be that the chicken she is cooking may have salmonella, that her children may not be safe while driving to school, or that her family will not have enough money to pay for expenses. As Melinda put it during her initial evaluation, “I never seem to stop worrying.” In addition to her chronic worrying, Melinda experiences ongoing physical discomfort, including headaches, nausea, sweating, and occasional shortness of breath. She does not, however, experience full-blown panic attacks. Her treatment consisted of a combination of relaxation (specifically, progressive muscle relaxation practiced on a twice-daily basis) and cognitive therapy (identifying and challenging worry-related thoughts and replacing them with more productive coping responses). For example, she applied a type of “probability analysis” to her constant fear that her children would be harmed in a car accident and produced a coping response along the lines of “It is unlikely that they will be in an accident” as well as “My son is an excellent driver.” Following completion of treatment, Melinda reported that her family saw a tremendous improvement in terms of her chronic worrying. While worry-type thoughts would still occasionally enter her consciousness, she now felt that she had effective tools for dealing with them.

PSYCHOSOCIAL THERAPY

Supportive and psychodynamic therapies have been widely used in the treatment of generalized anxiety disorder, panic disorder, and posttraumatic stress disorder, although further research is needed to establish their efficacy.

According to the 1998 American Psychiatric Association's Practical Guidelines for the Treatment of Patients with Panic Disorder, psychosocial therapies may be of use. They have some application to the other anxiety disorders discussed in this section. A summary of the relevant areas of the report follows.

Psychodynamic Psychotherapy Psychodynamic psychotherapy is based on the concept that symptoms result from mental processes that may be outside of the patient's conscious awareness and that elucidating these processes can lead to remission of symptoms. Moreover, in order to lessen the patient's vulnerability to panic, the psychodynamic therapist considers it necessary to identify and alter core conflicts. The goals of psychodynamic psychotherapy may be more ambitious and require more time to achieve than those of a more symptom-focused treatment approach. There are some case

reports of brief dynamic psychotherapies that took no longer than cognitive-behavioral therapy to achieve reasonable treatment goals for patients with panic disorder.

In psychodynamic psychotherapy, the successful emotional and cognitive understanding of the various elements of psychic conflict (impulses, conscience, internal standards that are often excessively harsh, psychological defense patterns, and realistic concerns) and reintegration of these elements in a more realistic and adaptive way may result in symptom resolution and fewer relapses. To achieve this insight and acceptance, the therapist places the symptoms in the context of the patient's life history and current realities and extensively uses the therapeutic relationship to focus on unconscious symptom determinants.

Combined Treatments Investigators have examined use of the combination of medication and cognitive behavior therapy for patients with panic disorder and agoraphobia. Several short-term treatment studies have shown that the combination of the tricylic medication imipramine (Imipramine) with one component of cognitive behavior therapy, behavioral exposure, may be superior to either treatment alone. Another study showed that selective serotonin reuptake inhibitors, such as paroxetine (Paxil), plus cognitive therapy worked significantly better for patients with panic disorder than cognitive therapy plus placebo. There has been one study of the combination of psychodynamic psychotherapy with medication. This study suggested that psychodynamic psychotherapy may improve the long-term outcome of medication-treated patients.

Group Therapy Reports on group therapy in the treatment of panic disorder have consisted primarily of cognitive behavioral approaches; the improvements with group cognitive behavior therapy were comparable to those in studies of individually administered cognitive behavior therapy and pharmacological treatment.

Patient Support Groups Patient support groups are very helpful for some patients suffering from panic disorder. Patients have the opportunity to learn that they are not unique in experiencing panic attacks and to share ways of coping with the illness. Support groups may also have a positive effect in encouraging patients to confront phobic situations. Finally, family members of patients with panic disorder may benefit from the educational aspects of patient support groups. In deciding to refer a patient to a support group, however, it is imperative that the psychiatrist obtain information about the nature of the group and the credentials of its leader. Support groups are not a substitute for effective treatment but are complementary to it.

FUTURE DIRECTIONS

There has been significant progress in recent years in the psychological treatment of anxiety disorders. Most of this progress can be linked to research on cognitive behavioral therapies, which are symptom-directed and relatively short-term. A common feature of all these treatments is active exposure to distressing stimuli, either in vivo or in imagination. These exposures vary depending on the type of anxiety disorder (e.g., the patient with social phobia is exposed to distressing social stimuli, such as parties or public-speaking events). Psychological treatments also include various forms of cognitive therapy that target irrational ideas emblematic of a particular disorder (e.g., the catastrophic fears of a

patient with panic disorder). In all cases the goal is to assist the patient to endure discomfort by regularly practicing stress-reducing techniques.