brand logo

Am Fam Physician. 1998;57(10):2405-2412

See editorial on page 2328.

See related patient information handout on panic disorder and agoraphobia, written by the authors of this article.

Panic disorder is a distressing and debilitating condition with a familial tendency; it may be associated with situational (agoraphobic) avoidance. The diagnosis of panic disorder requires recurrent, unexpected panic attacks and at least one of the following characteristics: persistent concern about having an additional attack (anticipatory anxiety); worry about the implications of an attack or its consequences (e.g., a catastrophic medical or mental consequence) and making a significant change in behavior as a consequence of the attacks. A variety of pharmacologic interventions is available, as are non-pharmacologic cognitive or cognitive-behavioral therapies that have demonstrated safety and efficacy in the treatment of panic disorder. Early detection and thoughtful selection of appropriate first-line interventions can help these patients, who often have been impaired for years, regain their confidence and ability to function in society.

Panic disorder is an anxiety disorder characterized by unexpected panic attacks. It is often associated with situational (agoraphobic) avoidance stemming from fear of further attacks.1 Epidemiologic studies suggest that panic disorder, with or without agoraphobia, has a lifetime prevalence between 1.5 and 3.0 percent1 and a familial tendency. It can run a chronic, relapsing course and can produce significant disability and personal distress. Panic disorder is commonly seen in the family practice setting, but it often eludes detection or is misdiagnosed because its clinical presentation mimics that of other medical conditions.2 A large body of evidence shows that panic disorder responds to various pharmacotherapies and to cognitive and cognitive-behavioral therapies. Early recognition and prompt, appropriate treatment are the keys to managing this disorder effectively.

Identifying Panic Disorder With and Without Agoraphobia

Panic disorder is characterized by the unexpected, “out of the blue” panic attack. A panic attack is defined as a discrete episode of intense symptoms that peak within 10 minutes and primarily involve sympathetic nervous system manifestations. According to criteria given in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV),1 a panic attack must include at least four of the symptoms listed in Table 1.

Neurologic symptoms
Dizzy, light-headed or unsteady feeling
Cardiac symptoms
Chest pain or discomfort
Palpitations, heart pounding or tachycardia
Respiratory symptoms
Shortness of breath
Feeling of smothering or choking
Gastrointestinal symptoms
Abdominal distress
Psychologic symptoms
Fear of losing control, going crazy or dying
Miscellaneous symptoms
Chills or hot flushes

A diagnosis of panic disorder is made if the patient has experienced recurrent, unexpected panic attacks and shows at least one of the following characteristics: (1) persistent concern about having another attack (anticipatory anxiety); (2) worry about the implications of an attack or its consequences (e.g., suffering a catastrophic medical or mental consequence), or (3) a significant change in behavior related to the attacks.

In clinical populations, panic disorder is usually accompanied by agoraphobia. Agoraphobia refers to avoidance behavior motivated by fear of having another panic attack. It may involve activities that patients fear could provoke an attack, situations where escape may not be readily available or routine activities during which patients are not accompanied by a “safe person” whom they believe could help in case of an attack. Table 2 lists common types of agoraphobic fear and avoidance.

Being far away from home
Being without the company of a “safe” person
Physical exertion that patients believe could provoke a panic attack
Going to places where escape is not readily available (e.g., restaurants, theaters, stores, public transportation)
Places where embarrassment could be a consequence of suffering a panic attack (e.g., social gatherings)
Ingesting substances that patients believe could provoke panic (e.g., foods, medicines, alcohol, caffeine)

Clinical Management of Panic Disorder

Patients presenting with panic-like symptoms should receive a thorough initial evaluation that goes beyond assessment of their primary somatic complaints. Areas of initial evaluation are outlined in Table 3. Several authors3,4 have recommended a specific work-up for these patients to reduce unnecessary assessments. Panic disorder can be treated effectively with pharmacotherapy, cognitive and cognitive-behavioral therapies or a combination of therapies.

Criteria for an unexpected panic attack
Agoraphobic avoidance
Use of caffeine and other anxiety-provoking substances
Substance-use history
Medical history to eliminate organic etiology
Psychiatric comorbidity (e.g., depression, interpersonal conflicts)
Previous assessments and treatments (psychiatric, medical)
Family history

The National Institutes of Health Consensus Development Conference on Treatment of Panic Disorder5 recommends that patients who are diagnosed with panic disorder should be provided with a description of indicated treatment options and the advantages and disadvantages of each option. Treatment selection should then be made with the patient's input and in consideration of the severity of the presenting complaints, and the patient's specific history and preferences. The following sections outline treatment options for patients with panic disorder and their known advantages and disadvantages. Considerations for selecting treatment also are presented.

Drug Therapy

Table 4 lists pharmacologic agents used to treat panic disorder and their common therapeutic dosage ranges.

DrugDosage range
Tricyclic antidepressants
Imipramine (Tofranil)50 to 300 mg per day
Clomipramine (Anafranil)25 to 250 mg per day
Nortriptyline (Pamelor)25 to 100 mg per day
Desipramine (Norpramin)25 to 300 mg per day
Fluoxetine (Prozac)20 to 80 mg per day
Paroxetine (Paxil)10 to 50 mg per day
Sertraline (Zoloft)50 to 200 mg per day
Fluvoxamine (Luvox)50 to 300 mg per day
Phenelzine (Nardil)45 to 90 mg per day
Tranylcypromine (Parnate)30 to 60 mg per day
Alprazolam (Xanax)2 to 10 mg per day
Lorazepam (Ativan)2 to 6 mg per day
Clonazepam (Klonopin)1 to 3 mg per day

Tricyclic Antidepressants

Imipramine (Tofranil) is the medication for panic disorder that has been most thoroughly studied, with at least 10 double-blind, placebo-controlled studies supporting its efficacy in the acute treatment of panic disorder.6 Clomipramine (Anafranil) has shown similar results in several double-blind trials as well. Other tricyclic antidepressants that have shown promise are listed in Table 4.

The onset of therapeutic action for tricyclic antidepressants typically takes three to four weeks. The average length of treatment is approximately six months but depends on several factors, including the efficiency with which panic suppression is achieved and agoraphobic avoidance, if any, is overcome. In obtaining an optimal response, the physician may find it helpful to assess plasma levels. For example, a therapeutic response should be evident at a level greater than 150 ng per mL (imipramine and desipramine [Norpramin] combined) in patients receiving imipramine.

Approximately one fourth of patients cannot tolerate the side effects of tricyclic antidepressants. Side effects are commonly anti-cholinergic (constipation, dry mouth, blurred vision and urinary retention), histaminergic (sedation and weight gain) or adrenergic (orthostatic hypotension).

One of the most burdensome adverse effects for patients with panic disorder, who often fear their own bodily sensations, is the “activation syndrome” that occurs on initial titration in approximately 25 to 40 percent of patients. The syndrome often can be mitigated by education, reassurances and initiating a low starting dosage (e.g., 10 mg of imipramine per day), then increasing gradually and flexibly at a rate of approximately 10 mg every two to three days until a dosage of 50 to 75 mg is achieved. An increment of 25 mg every two to four days from that point is usually well tolerated.

Since patients with panic disorder are often very sensitive to side effect symptoms, they may need more reassurance throughout pharmacotherapy than other patients. Physicians should also be aware that a withdrawal syndrome following abrupt cessation of these agents has been described.7

Imipramine and clomipramine are considered first-line treatment options for panic disorder. Some advantages and disadvantages of these agents are listed in Table 5.

Antidepressant effect+++
Generic form available+++
Established efficacy++/−++
Favorable side-effect profile+++
Not habit forming+++N/A
Maintenance of gains+
Potential toxicity+
Weight gain++
Sexual dysfunction++++
Anticholinergic effects+
Orthostatic hypotension++
Cardiovascular effects+
Withdrawal syndrome+/−+/−+
Multiple dosing/sessions+++
Delayed onset++++
Limited availability+
Dietary restrictions+

Selective Serotonin Reuptake Inhibitors

Although clinical trials have demonstrated the effectiveness of selective serotonin reuptake inhibitors (SSRIs) in treating depression, initial acceptance of these agents for treating panic disorder preceded well-designed studies that supported their efficacy.6 Controlled studies are confirming that initial enthusiasm.

Fluvoxamine (Luvox) has shown strong improvement rates in several double-blind and placebo-controlled studies of patients requiring acute treatment.811 Initial controlled studies of other SSRIs such as fluoxetine (Prozac), sertraline (Zoloft) and paroxetine (Paxil) are showing promise as well.12

Fewer patients drop out of SSRI therapy than tricyclic antidepressant therapy, suggesting that the SSRIs are slightly better tolerated than the tricyclics. Common side effects of SSRIs include sleep disturbance, headaches, gastrointestinal problems and sexual dysfunction. As with tricyclic antidepressants, beginning with a low starting dosage (e.g., 10 mg fluoxetine per day) and then gradually titrating the dosage upward to therapeutic levels may help minimize adverse effects. A withdrawal reaction has occasionally been described with abrupt cessation of SSRI therapy.

The SSRIs are considered appropriate first-line treatment for panic disorder, especially in patients with comorbid depression.6 Their advantages and disadvantages are summarized in Table 5.

Monoamine Oxidase Inhibitors

The monoamine oxidase inhibitors (MAOIs) are known for their effectiveness in treating atypical depression and social phobia, but they also have shown benefit in treating anxiety states, including panic disorder. Phenelzine (Nardil), in particular, has been proved efficacious in both controlled and open trials.13,14 Reversible inhibitors of monoamine oxidase A-enzyme such as moclobemide and brofaromine (which require no dietary restrictions) have shown promise in initial clinical trials overseas but are not available in the United States.6

Side effects of MAOIs include orthostatic hypotension, weight gain, sexual dysfunction and insomnia. When taking any nonspecific irreversible MAOI, patients must maintain a restrictive tyramine-free diet, and hypertensive crisis is risked if adherence to that diet is not maintained. The MAOIs also introduce a risk for serious drug-drug interactions (e.g., sympathomimetic drugs and meperidine [Demerol]). These risks lead many patients to refuse treatment with MAOIs, and many physicians reserve MAOIs for use in patients who do not respond to other therapies.

Although MAOIs are not regarded as a first-line treatment for panic disorder, they are considered appropriate therapy for patients who do not respond to other first-line agents like tricyclic antidepressants or SSRIs, and for patients with panic disorder accompanied by atypical depression or comorbid social anxiety.6 Advantages and disadvantages of MAOIs are summarized in Table 5.


Large-scale, controlled outcome studies have shown that benzodiazepines are clinically effective in the treatment of panic disorder.15,16 A unique advantage of the benzodiazepines is their quick onset of action relative to alternative agents (i.e., hours versus weeks), making them the only pharmacotherapeutic option for managing acutely distressed patients. Another advantage of benzodiazepines may be their broader spectrum of anxiolytic action, which extends beyond the suppression of panic attacks to amelioration of generalized anxiety.17 Benzodiazepines show the most tolerable side effect profile (most commonly sedation) of the antipanic medications, which may account for the routinely lower dropout rate compared with that of the antidepressants.

The principal drawback of benzodiazepines, particularly short-acting medications such as alprazolam (Xanax), involves their ability to produce physical dependency,18 manifested by a withdrawal syndrome on abrupt discontinuation. Even with gradual tapering, it may be difficult for some patients to discontinue benzodiazepine therapy.19 The benzodiazepines have also been known to produce interdose rebound anxiety requiring multiple daily dosing.

Although benzodiazepines are considered an appropriate first-line treatment in certain cases of panic disorder (e.g., when rapid symptom relief is needed), discontinuation difficulties have relegated these medications to a second- or third-line consideration. The most common use for benzodiazepines is to stabilize severe initial symptoms until another treatment (e.g., an SSRI or cognitive behavioral therapy) becomes effective.

Benzodiazepines are not indicated for use in patients who have a history of substance abuse or dependence, or as a first-line, sole intervention in patients with comorbid depression. Advantages and disadvantages of benzodiazepines in the treatment of panic disorder are summarized in Table 5.

Other Agents

Several other agents have been studied and have shown poor to mixed results or are undergoing empiric study for the treatment of panic disorder. Table 62027 lists these agents and briefly summarizes the evidence for their efficacy.

Propranolol (Inderal)Studies in patients with panic disorder have shown mixed results. Recommendations for its use as a sole treatment for panic disorder are controversial.20 Limited usefulness. May provide adjunctive relief of severe manifestations of catecholamine hyperactivity in some patients.
Nefazodone (Serzone)An open-label, 8-week trial of nefazodone in 14 patients showed efficacy and tolerability in patients with panic disorder and panic disorder with comorbid depression or depressive symptoms.21 Promising initial support warranting further study.
Buspirone (BuSpar)Two open and two controlled trials conducted. The open trials produced conflicting results. The controlled studies showed promise but have been criticized for methodologic problems.22 Efficacy of buspirone in panic disorder remains unresolved.
Venlafaxine (Effexor)One report of patients treated with low doses of venlafaxine showed favorable results.Very preliminary support. Needs further study.
InositolA controlled study (4-week, double-blind, crossover design) of inositol, a glucose isomer, found it more effective than placebo in reducing frequency and severity of panic attacks and severity of agoraphobia.23 Promising initial support warranting further study.
Trazodone (Desyrel)One study found that trazodone was less effective and less well tolerated than alprazolam (Xanax) or imipramine (Tofranil).24 Another study suggested some promise.25 Mixed results. Needs further study.
Verapamil (Calan, Isoptin)Verapamil has shown statistically significant but clinically modest improvements in patients with panic disorder.26 Initial support for limited effectiveness. Needs further study.
AnticonvulsantsSome emerging data support the use of anticonvulsants like valproic acid (Depakene) in the treatment of panic disorder.27 Promising initial support warranting further study.

Comparative Outcome of Pharmacotherapy

Not all agents indicated for the treatment of panic disorder have been subjected to head-to-head comparison. Meta-analytic techniques, which make use of standardized, “effect size” scores, allow comparisons of acute treatment response on relevant outcome measures across studies. A recent meta-analysis28 of 32 randomized, prospective, double-blind, placebo-controlled studies of imipramine, clomipramine, alprazolam, fluvoxamine, paroxetine and zimelidine found that all of these agents have proved to be superior to placebo. The SSRIs produced an effect size score that was significantly superior to that of imipramine and alprazolam. A trend favoring alprazolam over imipramine was also evident, although the trend did not reach statistical significance. These and other data have led some investigators12 to suggest that SSRIs are emerging as the drugs of first choice in the treatment of panic disorder.

Cognitive and Cognitive-Behavioral Treatments

Contrary to common clinical practice, existing evidence does not indicate that general, supportive psychotherapy used alone is an appropriate intervention in the treatment of panic disorder. Although a particular form of psychotherapy called emotion-focused treatment has shown initial promise in a recent empiric study,29 it warrants further, controlled study before it can be recommended as an evidence-based treatment option for panic disorder.

Cognitive30 and cognitive-behavioral therapies31 have received strong empiric support through numerous controlled clinical trials and are the psychotherapeutic treatments of choice for patients with panic disorder. With few exceptions, acute treatment improvement rates associated with these therapies range from 80 to 90 percent of patients.32

The major components of cognitive-behavioral therapies are outlined in Table 7.

Education about the onset and development of panic disorder and agoraphobia
Training in symptom management skills for anxiety, including training in relaxation methods and diaphragmatic breathing
Cognitive restructuring to change thought processes that initiate and maintain panic attacks and agoraphobic fears and avoidance behavior
Exposure simulations involving the attempted induction of bodily sensations feared by patients with panic disorder
Situational exposure aimed at eliminating agoraphobic avoidance behavior

Earlier behavior treatments tended to emphasize situational exposure aimed at reducing agoraphobic avoidance, rather than the panic attacks themselves. Newer treatments target both. Recent meta-analyses33 of cognitive-behavioral therapy quote an effect size (0.68) that compares favorably with those of pharmacotherapy (0.47) and combined treatments (0.56). Although various approaches to cognitive-behavioral therapy are available, those involving cognitive restructuring and exposure in vivo to feared stimuli have yielded the strongest effect sizes (effect size: 0.88) in meta-analyses.33

A recent review of long-term studies reported that approximately 75 percent of treated patients remained improved years after treatment had ended.34

Advantages and disadvantages of cognitive-behavioral therapy are summarized in Table 5.

Treatment Selection

Debate continues about whether panic disorder should be treated initially with cognitive therapy or cognitive-behavioral therapy, pharmacotherapy or a combined approach. Although acute treatment effect sizes may vary between treatment options, physicians and patients must consider several factors that go beyond success rates in acute treatment when selecting treatment. These decisions involve weighing the advantages and disadvantages of each treatment option and how well the options match the patient's presentation, preferences, and personal and financial resources. Although treatment selection guidelines vary, some considerations can be offered.

Regarding the choice of pharmacologic treatment, an evidenced-based approach suggests that the SSRIs are an appropriate first consideration. Although tricyclic antidepressants show similar success rates for acute treatment, their side-effect burden has been greater than that of SSRIs. A high-potency benzodiazepine given at the minimum therapeutic dose may be a useful adjunct to antidepressant therapy if prompt relief is indicated. However, the treatment plan should include discontinuing the benzodiazepine when the antidepressant's maximal effects are expected (i.e., approximately four to six weeks), to minimize the risks of discontinuation difficulties associated with the benzodiazepines. Although benzodiazepines are considered an appropriate intervention after SSRIs and tricyclic antidepressants have failed, they should not be prescribed if a history of or current comorbid substance abuse is suspected or if the patient shows comorbid depression. MAOIs are also an appropriate consideration when comorbid depression or social phobia is evident, although they remain a second choice given the risks they pose.

Cognitive-behavioral therapy is an appropriate first-line consideration in patients with mild to moderate panic disorder or panic disorder with situational avoidance. Although benzodiazepines can be combined temporarily with cognitive-behavioral therapy for prompt relief of severe symptoms, they have been known to interfere with cognitive-behavioral therapy, so their adjunctive use should be minimized. In situations where severe agoraphobic avoidance precludes participation in cognitive-behavioral therapy, consideration should be given to the combined use of SSRI or tricyclic antidepressant pharmacotherapy with cognitive-behavioral therapy. Cognitive-behavioral therapy has also been effective in patients who do not respond to pharmacotherapy and can be used in this role as well.

Acute relapse is common when pharmacotherapy for panic disorder is discontinued.35 Since recent studies have shown that cognitive-behavioral therapy can effectively reduce discontinuation and relapse difficulties associated with the discontinuation of pharmacotherapy, its adjunctive use with any pharmacotherapy should be considered to help reduce these risks. Other considerations for selecting a first-line treatment include the patient's preference for an approach that includes medication versus one that does not, as well as the availability of cognitive-behavioral treatment in the community.

Once treatment is selected, patients should be monitored periodically. When stabilized, patients should be encouraged to reenter previously avoided situations gradually, regardless of the treatment approach being used. If the treatment response is inadequate after approximately eight weeks of therapy, alternatives should be reconsidered.

Finally, patients with panic disorder often need sensitive clinical management. Many of these patients have been ill for several years and tend to have a history of varied, ineffective and failed treatments. Establishing a therapeutic alliance with patients, as described in Table 8, is an important aspect of any treatment selected.

Establish a therapeutic relationship by providing the following: respectful attention to patient's concerns, realistic reassurance and instillation of hope, and a willingness to be available in case of unexpected problems.
Provide the patient with some knowledge about panic disorder (e.g., it is common, it is not life-threatening, it is treatable).
Discuss the components of panic disorder (panic attacks, anticipatory anxiety and avoidance) and how treatment targets changes in each component.
Explore and discuss patient's concerns about medications.
Explain treatment options and their advantages and disadvantages.

Continue Reading

More in AFP

More in PubMed

Copyright © 1998 by the American Academy of Family Physicians.

This content is owned by the AAFP. A person viewing it online may make one printout of the material and may use that printout only for his or her personal, non-commercial reference. This material may not otherwise be downloaded, copied, printed, stored, transmitted or reproduced in any medium, whether now known or later invented, except as authorized in writing by the AAFP.  See permissions for copyright questions and/or permission requests.