Chapter 6 Antianxiety Agents

 

 

Chapter 6. Antianxiety Agents

INTRODUCTION

Anxiolytic agents—usually defined in the past as chiefly the benzodiazepines—are the most

commonly used psychotropic drugs. The vast majority of prescriptions for these medications are

issued by internists, family practitioners, and obstetricians. Psychiatrists write less than 20% of

the prescriptions for anxiolytics in this country, reflecting, in part, the fact that most anxious

patients never see psychiatrists. Moreover, anxiolytics are prescribed for a wide variety of patients

who do not have a primary anxiety disorder—namely, patients who present to primary care

physicians with somatic complaints or true somatic disease.

Antianxiety agents may be divided into many subclasses, of which the benzodiazepines are the

most frequently prescribed. Several of the subclasses of anxiolytics (e.g., benzodiazepines) include

agents marketed primarily as hypnotics (e.g., flurazepam). In this manual, we have separated the

pharmacological treatments of anxiety from those of insomnia. The distinction, however, is rather

artificial, because almost any sedative or antianxiety drug can be used at a low dose in the daytime

for anxiety and at a similar or higher dose for difficulty in sleeping.

The first major anxiolytic group, the barbiturates, were developed as sedative-hypnotic and

antiepileptic agents and were first introduced in the early 1900s. These drugs are also discussed in

the chapter on hypnotics (see Chapter 7: “Hypnotics”). Meprobamate, a carbamate derivative, was

introduced almost 60 years later as a sedative-anxiolytic agent. Although use of the two

classes—barbiturates and carbamates—has waned in recent decades, they are still more commonly

prescribed than would be imagined; in the late 1980s, meprobamate and phenobarbital represented

approximately 7% of the anxiolytic market.

Benzodiazepines, introduced in the early 1960s, dramatically changed the pharmacological

approach to anxiety. First developed as muscle relaxants, their anxiolytic-hypnotic properties,

wider safety margin in overdose, and potential to elicit physical dependence quickly became

apparent. Buspirone, a serotonin1A (5-HT1A) agonist with some mixed dopaminergic effects, was

released in the United States in 1987 for use in anxiety. Its use by psychiatrists in the treatment of

anxiety and related conditions was less than its use in primary care and nursing home settings (see

“Buspirone” section in this chapter; Cole and Yonkers 1995).

Increasingly, anticonvulsants are being used in the treatment of anxiety states. Agents such as

gabapentin and pregabalin may be alternatives or adjuncts to the more commonly used

antidepressants and benzodiazepines in treating some anxiety disorders.

Less widely used pharmacological approaches to anxiety include antihistamines and autonomic

agents (e.g., -blockers). The former have primarily a general sedative action; the latter, which are

more commonly used than antihistamines, act by blocking peripheral or central noradrenergic

activity and many of the manifestations of anxiety (e.g., tremor, palpitations, sweating). Several of

the phenothiazines also have indications in anxiety, although in the United States they have

become less widely used in recent years for this purpose; the atypical antipsychotics also appear to

have antianxiety effects. Use of the atypical antipsychotics for some anxiety disorders, such as in

the augmentation of treatment-resistant obsessive-compulsive disorder (OCD), generalized anxiety

disorder (GAD), and panic disorder, is now growing, since these agents appear to have a lower risk

of eliciting tardive dyskinesia.

Many newer antidepressants (mainly selective serotonin reuptake inhibitors [SSRIs] and

serotonin-norepinephrine reuptake inhibitors (SNRIs) such as venlafaxine) have taken center stagePrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

2 of 35

31/01/2009 06:51

in the treatment of the whole range of anxiety disorders but not in the treatment of insomnia.

However, mirtazapine and trazodone are commonly used as hypnotic agents. Clomipramine is the

only tricyclic antidepressant (TCA) shown to be effective in the treatment of OCD. However, all the

SSRIs are also presumably effective treatments for OCD (see “Obsessive-Compulsive Disorder”

section later in this chapter). The SSRIs and related or unrelated newer drugs (e.g., gabapentin,

venlafaxine) may already have become the primary drugs used by psychiatrists for the treatment of

specific anxiety disorders, whereas benzodiazepines may still be the first drugs prescribed by

primary care physicians.

In addition to this major shift in the patterns of treatment of anxiety disorders, there has been a

major upsurge of interest in cognitive-behavioral therapy (CBT), whose efficacy has been shown in

well-designed studies, mainly done by psychologists. Specific or semispecific programs have been

designed to address the symptoms and treatment requirements of individual anxiety disorders.

Almost all these programs, however, have elements of desensitization, exposure, and cognitive

restructuring and include having the patient demonstrate new learning in real-life situations.

As a very general overview, benzodiazepines often work fastest in relieving symptoms,

antidepressants (SSRIs and newer agents) take several weeks, and CBT may take 3 months or

longer. There is some evidence that patients who show improvement while participating in CBT

programs maintain the improvement longer after treatment is stopped than do patients who show

improvement while receiving drug therapies (Barlow et al. 2000).

We begin the chapter by discussing the use of benzodiazepines in treating general and panic

anxiety. We then address the use of antidepressants in treating other anxiety disorders, such as

posttraumatic stress disorder (PTSD), social phobia, and OCD, and body dysmorphic disorder.

Finally, we discuss the use of other classes of medication for anxiety and other conditions, such as

catatonia, a syndrome uniquely responsive to sedative drugs and electroconvulsive therapy (ECT).

Since the SSRIs have already been considered in some detail in Chapter 3 (“Antidepressants”), we

give them less detailed attention here. Several of the SSRIs and venlafaxine have been approved by

the U.S. Food and Drug Administration (FDA) for one or more specific anxiety diagnoses, including

paroxetine for GAD, PTSD, OCD, panic disorder, premenstrual dysphoric disorder (PMDD)

(controlled-release form only), and social phobia; fluoxetine for OCD, bulimia, panic disorder, and

PMDD; sertraline for OCD, panic disorder, PMDD, social anxiety disorder, and PTSD; escitalopram

for GAD; and venlafaxine (extended-release form) for GAD and social anxiety disorder). It is our

position that until studies clearly show differences in efficacy between these drugs in the treatment

of specific anxiety disorders, all SSRIs and venlafaxine are probably reasonably effective across the

entire range of anxiety disorders. Their use in such conditions may require adjustments, but these

apply to the specific disorder and the whole class of SSRIs. For example, use very low dosages

(particularly initially) in patients with panic disorder with agoraphobia; use higher dosages and

wait even longer for clinical response in patients with OCD.

BENZODIAZEPINES

Indications

In addition to anxiety, benzodiazepines are indicated for muscle tension, insomnia, status

epilepticus (diazepam), myoclonic epilepsy (clonazepam), preoperative anesthesia, and alcohol

withdrawal. One benzodiazepine, the triazolobenzodiazepine alprazolam, is also indicated for

anxiety associated with depression (as is lorazepam), and some studies have shown that

alprazolam also parallels imipramine and phenelzine in having both antipanic and antidepressant

properties (see Chapter 3: “Antidepressants”). Clonazepam has also been shown to have some

antipanic effects, as has lorazepam.

Benzodiazepines (e.g., diazepam, clonazepam, alprazolam): overview

Efficacy Generalized anxiety (FDA approved)

Panic disorder (FDA approved for alprazolam)Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

3 of 35

31/01/2009 06:51

Benzodiazepines (e.g., diazepam, clonazepam, alprazolam): overview

Insomnia (FDA approved)

Seizure disorder (FDA approved for clonazepam)

Muscle relaxation

Anesthesia

Side effects Sedation

Lethargy

Dependence/Withdrawal

Safety in

overdose

Safe in overdose up to 30 times the normal daily dose. Usual symptoms of overdose

include sedation, drowsiness, ataxia, and slurred speech. May result in respiratory

depression in combination with other CNS depressants. Management includes gastric

lavage, forced emesis, and assisted ventilation.

Dosage and

administration

Varies by benzodiazepine and indication; see Table 6–1.

Discontinuation Taper by no more than 25% of total dose per week after long-term administration.

Withdrawal includes insomnia, agitation, anxiety, and, rarely, seizures.

Drug interactions Additive CNS depression with ethanol, barbiturates, and other CNS depressants

Drugs that triazolo-benzodiazepine levels include:cytochrome P450 3A4 inhibitors,

ketoconazole, fluconazole, nefazodone

Drugs that triazolo-benzodiazepine levels include:carbamazepine

Note. CNS = central nervous system; FDA = U.S. Food and Drug Administration.

Probably all currently available benzodiazepines are useful in treating both chronic anxiety and

anxiety secondary to life stresses or medical conditions. The definition of GAD in DSM-IV-TR

(American Psychiatric Association 2000) is probably too restrictive to cover all the forms of anxiety

for which benzodiazepines can be helpful. It is even likely that there are “double anxiety”

disorders, analogous to the concept of double depression; some patients have lifelong mild to

moderate anxiety symptoms with episodic periods of worsening during which they seek therapy.

Panic disorder with or without agoraphobia is a chronic, fluctuating condition; some patients

experience episodes of illness (like depressive episodes), whereas other patients have mild,

infrequent attacks during some life phases and incapacitating symptoms at other times. Alprazolam

is the only benzodiazepine officially deemed effective and well studied in the treatment of panic

disorder with or without agoraphobia, although other benzodiazepines may also be effective.

Both panic and GAD are conditions often accompanying other disorders, such as major depression,

PTSD, and borderline personality disorder, as well as other specific anxiety disorders (e.g., social

phobia).

It is likely that patients seen in primary care settings may show even more complex mixtures of

milder anxiety and depressive symptoms than patients seen by psychiatrists. The old-fashioned

wastebasket diagnosis of mixed anxiety and depression may still deserve recognition because of

the prevalence of such a presentation, even though recent diagnostic systems (DSM-III-R

[American Psychiatric Association 1987] and DSM-IV [American Psychiatric Association 1994])

have not fully endorsed the condition. Such disorders could explain why antidepressants and

antianxiety agents often work in the same types of patients (Rickels and Schweizer 1995).

Clonazepam has been reported to speed response in patients with major depression treated with

fluoxetine (Smith et al. 1998) and in panic disorder patients treated with sertraline (Goddard et al.

2001). The drug appeared to have a calming effect and to offset any anxiogenic effect associated

with initiation of the SSRI. The benzodiazepine was used for brief periods (about 3 weeks) at doses

of 0.5–1.5 mg hs and was then discontinued. Such uses are discussed in Chapter 9 (“Augmentation

Strategies for Treatment-Resistant Disorders”).Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

4 of 35

31/01/2009 06:51

Pharmacological Effects

In recent years there has been considerable attention to the mode of action of benzodiazepines,

spurred on by the identification of specific receptor sites. These sites, found in various brain

regions, are coupled to -aminobutyric acid (GABA) receptors. This receptor complex appears to

mediate the anxiolytic, sedative, and anticonvulsant actions of the benzodiazepines. The location of

specific receptors may be related to the relative anticonvulsant, anxiolytic, or sedative properties of

the various benzodiazepines.

Some pharmacologists have hypothesized that it may be possible to develop new compounds that

either bind more specifically to certain receptors or act as partial agonists to produce anxiolysis

without sedation. These approaches are being explored. Thus far, drugs specifically binding to the

benzodiazepine1 receptor (e.g., zolpidem) do not seem particularly unique. Partial agonists or

drugs with more specific binding could substantially reduce the risk for tolerance, dependence, and

withdrawal effects. Unfortunately, neither partial agonists nor more uniquely binding drugs appear

so far to be different from available benzodiazepines.

The triazolobenzodiazepine alprazolam appears to also have effects on noradrenergic systems,

causing downregulation of postsynaptic -adrenergic receptors in reserpine-treated mice and

increasing the activity of the N protein in humans (the protein that couples the postsynaptic

receptor to the intraneuronal energy system). These effects may help to explain the drug’s

antipanic and moderate antidepressant effects beyond the effects mediated by the

benzodiazepine-GABA receptor complex.

Adinazolam, a related drug now unlikely to become available, appears to enjoy more pronounced

effects on noradrenergic and probably serotonergic systems than does alprazolam (see Chapter 3:

“Antidepressants”). These (and probably other) benzodiazepines may also exert downregulating

effects on corticotropin-releasing factor (CRF), a peptide that initiates the

hypothalamic-pituitary-adrenal (HPA) axis stress response and may affect central

catecholaminergic systems as well. Thus, some benzodiazepines exert extremely complicated

neurochemical effects.

Subclasses

The anxiolytic benzodiazepines are commonly divided into three subclasses on the basis of

structure: 2-keto (chlordiazepoxide, clonazepam, clorazepate, diazepam, halazepam, prazepam,

and the hypnotic flurazepam); 3-hydroxy (lorazepam, oxazepam, and the hypnotic temazepam);

and triazolo (alprazolam, adinazolam, estazolam, and the hypnotic triazolam) (see Figure 6–1 and

Table 6–1).

Figure 6–1.Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

5 of 35

31/01/2009 06:51

Chemical structures of anxiolytic benzodiazepines.

Table 6–1. Benzodiazepines: names, formulations and strengths, and anxiolytic dosage range

Generic name

Brand namea

Formulations and strengths Anxiolytic dosage

range (mg/day)b

2-Keto

chlordiazepoxide Librium Capsules: 5, 10, 25 mg

Powder for injection: 100-mg ampule with 2-mL

diluent

15–40

50–100 im

clorazepate Tranxene Tablets: 3.75, 7.5, 15 mg 15–40

Tranxene-SD

(single-dose)

Tablets: 11.25, 22.5 mg 11.25–45

diazepam Valium Tablets: 2, 5, 10 mg

Oral solution: 5 mg/5 mL (30-mL)

Injection: 5 mg/mL (2-mL prefilled syringe;

2-mL ampule; 1-mL, 2-mL, 10-mL vials)

5–40

clonazepam Klonopin Tablets: 0.5, 1, 2 mg

Wafers: 0.125, 0.25, 0.5, 1, 2 mg

3-Hydroxy

lorazepam Ativan Tablets: 0.5, 1, 2 mg

Oral solution: 2 mg/mL (30-mL)

1–6Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

6 of 35

31/01/2009 06:51

Generic name

Brand namea

Formulations and strengths Anxiolytic dosage

range (mg/day)b

Injection: 2 mg/mL, 4 mg/mL (both in 1-mL

prefilled syringe and single-dose vial and 10-mL

multidose vial)

1–2

oxazepam Serax Capsules: 10, 15, 30 mg 15–120

Triazolo

alprazolam Xanax Tablets: 0.25, 0.5, 1, 2 mg

Disintegrating tablets: 0.25, 0.5, 1, 2 mg

Oral solution: 1 mg/mL (30-mL)

1–4

alprazolam XR Xanax XR 1, 2, 3 mg

aExcept for Tranxene-SD, the benzodiazepines shown are available in generic form.

bApproximate dosage ranges. Some patients will require higher dosages; others may respond to dosages

below the range.

The pharmacokinetic properties (i.e., half-lives) vary among these classes, in part reflecting

differences in their modes of drug metabolism, as summarized in Table 6–2. The 2-keto drugs and

their active metabolites are oxidized in the liver, and because this process is relatively slow, these

compounds have relatively long half-lives. For example, the half-life of diazepam is approximately

40 hours. One active metabolite (desmethyldiazepam) has an even longer half-life (about 60

hours). Moreover, because desmethyldiazepam is further metabolized to oxazepam, which is also

active as an anxiolytic (Table 6–1), diazepam imparts long-range sedative and anxiolytic effects.

The half-life of clonazepam is approximately 40 hours. Many of the marketed 2-keto drugs are

prodrugs—they are themselves inactive but eventually form active metabolites. Thus, prazepam,

clorazepate, and halazepam are mere precursors to desmethyldiazepam, as is diazepam.

Differences among these specific 2-keto compounds revolve around the rates of absorption and the

specific active metabolites formed.

Table 6–2. Benzodiazepines: absorption and pharmacokinetics

Generic name Oral absorption Major active components Approximatehalf-life (hours)a

2-Keto

chlordiazepoxide Intermediate chlordiazepoxide 20

desmethylchlordiazepoxide 30

demoxepam Unknown

desmethyldiazepam 60

clorazepate Fast desmethyldiazepam 60

diazepam Fast diazepam 40

desmethyldiazepam 60

methyloxazepam 10

halazepam Intermediate desmethyldiazepam 60

prazepam Slow desmethyldiazepam 60

3-Hydroxy

lorazepam Intermediate lorazepam 14

oxazepam Slow to intermediate oxazepam

9

TriazoloPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

7 of 35

31/01/2009 06:51

Generic name Oral absorption Major active components Approximatehalf-life (hours)a

alprazolam Intermediate alprazolam 14

alprazolam XR

aBased on ranges of half-lives reported in young, psychiatrically and physically healthy volunteers.

In contrast, the 3-hydroxy compounds are metabolized via direct conjugation with a glucuronide

radical, a process that is more rapid than oxidation and does not involve the formation of active

metabolites. The two major examples of this subclass are oxazepam and lorazepam, which have

considerably shorter half-lives (9 and 14 hours, respectively) than do their 2-keto counterparts.

Similarly, the hypnotic temazepam has a half-life (8 hours) that is much shorter than flurazepam’s.

The triazolo compounds are also oxidized; however, they appear to have more limited active

metabolites and thus relatively shorter half-lives. The half-life of alprazolam is about 14 hours;

adinazolam, 2 hours; and N-desmethyladinazolam (adinazolam’s active metabolite), 4 hours; the

half-life of the hypnotic triazolam is 3–4 hours.

The pharmacokinetic properties of benzodiazepines that are oxidized in the liver may be affected

by other medications. Of particular note, nefazodone, fluoxetine, fluvoxamine, sertraline,

cimetidine (Tagamet), and contraceptive pills inhibit the liver oxidative enzyme cytochrome P450

3A3/4 and thus slow the degradation of the 2-keto and triazolo compounds. Clinicians should keep

this in mind in treating anxious patients who are also taking these drugs. Fluoxetine’s effects on

alprazolam metabolism have not appeared to be clinically meaningful.

Other differences among benzodiazepines revolve around their rates of absorption and distribution.

For example, although prazepam and clorazepate are similar in structure and both are prodrugs of

desmethyldiazepam, the two differ in terms of the metabolic processes required for absorption and,

thus, in the rates at which they appear in blood (Table 6–2). Clorazepate and diazepam are rapidly

absorbed and produce peaks in plasma levels more quickly than does prazepam, whose absorption

is mediated via slower processes. Halazepam’s conversion to desmethyldiazepam is even slower.

The lipophilic and hydrophilic properties of these drugs also vary, resulting in pronounced

differences in how quickly they work and for how long. Drugs that are more lipophilic (e.g.,

diazepam) enter the brain more quickly, “turning on” the effect promptly, but “turning off” the

effect more quickly as well as they disappear into body fat. Less lipophilic compounds (e.g.,

lorazepam) produce clinical effects more slowly but may provide more sustained relief. These

properties are largely independent of pharmacokinetics. Some drugs with long half-lives (e.g.,

diazepam) can also be highly lipophilic, providing rapid relief but for shorter periods than one

might predict from half-life data alone. In contrast, lorazepam is less lipophilic and turns on and off

more slowly, potentially providing more sustained effects, despite its shorter half-life compared

with diazepam. In short, traditional half-life pharmacokinetics can be misleading and tell only a

part of the story of how drugs work.

In addition, investigators have begun to pay more attention to relative receptor affinity, a property

that may play a more important role than was previously thought in determining the duration of

action. High-potency benzodiazepines, such as lorazepam and alprazolam, may have such high

receptor affinity that withdrawal symptoms may be far more intense than might be expected from

inspecting other variables such as half-lives. Interestingly, oxazepam, which is similar in lipid

solubility and half-life to lorazepam, appears to produce fewer withdrawal symptoms. This position

has been most eloquently stated by Lader (1982) in the United Kingdom. Unfortunately, there are

few data to confirm or refute this assertion.

Although several of the benzodiazepines are available for parenteral use (see Table 6–1), there is

wide variability in the absorption properties of these compounds when given intramuscularly. For

example, lorazepam is relatively rapidly absorbed when given intramuscularly. In contrast,

chlordiazepoxide and diazepam are slowly absorbed. Lorazepam has become increasingly popular

as an adjunctive treatment for agitation in acutely psychotic patients and also often relievesPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

8 of 35

31/01/2009 06:51

catatonic and depressive stupor. Oral concentrate forms of some benzodiazepines, such as

diazepam, lorazepam, and alprazolam, are available in the United States. In addition, lorazepam

and clonazepam tablets are given sublingually in some emergency room situations to promote rapid

absorption of the drug from the oral mucosa. Clonazepam is available as a rapidly disintegrating

sublingual wafer, and some anxious patients or patients with panic find the sublingual wafer quite

helpful. In one study alprazolam given by nasal spray at the onset of panic attacks proved effective

in aborting them. To our knowledge, this delivery route has not been studied further.

Dosage and Administration

The efficacy of benzodiazepines in treating patients with symptomatic anxiety or diagnosable

anxiety disorders has been established in double-blind, random-assignment comparisons with

placebo. When treating a GAD patient, the clinician should begin with a benzodiazepine (e.g.,

diazepam at approximately 2 mg tid, with increases as needed to a maximum regular daily dose of

40 mg). A modal dosage of diazepam for GAD is 15–20 mg/day. Chlordiazepoxide has a much wider

dosage range: the recommended starting dosage is 5–10 mg po tid, with a maximum of 60 mg/day

for anxiety. The dosage of chlordiazepoxide for acute alcohol withdrawal is much higher: 50–200

mg/day. Generally, clinicians prescribe chlordiazepoxide 25 mg every 1–2 hours until symptomatic

relief or sedation occurs, up to a maximum dosage of 200 mg/day. For lorazepam, the starting

dosage is 0.5 mg tid, with titration upward as needed to 6 mg/day. Higher daily doses are approved

but are frequently associated with intense sedation. Dosage ranges for the anxiolytic

benzodiazepines are listed in Table 6–1.

Clonazepam is generally started at 0.5–1 mg/day. Dosages up to 4 mg/day are sometimes needed

to control panic attacks, but most patients do well with 1-2 mg/day. Clonazepam wafers can also

be given in doses as small as 0.125 mg to treat an acute panic attack. The ideal dosing regimen,

however, prevents panic attacks and circumvents the need to use sublingual wafers. Clonazepam,

like other benzodiazepines, works much faster than antidepressants. We will often treat a panic

disorder for the first 4–6 weeks with a benzodiazepine such as clonazepam while simultaneously

starting an antidepressant. Studies have suggested that the addition of clonazepam to an SSRI

speeds up treatment effects in panic disorder (Pollack et al. 2003).

The use of alprazolam in panic disorder patients often requires higher dosages than those used in

GAD. Currently, alprazolam is approved in dosages up to 10 mg/day, but generally 4–5 mg/day or

less is used. In our early studies on depression, we used the much higher dosage regimen, but we

have been impressed that patients generally do not require more than 4 mg/day for a response,

and some are even oversedated at 2–3 mg/day. Because of concern about dependence, this drug

should be used at the lowest effective dose possible. There is evidence from controlled, fixed-dose

studies that blood levels of alprazolam in the range of 20–40 ng/mL are optimal for improvement in

patients with panic disorder. At higher plasma levels (40–60 ng/mL), some additional patients may

improve, but sedative-type side effects and ataxia increase (Greenblatt et al. 1993). Alprazolam is

available in an extended-release formulation (XR) for qd or bid dosing (Glue et al. 2006). The XR

form generally allows for once-a-day dosing and may mitigate some of the withdrawal associated

with missed doses. However, alprazolam XR can still result in withdrawal symptoms. Several

biotechnology start-up companies are attempting to develop more rapidly absorbed or even

longer-acting formulations.

The starting dosage of alprazolam in both GAD and panic disorder should be 1.5 mg/day or less,

given in divided doses, with a gradual increase in dosage as tolerated by the patient. In treating

patients with panic disorder, alprazolam dosage may be increased, to block not only panic attacks

but anticipatory anxiety as well. This often requires higher dosages (4–5 mg/day) in the first 6 or

more weeks. Over time, however, as patients overcome their anticipatory anxiety, the dosage can

be reduced to 2–3 mg/day for continued blocking of panic attacks. Although alprazolam was

thought to have unique antipanic properties, subsequent reports have indicated that lorazepam,

clonazepam, and diazepam may all be effective in ameliorating or preventing panic symptoms.

Clonazepam is used at dosages of 1–3 mg/day.Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

9 of 35

31/01/2009 06:51

Although TCAs, monoamine oxidase inhibitors (MAOIs), and SSRIs are all probably as effective as

alprazolam in treating patients with panic disorder with agoraphobia, only the benzodiazepines

provide rapid relief; the other drug groups take at least 4–6 weeks, compared with 1 week or less

for alprazolam. With the more conventional antidepressants, more patients drop out early in

treatment because of side effects, and there is a general belief that some panic patients are even

more sensitive to antidepressant drug side effects than are depressed patients. In patients who

have occasional bouts of moderate anxiety occurring only every few days or weeks,

benzodiazepines may be preferable to maintenance antidepressants, since they work as prn

medication. Diazepam’s ability to act rapidly without prolonged sedation makes it particularly

useful in such situations with patients not prone to drug abuse. Other benzodiazepines can also be

used in this manner, of course. Oxazepam’s slow absorption rate raises concern about its utility as

a prn medication. However, patient acceptance of oxazepam is fairly good, and its low abuse

liability makes it a reasonable choice for some patients.

In patients with long histories of a particular kind of panic disorder with agoraphobia, 6-month and

18-month courses of alprazolam therapy were studied in terms of relapse/recurrence rates after

the alprazolam was tapered and stopped; the longer period was found to be associated with

symptomatic relief (Ballenger et al. 1993).

One major area of debate revolves around how long to use these drugs for patients with significant

anxiety. For patients whose anxiety is very acute and related to specific stressors, use of these

agents should be directed at reduction of acute symptoms, and thus prolonged use beyond 1–2

weeks is generally not required or advised. In patients whose anxiety symptoms are of several

months’ or greater duration, we recommend treatment for 4–6 weeks at doses that provide relief,

then reduction of dosage to the minimum needed for maintenance for the next few months,

followed by discontinuation when possible. Patients who meet DSM-IV-TR criteria for GAD have by

definition an even more chronic condition and require even longer treatment (e.g., 4–6 months or

longer) before discontinuation is attempted. For these patients, SSRIs may be the preferred agent

to use first. Unfortunately, psychiatrists often first see patients with histories compatible with GAD

only after they have received benzodiazepines for years from other physicians; there are many

such patients who obtain relief from these drugs but who relapse when the drugs are stopped.

Further, because many patients seem to do well on reasonable dosages over longer periods, the

clinician may be faced with the difficult decision of how long to maintain use of the benzodiazepine.

This dilemma is intensified by the observation that tolerance can develop to some effects of

benzodiazepines (e.g., hypnosis), suggesting that the apparent relief experienced by patients could

reflect a nonspecific psychological effect.

Although tolerance can occur, it is our belief that most patients do not develop tolerance but are

still responding. We base this observation on the number of patients we have seen over the years

who have functioned well on a given daily dose of benzodiazepine and have not found themselves

escalating their total daily intake. Longer-term data from alprazolam studies indicate that panic

patients do not escalate their daily dosages but, rather, frequently lower them over time. There

does not appear to be a loss of efficacy of alprazolam in patients followed for up to 1 year. It is our

impression that animal and human models of tolerance may not be totally applicable to chronic

anxiety per se. Rather, such models emphasize self-administration of a drug or drug-induced ataxia

produced in “normal” specimens but do not fully take into account the biological and clinical status

of the anxious patient. If possible, the clinician should attempt gradually to taper benzodiazepines,

using psychotherapy, behavior therapy, or other drug therapies to help patients deal with their

anxiety (see “Withdrawal” subsection later in this chapter). Some patients, however, may require

continued benzodiazepine therapy. Overlapping an SSRI with the benzodiazepine for several

weeks, in the hope that the patient will feel still further relief of symptoms, before tapering the

benzodiazepine may be a useful strategy but has not been systematically studied.

True longer-term harmful effects of benzodiazepines have not been convincingly described. For

example, Lader (1982) reported computed tomography (CT) scan abnormalities in a series of

patients who had taken benzodiazepines for a long time. Although these observations could bePrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

10 of 35

31/01/2009 06:51

interpreted as indicating that these drugs produce organic or structural changes in brain tissue (as

chronic alcohol use does), an equally acceptable explanation is that some anxious patients who

require chronic treatment with benzodiazepines may have neuropsychiatric disorders, as evidenced

by CT scan abnormalities. A study by Lucki et al. (1986) of patients on chronic long-term

benzodiazepine treatment failed to show significant cognitive impairment on psychometric tests.

The most recent controlled study in this area failed to find significant long-term cognitive effects

for alprazolam XR in panic patients (Gladsjo et al. 2001). However, some investigators believe such

impairment can occur, particularly in elderly patients. Certainly elderly medical patients taking

long-acting benzodiazepines (or antidepressants) are at increased risk for falls leading to hip

fractures.

Are benzodiazepines addictive? Do they produce withdrawal symptoms? Studies in animals have

indicated that benzodiazepines can reinforce use and can produce physical dependence and

tolerance. Available survey and treatment facility data suggest that benzodiazepines are rarely

sought after or craved in the sense that heroin and cocaine are. Rather, they are used as part of a

polysubstance abuse pattern to modulate the effects of primary drugs of abuse (e.g., cocaine) or as

backup drugs when more euphoriant drugs are not available.

Risk factors for benzodiazepine abuse include a history of alcohol or other substance abuse and the

presence of a personality disorder. In patients with a history of substance abuse, benzodiazepines

should generally not be prescribed routinely. However, in some recreational drug users with

anxiety, they can be used, particularly if trials of other agents have failed. In patients with an Axis

II disorder, benzodiazepines should be administered only if necessary and for brief periods at low

doses. Benzodiazepine dependence is mainly, or at least partially, an iatrogenic problem in that

patients receive the medication from physicians originally for legitimate reasons but may then take

it for too long or at too large a dosage. Possible length of treatment with these drugs should be

thought out in advance of their prescription, and longer-term trials should be monitored carefully.

Given the current medicolegal climate and the existence of experts and medical boards devoted to

“pharmacologic Calvinism,” physicians planning to recommend that patients, particularly those

with a past history of substance abuse, continue taking benzodiazepines for long periods should get

outside consultation to help in confirming or changing their treatment plans.

Withdrawal

Should one withdraw benzodiazepines from patients who have taken them regularly over longer

periods? As a rule, this approach is sensible; reduction should be made at a maximum rate of

approximately 10% per day. In their classic study of benzodiazepine withdrawal, Rickels et al.

(1983) noted that when benzodiazepines were abruptly discontinued under double-blind

conditions, withdrawal symptoms were demonstrated in more patients who had been taking

benzodiazepines for more than 8 months (43%) than patients who had been taking them for

shorter periods (5%). In a subsequent study, this group reported similar rates of withdrawal

symptoms in patients who had been receiving maintenance therapy with clorazepate for 6 months

(Rickels et al. 1988). Patients who had received prolonged treatment with buspirone did not

experience withdrawal symptoms when the buspirone was discontinued. Pecknold et al. (1988)

reported that sudden discontinuation after an 8-week trial of alprazolam resulted in symptoms of

anxiety in about 35% of panic disorder patients. Some of these patients may have experienced

reemergence of their panic symptoms rather than withdrawal.

Common symptoms of benzodiazepine withdrawal include jitteriness, anxiety, palpitations,

clamminess, sweating, nausea, confusion, and heightened sensitivity to light and sound. Seizures

represent the most worrisome of withdrawal reactions, but fortunately they are generally rare. No

patients in the 1983 study by Rickels et al. experienced seizures. Seizures with abrupt diazepam

withdrawal occur about 5–7 days after the drug is stopped and not within 24 hours, reflecting the

long half-lives of both diazepam and desmethyldiazepam. With shorter-acting drugs (e.g.,

lorazepam and alprazolam), withdrawal symptoms emerge more rapidly—within 2–3 days. Thus,

with diazepam, physicians cannot be confident that seizures will not occur unless the patient hasPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

11 of 35

31/01/2009 06:51

stopped taking the drug for at least 1 week. Any signs of withdrawal (even at day 5) should be

reviewed carefully, and consideration should be given to reinstituting the drug and then

withdrawing it more gradually. A few days after discontinuation of benzodiazepines, some patients

reexperience their original anxiety symptoms, but in a more severe form (so-called rebound

anxiety). (In the case of hypnotics, this takes the form of rebound insomnia.) This syndrome is

generally transient, usually lasting 48–72 hours.

As Rickels and Schweizer (1995) suggested, withdrawal symptoms usually occur during

benzodiazepine tapering and in approximately the week following cessation and have passed after

the patient has been without benzodiazepines for 3 weeks. Symptoms of the preexisting anxiety

disorder usually reemerge more rapidly.

It is becoming increasingly clear that many patients stop adhering to withdrawal regimens for

benzodiazepines early in the process, well before appreciable withdrawal symptoms could occur

(see Rickels et al. 1999 and the entire supplement of the journal in which their article appears for

more details). This strongly suggests a psychological belief that the medications are necessary and

a strong overreaction to somatic symptoms related to anxiety. Not surprisingly, CBT of the kind

used in treating panic disorder with agoraphobia and instituted well before tapering of the

benzodiazepines can be very effective in facilitating tapering and allowing the patient to become

free of these drugs (Spiegel 1999).

In a follow-up of patients who had been in benzodiazepine tapering trials 2–3 years earlier, Rickels

et al. (1999) found patients who were still free of benzodiazepines to be less symptomatic than

patients who failed to complete the taper or had returned to benzodiazepine use. These data

unfortunately do not unscramble cause and effect. One wonders whether, in these days of powerful

advertising for the use of SSRIs in a variety of anxiety disorders and of a general feeling that

benzodiazepine use invites “addiction” (not just physical dependence), benzodiazepines are being

underutilized and whether some less safe drugs, such as meprobamate, glutethimide, or

barbiturates, may be inappropriately replacing them when trials with SSRIs fail. This may be the

case in some states (e.g., New York) where benzodiazepines require triplicate prescriptions.

Factors that make benzodiazepine withdrawal more difficult include higher daily dose, shorter

half-life, longer duration of prior benzodiazepine therapy, and more rapid taper. At the patient

level, a diagnosis of panic disorder, higher pretaper levels of anxiety or depression, more

personality disorder, and concomitant alcohol or substance abuse make tapering more difficult.

Successful tapering of benzodiazepines is often demanding of the clinician’s time and energy

(Rickels et al. 1999). The clinician needs to be available at all hours to provide counseling, support,

and reassurance.

Generally the first half of the benzodiazepine dose can be tapered over a 4-week period, but

tapering the remaining half of the dose can be a prolonged process. Patients may need to be kept at

the 50% dose for several months before further tapering is attempted. However, many patients

drop out of tapering attempts very early before any significant withdrawal symptoms occur, mainly

for personality reasons.

So far, adjunctive therapy with other nonbenzodiazepine drugs—carbamazepine, trazodone,

valproate, buspirone, and imipramine—has not been shown to be of help in decreasing withdrawal

symptoms, although some of these drugs—imipramine and valproate—may enable some patients to

remain free of benzodiazepines for a few weeks.

If a patient is to undertake benzodiazepine withdrawal with any real prospect of success,

prewithdrawal symptoms of anxiety and depression need to be actively treated with either

pharmacological or psychological therapies.

Panic control treatment (PCT), a type of CBT, is an educational-experiential approach aimed at

having patients learn to tolerate somatic symptoms of panic without undue anxiety. A controlled

study has shown that use of PCT in combination with a very slow and cautious benzodiazepine

taper (0.125 mg of alprazolam every 2 days for patients taking more than 1 mg/day initially, orPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

12 of 35

31/01/2009 06:51

0.25 mg every 8 days once the dosage has been reduced to 1 mg/day) is effective. In another

study, the taper was even slower. In both studies, PCT plus very slow taper was substantially

better than medical management plus slow taper, although the sample sizes were small. Most of

the patients whose benzodiazepine use had been successfully tapered with PCT were still free of

benzodiazepines 3 years later (Spiegel 1999). Some informal clinical experience suggests that CBT

should be continued for at least a few weeks after benzodiazepine taper is completed for better

long-term success.

As noted later in this chapter, other forms of CBT are now being successfully used in treating most

of the anxiety disorders and may well prove useful in the primary treatment of drug-free patients

as well as in assisting benzodiazepine withdrawal.

Side Effects

Compared with many other classes of psychotropic agents, benzodiazepines have relatively

favorable side-effect profiles. The most common side effect is sedation, which is in part related to

dose and can be managed by reducing it. Other effects include dizziness, weakness, ataxia,

anterograde amnesia (particularly with the short-acting benzodiazepines [e.g., triazolam]),

decreased motoric performance (e.g., driving), nausea, and slight hypotension. Falls in elderly

patients have been reported to be related to the use of longer-acting benzodiazepines as well as to

the use of antidepressants (see Chapter 12: “Pharmacotherapy in Special Situations”). In the

popular press, there have been reports of severe dyscontrol syndromes in patients taking certain

benzodiazepines, particularly triazolam. We personally have not encountered any such syndrome in

our clinical practice.

Overdose

Fortunately, these drugs have a relatively wide safety margin, and deaths due to benzodiazepine

ingestion alone are rare. Most deaths that have involved these drugs have been associated with

concomitant ingestion of other agents (e.g., alcohol or TCAs).

ANTIDEPRESSANTS

Since the SSRIs have already been considered in some detail in Chapter 3 (“Antidepressants”), we

give them less detailed attention here. Although several of the SSRIs have been FDA approved for

one or more specific anxiety diagnoses (e.g., paroxetine for social anxiety, GAD, OCD, panic

disorder, PMDD, and PTSD; sertraline for panic disorder, OCD, PMDD, social anxiety disorder, and

PTSD), it is our position, as noted earlier, that until studies clearly show differences between these

drugs in treating specific anxiety disorders, all SSRIs are probably reasonably effective across the

whole range of anxiety disorders. Their use for such conditions may require adjustments, but these

apply to the specific disorder and the whole class of SSRIs. (For example, in treating patients with

panic with agoraphobia, begin with very low dosages; in treating OCD patients, use higher dosages

and wait even longer for clinical response.)

Agoraphobia and Panic

Several antidepressants exert major antianxiety effects. Imipramine was first reported by Klein and

various colleagues in the 1960s (Klein 1967) to have potent anxiolytic effects in agoraphobic

patients with panic. Clinically, it appears that most, if not all, TCAs and SSRIs exert similar

antipanic effects. In addition, the MAOI phenelzine is a potent antipanic agent, as probably are the

other MAOIs and trazodone. However, not all antidepressants are as effective in treating panic.

Most notably, bupropion does not appear to exert antipanic effects as reliably. The noradrenergic

effects of various antidepressants (particularly the TCAs and MAOIs) on the locus coeruleus

generally have been invoked to explain their antipanic activity. Whether such a mode of action

explains the possible antipanic effects of trazodone is unclear.

Fluoxetine and the other SSRIs appear to block panic attacks. Indeed, paroxetine and sertraline

now have FDA approval for use in panic disorder. Generally speaking, doses for paroxetine are

higher in treating panic than in treating major depression. The starting dosage is 10 mg/day, andPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

13 of 35

31/01/2009 06:51

the therapeutic dosage range is 40–60 mg/day.

Early on, the general rule of thumb was that panic patients required only low doses of TCAs (e.g.,

50 mg/day of imipramine) for a response to occur. Over the years, it became more evident that, as

in depression, many panic patients require relatively higher doses of TCAs or MAOIs, although a

small proportion are very sensitive to TCAs, tolerating only 10–25 mg/day of imipramine. We

recommend, when indicated, using the general dosage regimens of TCAs that are used for

depression (see Chapter 3: “Antidepressants”).

Klein (1993) proposed that TCAs are effective in panic disorder by affecting a supersensitive

threshold for feeling smothered. In a second international collaborative study comparing

imipramine with alprazolam and placebo in panic patients, patients with prominent respiratory

panic symptoms (e.g., shortness of breath, choking feelings) showed more improvement with

imipramine, whereas patients who did not have these symptoms during panic attacks did better

with alprazolam. We know of no similar data on SSRIs.

It appears that patients with panic disorder may be quite sensitive to the stimulatory properties of

some SSRIs and require extremely low dosages of fluoxetine (5–10 mg/day) at initiation of

treatment. Dosages of 20 mg/day of fluoxetine are frequently effective.

Generalized Anxiety Disorder

Past studies have pointed out that TCAs also exert effects in GAD. In one major study, imipramine

was as effective at 4–6 weeks as the benzodiazepine chlordiazepoxide in patients with this

disorder. However, in the first 2 weeks, the benzodiazepine was more effective. More recent

studies have led to FDA approvals for venlafaxine and paroxetine in the treatment of GAD.

Given the current DSM-IV-TR criteria for GAD as a chronic disorder—one that probably requires

long-term treatment—the role of benzodiazepines has been reduced to the short-term relief of

symptoms, if clinically necessary, while a longer-term non-dependence-inducing therapeutic

program is implemented. Currently, venlafaxine, escitalopram, and paroxetine are FDA approved

for use in GAD, but probably all SSRIs are effective, and all the newer drugs are probably more

benign than the older TCAs, though we know of no direct comparisons between newer and older

antidepressants in treating this condition. CBT approaches are available for the symptoms of GAD.

Again, we know of no direct comparisons of specific drugs and specific CBT approaches for GAD, but

the use of multimodal approaches is a function of the cost, the patient’s response to the initial

treatment trial, and the availability of skilled CBT therapists in the geographic area.

Social Phobia

Seriously symptomatic patients with social phobia experience marked anxiety in a range of “social”

situations, such as eating in public, signing checks, public speaking, and even being in large groups.

The condition, as defined in DSM-IV-TR, may also include more limited fears of performing or

speaking in public, often called performance anxiety. Performance anxiety is less incapacitating

than generalized social phobia but may affect an area vital to a patient’s career or interests. There

is reasonable evidence that milder degrees of performance anxiety studied in volunteers (e.g.,

music students) respond to -blockers given a couple of hours before the performance. Several

-blockers, including propranolol, oxprenolol, alprenolol, and atenolol, have been a bit more

effective than placebo in individual controlled studies. Atenolol is cardioselective and may not cross

the blood-brain barrier easily, suggesting that -blockers may act, at least in part, by suppressing

tachycardia and tremor.

Social phobia has been increasingly well studied over the past few years. This disorder is often

comorbid with a variety of other Axis I disorders, such as panic disorder, major depressive

disorder, body dysmorphic disorder, and substance abuse. Because social anxiety disorder appears

to begin in childhood or adolescence, early identification and treatment may be of special value in

preventing the onset of other complicating disorders.

At present, there is good evidence that a variety of drug therapies are more effective than placebo,Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

14 of 35

31/01/2009 06:51

with the conventional MAOI phenelzine being the most effective and best studied of the

first-generation antidepressants. Moclobemide, a short-acting, reversible MAOI available in Canada

and Europe, appears to be somewhat less effective. Brofaromine, another reversible MAOI, was

reported to not separate from placebo in social phobia. Clonazepam has been well studied by

Davidson’s group (Davidson 2000) and is substantially more effective than placebo. Alprazolam, in

other studies, has had a weaker effect. Paroxetine has been very well studied and enjoys an

FDA-approved indication for the treatment of social phobia. Dosages from 20 to 60 mg/day

appeared equally effective in one multisite study. Other SSRIs are presumably also effective, and

several have shown efficacy in smaller double-blind or open studies. -Blockers (or at least

atenolol) are not much better than placebo in controlled studies in the treatment of generalized

social phobia, despite their efficacy in ameliorating performance anxiety.

There are only a few small studies comparing CBT with pharmacotherapy, and these show

equivalent efficacy. Interesting reasons have been suggested as to why the “exposure” socially

phobic patients receive to anxiety-producing situations in the course of their lives does not

extinguish the anxiety; CBT programs have been developed that get around this problem and are

quite successful.

One tentative algorithm for treating new patients with generalized social phobia is to begin with a

benzodiazepine to reduce initial apprehension; then add an SSRI, stopping the benzodiazepine after

the SSRI has begun to work; and then add CBT before phasing out the SSRI.

Duration of treatment for patients with generalized social anxiety has not been well studied. Since

the condition is generally chronic, pharmacotherapy of at least a year’s duration with evidence that

the patient can handle previously feared situations seems reasonable before attempts are made to

taper medication. Some patients may require medication indefinitely.

It should be noted that gabapentin at dosages averaging about 3,000 mg/day has been shown to

be more effective than placebo in treating social anxiety disorder. The reason for its presumed

efficacy is unclear.

For patients who have not responded to SSRIs and CBT, an MAOI trial is obviously the next step.

Clomipramine may also be effective in treating social phobia.

On the basis of the available evidence, mild social and/or performance anxiety could be treated

first with a -blocker, and patients with more severe impairment could be given an SSRI before a

trial of phenelzine. Clomipramine could be tried in patients adverse to trying MAOIs who have not

had a response to SSRIs either because of discontinuation due to side effects or because of lack of

improvement. If both clonazepam and alprazolam are effective in treating social phobia,

clonazepam should be tried first because of its longer half-life. Benzodiazepines probably rapidly

exert an effect on social anxiety; SSRIs or MAOIs may take weeks to obtain a full effect.

Body Dysmorphic Disorder

Patients with body dysmorphic disorder are preoccupied with the belief that one or more of their

physical features—nose, skin, ears, hair, and so forth—are so “bad” that others will reject them

because of the abnormality. Such patients may spend hours every day scrutinizing their presumed

blemishes and suffer greatly in the process. Behavioral approaches, some as simple as turning all

mirrors to the wall so patients cannot inspect themselves, may help. Phillips et al. (1994) originally

reported that the SSRIs, often in higher doses, were quite effective in treating this condition.

Further, they found that patients with the delusional variant of this syndrome also responded to

SSRIs but did not respond to neuroleptics. More recently, her group reported a 53% response rate

with fluoxetine compared with only 18% with placebo (Phillips et al. 2002). Phillips has been

unable to replicate earlier open studies suggesting that pimozide was the treatment of choice for

monosymptomatic hypochondriacal delusions.

It is worth noting that several conditions classified as anxiety disorders in DSM-IV-TR, as well as

body dysmorphic disorder (Phillips et al. 2001), can exist as sole psychiatric disorders or asPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

15 of 35

31/01/2009 06:51

comorbid conditions for years without ever being reported, even to the psychiatrist treating the

patient for depression or some other, more overt Axis I disorder. Only when the patient is

specifically asked about the manifestations of the conditions, such as body dysmorphic disorder,

OCD, PTSD, panic with agoraphobia, or social phobia, will it be evident that the patient has one or

more of these disorders.

As an example, a patient one of us had known for several years and had seen during several

hospital admissions for acute, recurrent psychotic depression finally admitted to a chronic

life-altering conviction that his nose was very ugly. We have had patients admit to manifestations

that fulfill DSM criteria for other conditions only when a full, formal diagnostic interview (such as

the Structured Clinical Interview for DSM-IV) was done months or years into treatment.

Posttraumatic Stress Disorder

Posttraumatic stress disorder is considered an anxiety disorder with a presumed cause: exposure to

a markedly traumatic event or series of events. In general, studies show that about 20% of persons

exposed to a severe stress develop the disorder. Probably a variety of factors predispose

individuals to develop the full syndrome and to need treatment, including preexisting exposure to

other stresses, prior psychiatric disorders, and the severity and prolongation of the traumatic

experience(s). Group cohesion during major stresses may offer some protection. Most patients with

PTSD have one or more comorbid psychiatric disorders; depression, various other anxiety

disorders, dissociative disorders, and substance abuse are all relatively common and both increase

morbidity and complicate treatment. Further, PTSD syndromes occurring after single adverse

events—rape, fire, motor vehicle accident, volcanic eruption—may be different from PTSD seen in

combat veterans or in patients who were sexually or physically abused as children.

Although one might assume that PTSD is a close chronological sequela of the bad event (and it

sometimes is), some patients may be fine or at least may function well for months or years until

some factor—known, hypothetical, or unknown—causes a sudden emergence or reemergence of the

typical patterns of symptoms. An episode of PTSD has been estimated to last around 7 years;

however, some patients adapt to the memories and symptoms and recover early, whereas others

have symptoms that vary in intensity but seem to persist for decades. PTSD was reported in some

World War I veterans in studies published as late as the 1980s.

It is generally presumed that patients with PTSD have overactive autonomic nervous systems and

that their cortisol responses to stress or dexamethasone are blunted—the opposite of the status in

severe depression. Some patients with PTSD feel much better after being given a single dose of an

opiate antagonist (nalmefene), whereas others feel much worse (Glover 1993). In one study of

PTSD patients, a serotonin agonist, m-chlorophenylpiperazine, induced an exacerbation of PTSD

symptoms, while yohimbine, an adrenergic agonist, induced panic and flashback symptoms

(Southwick et al. 1997). In a similar small study, the pure benzodiazepine antagonist flumazenil

relieved symptoms in PTSD patients (Coupland et al. 1997).

Further, most published clinical trials of treatments for PTSD involved outpatients who were not

taking any medications and whose illness is presumably less severe than that of patients in their

seventh inpatient psychiatric hospital admission whose condition has worsened despite having

received prolonged trials of a variety of concomitant medications and, often, psychosocial

therapies. In these days of limited hospital stays, it is no longer an option to withdraw a patient’s

numerous drug therapies to see whether any are helping or hurting. At least at McLean Hopsital,

such readmissions for severe PTSD usually involve women with histories of childhood sexual abuse

who meet the criteria for, in addition to PTSD, a number of other psychiatric disorders. These

patients often dissociate a lot; have auditory hallucinations, severe insomnia, severe depression,

and recurrent substance abuse; and exhibit self-injurious behavior. They are likely to be taking

clonazepam, valproic acid, gabapentin, olanzapine, SSRIs, bupropion, clonidine, or lamotrigine, as

well as Fioricet for headaches, and despite their medications they still feel terrible. It may be that

severe symptoms and self-destructive behavior elicit overprescribing in psychiatrists and that the

presence of such patients on psychiatric wards encourages other, more naive patients to emulatePrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

16 of 35

31/01/2009 06:51

an expanding repertoire of symptoms. Although some reported trauma histories sound rather

incredible, most seem valid and are likely true.

Given all the above factors, the most widely studied drugs for the treatment of PTSD are the SSRIs.

They are often more effective than placebo, though only 40%–50% of patients taking SSRIs for

PTSD show major improvement. The older MAOIs and brofaromine, a short-acting reversible MAOI

no longer in development, may be more often effective than SSRIs, but these drugs’ serious

adverse effects in unstable, self-destructive outpatients make it hard to risk prescribing them.

Using nefazodone was found to be helpful in several small open studies, and lamotrigine was

helpful in one small placebo-controlled study. Uncontrolled studies of moclobemide, fluvoxamine,

paroxetine, mirtazapine, venlafaxine, valproate, and carbamazepine have all shown some benefit.

Both paroxetine and sertraline were significantly more effective than placebo in PTSD patients, and

both now have FDA approval for the disorder. It may be that anticonvulsants are better for rage

and instability, whereas antidepressants are better for depression and anxiety. Clonidine and

guanfacine are said to suppress nightmares. These have not been well studied (Pearlstein 2000).

Most PTSD patients with auditory hallucinations and “paranoid” ideas (e.g., fear of being attacked

when exposed in public places) have found the typical antipsychotics (e.g., perphenazine 16

mg/day) helpful, though the psychotic-like symptoms do not completely abate. The newer atypical

antipsychotics are less well studied. Many patients object vociferously to the weight gain.

Ziprasidone is too new to be judged. Quetiapine has been used in a fair number of patients at

McLean Hospital without any compelling results.

Autonomic drugs—clonidine, guanfacine, propranolol—have been suggested for treating PTSD

symptoms without there being any clear evidence of efficacy. One wonders whether methadone

clinic patients, particularly veterans, might be able to report on the effects of opiates on PTSD

symptoms.

One study showed that among recent trauma survivors, acute use of a benzodiazepine for several

weeks after the traumatic experience tended to increase the likelihood that PTSD would develop.

After PTSD is well established, patients often have severe initial insomnia and severe anxiety and

thus almost invariably are prescribed benzodiazepines. There is even suggestive evidence that

using alcohol regularly may delay onset of PTSD; thus, the role of sedatives in PTSD is quite

unclear.

In patients with severe depression with melancholic features plus PTSD, ECT will occasionally be

quite helpful.

In short, for milder or less intractable PTSD, SSRIs are the treatment of choice. In one study of

PTSD patients, longer-term treatment with fluvoxamine led to a better outcome, with more

improvement in PTSD symptoms after 1 year than had been seen at 6 weeks. Because PTSD in the

majority of patients seems to have a prolonged, chronic course, long-term use of any substantially

helpful medication seems sensible.

Current expert opinion on psychosocial therapies (Foa 2000; Hembree and Foa 2000) favors

exposure therapies (though frequent experiencing of flashbacks is clearly not helpful), perhaps

with some cognitive therapy aspects to encourage compliance. Certainly, PTSD patients often have

“bad” cognitions: typically that the world is much more dangerous than it really is and that they are

much more helpless and inadequate than they really are. Eye movement desensitization and

reprocessing (EMDR) therapy—having the patient reexperience traumatic memories while the

therapist’s fingers are waved back and forth before the patient’s eyes and the patient moves his or

her eyes from side to side—has been assessed in controlled trials and seems to be a form of

exposure therapy. Holding the eyes still does not affect the benefit achieved through EMDR.

All in all, the treatment of PTSD is now in the state that treatment of OCD was before the advent of

clomipramine. There is no very effective definitive pharmacotherapy. SSRIs are probably the best

studied, relatively effective medications, and here the resemblance to OCD occurs: both PTSD and

OCD are long-term conditions, and prolonged drug therapy, at least with the SSRIs, is likely to yieldPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

17 of 35

31/01/2009 06:51

better results over time. A wide variety of other agents are said to be helpful sometimes but are not

well studied. In PTSD, memory consolidation and stress arousal systems are probably

malfunctioning. Benzodiazepines are, perhaps, contraindicated, but many PTSD patients have

already been taking them for quite some time, and there is no evidence that discontinuing them

results in long-term improvement to compensate for the probable distress caused by withdrawal.

Behaviorial therapies, particularly limited exposure therapies with some cognitive therapy, may be

quite helpful, but perhaps more so in patients with more recent trauma.

Obsessive-Compulsive Disorder

When the second edition of this manual was published in 1991, only one drug, clomipramine, had

been approved by the FDA for use in treating OCD. By 1996, fluoxetine and paroxetine were

approved for use in treating both OCD and depression, and fluvoxamine was approved for use only

in OCD. Now there is evidence that essentially all SSRIs are effective for both depression and OCD.

Despite this overlap in indications, the evidence is strong that the way all the SSRIs work in OCD is

different from the way they work in depression. Depression tends to respond relatively rapidly—in

2–6 weeks—and most patients improve; there is a substantial placebo response. In OCD,

improvement is delayed and may take 6–12 weeks, only half the patients improve, and the placebo

response is smaller. The presence of coexisting or comorbid depression in OCD patients does not

affect the response of their OCD to serotonergic antidepressants. Generally, OCD patients do better

with higher doses of SSRIs (e.g., 40–80 mg of fluoxetine), whereas patients with depression

respond to 20 mg. Noradrenergic drugs (desipramine, nortriptyline, and bupropion), which are

quite effective in depression, are ineffective in OCD.

If one accepts that all currently prescribable SSRIs are effective in OCD, it is hard to make choices

among them. All cause a fair amount of sexual dysfunction. Some differences in their degrees of

binding to various receptors and their interactions with drug metabolism exist. It is hard to

compare efficacies of these drugs, even in meta-analyses, because the kinds of patients being

recruited have likely changed over time. In early clomipramine studies, most patients had never

had an adequate trial of an effective drug, and drug response rates and placebo rates were around

50% and 5%, respectively. Since then, placebo rates have risen and drug improvement rates have

dropped.

The principles of treating patients with any of the SSRIs remain clear: begin with a standard

antidepressant dose and increase gradually to three to four times that dose if improvement does

not clearly occur. Prepare the patient for at least an 8- to 12-week trial before changing drugs or

adding other drugs to the SSRI.

Over the years a variety of drugs have been tried alone or added to clomipramine or one of the

SSRIs to elicit or increase a therapeutic effect. So far, none—lithium, buspirone, L-tryptophan,

fenfluramine, neuroleptics, clonazepam—has had any regular, substantiated success, though case

reports suggest that an occasional patient may improve when one of these drugs is added.

Buspirone appeared very effective as an augmenter in open trials but did not separate from placebo

in a double-blind study. It should be noted that buspirone could elicit a serotonin syndrome when

added to an SSRI (although this was not commonly seen) and that L-tryptophan and fenfluramine

are no longer available in the United States. In the last few years, as evidence has increased on the

overlap of OCD with Tourette’s disorder, the newer atypical antipsychotics have been added to

SSRIs in patients with OCD, with impressive improvements in symptoms seen in occasional

patients. Such a response is perhaps more likely if the patient also has schizotypal features or has a

history or family history of tics. At this point, risperidone at dosages up to 3 mg/day seems safest

in terms of weight gain and oversedation, but orthostatic hypotension has occurred in older

depressed patients when risperidone and an SSRI were combined. So far, ziprasidone has not even

been looked at informally in this regard. Conversely, consider, as an illustration that almost any

drug or drug class stretches across DSM-IV-defined boundaries, a controlled study showing that

fluvoxamine was substantially more effective than placebo in about half of a group of male autisticPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

18 of 35

31/01/2009 06:51

adults, improving sociability as well as autistic compulsions (McDougle 1997). The bottom line is

that about half of patients with OCD are a good deal better while taking an SSRI or clomipramine

and about 15%–20% drop out of drug therapy because of side effects. Not an ideal situation.

On the other hand, CBT (Baer and Greist 1997) is about as effective as pharmacotherapy if

adequately done; relaxation therapy does not help, but therapies involving in vivo exposure and

ritual prevention are typically quite effective. OCD patients with cleanliness or counting rituals do

much better than patients with pure mental obsessions and no compulsive behaviors, hoarders, or

patients with compulsive slowness. Perhaps all OCD patients should get both CBT and medication.

However, therapists capable of doing CBT are mainly at major university centers with specialized

OCD programs, so availability of CBT (and cost, sometimes) is a major problem. Early efforts that

incorporate computerized programs and books that patients can use to carry out their own CBT

with backup from a live professional seem to be effective (Baer and Greist 1997).

With or without CBT, OCD patients should receive, in adequate dose and length, separate trials of at

least two of the SSRIs and of clomipramine before going on to more elaborate or complex

therapies. For extremely treatment-resistant OCD, either intravenous clomipramine therapy (not

available in the United States) or stereotactic psychosurgery (Jenike et al. 1995; Mindus and Jenike

1992), both of which have some reported benefit, could be tried.

If a patient has long-standing OCD symptoms and improves on a drug regimen, long-term

maintenance therapy on that regimen seems advisable. Patients who are taken off medications

tend to relapse fairly rapidly. With CBT, it is more likely that a series of treatments—say 12

sessions—with marked behavioral improvement will lead to maintained improvement after CBT is

stopped, though occasional “booster” sessions may be useful.

ANTICONVULSANTS

A number of anticonvulsants are being used with increased frequency for the treatment of anxiety

disorders. Many of these drugs were thought primarily to act on the GABAergic system and would

be expected to have anxiolytic properties. They also tend to have utility in the treatment of

neuropathic pain. Among these anticonvulsants are gabapentin, tiagabine, and pregabalin. The

anticonvulsants probably work slower than benzodiazepines and faster than antidepressants in the

treatment of anxiety. Among the advantages of the anticonvulsants relative to benzodiazepines is a

lower risk of dependence and withdrawal.

Gabapentin (Neurontin) has been used for many years in the treatment of bipolar disorder,

although very little evidence exists as to its efficacy in this disorder. Once thought to be primarily

GABAergic, it appears now that it binds to an 2- subunit of calcium channels found in brain, and

this is the more relevant mechanism of action. On the other hand, there is more evidence that

gabapentin is at least modestly effective in the treatment of social anxiety, including public

speaking, generalized anxiety, and panic disorder. Our experience is that gabapentin is much less

effective than either benzodiazepines or antidepressants in the treatment of anxiety. However,

gabapentin has few side effects or drug interactions and has not been associated with a risk for

dependence. Doses as low as 300–400 mg may be helpful in the treatment of social anxiety, but

most patients seem to need between 900 and 2,700 mg/day (in divided doses) for the treatment of

panic or more serious anxiety. The primary side effects are somnolence and fatigue.

Tiagabine (Gabitril) is a more potent drug than gabapentin and selectively inhibits GABA reuptake

and the GAT1 GABA transporter. Tiagabine has been studied primarily in GAD either as a

monotherapy or in combination with SSRIs. In one study, tiagabine was as effective as paroxetine

and more effective than placebo in the treatment of GAD. There are anecdotal reports of tiagabine’s

efficacy as an adjunctive agent in treating PTSD and in treating panic disorder. We usually start

tiagabine at 2 mg/day and work toward a target dosage of 2 mg bid up to 8 mg bid. While the

lowest doses are well tolerated, some patients complain of sedation, somnolence, cognitive

problems, or even feeling as if they were drunk at higher dosages.

Pregabalin (Lyrica), an analog of GABA that is related to gabapentin, has been extensively studiedPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

19 of 35

31/01/2009 06:51

in the treatment of GAD, neuropathic pain, and partial complex seizures. It appears to act even

more selectively on the 2- subunit of calcium channels found in brain than does gabapentin.

Pregabalin received FDA approval in late 2004 for use in treating neuropathic pain and epilepsy.

However, its approval for the treatment of generalized anxiety is still pending. At the time of this

writing, there have been at least five placebo-controlled trials which suggest that pregabalin is at

least as effective as alprazolam and venlafaxine and more effective than placebo in the treatment

of GAD. Pregabalin was more rapidly acting than venlafaxine and controlled somatic symptoms as

well as did alprazolam (Montgomery 2006). In addition, pregabalin looks about as effective as

venlafaxine in treating GAD but is better tolerated and probably more rapidly acting (Montgomery

et al. 2006). The starting dosage of pregabalin in the treatment of anxiety is expected to be 150

mg/day, with the dosage increasing to 300 mg/day. Most patients should do fine at 300–400

mg/day, but the dosage can be increased by 150 mg every few days to the maximum of 600

mg/day. Although the studies suggest pregabalin should be a good first-line agent for the

treatment of GAD, other promising anxiolytics, such as buspirone, did not live up to their

expectations. Still, we expect pregabalin will fill an important niche. We have seen some patients

who have either not tolerated or not responded to SSRIs who have done well with pregabalin.

Patients requiring immediate relief of anxiety in an emergency setting will still receive

benzodiazepines. However, pregabalin will be a more rapidly acting and probably better alternative

to the SSRIs for some patients. The most common side effects of pregabalin have been dizziness

and somnolence. This drug, like gabapentin and tiagabine, lacks the sexual side effects of most

antidepressants.

ANTIPSYCHOTICS

Antipsychotics have long been used as adjunctive agents in the treatment of anxiety associated

with disorders such as schizophrenia and bipolar disorder (Hirschfeld et al. 2006; Keck et al. 2006).

In addition, early work with agents such as trifluoperazine suggested that antipsychotics might be

as effective as benzodiazepines in the treatment of GAD. The atypicals, with their 5-HT2 effects,

would be expected to have some benefits in anxiety disorders. Most of the controlled studies of

antipsychotics in anxiety states have involved the adjunctive use of atypical agents to augment

antidepressant effects in OCD and PTSD. While the data are somewhat conflicting, most of the

studies suggest that atypical antipsychotics, especially risperidone, olanzapine, and quetiapine,

may be useful adjunctive agents in the treatment of refractory OCD and PTSD (Gao et al. 2006). In

addition, the atypicals have sometimes been said to be effective in the adjunctive treatment of

panic disorder and OCD.

We have seen instances when the atypicals have been well tolerated and effective as monotherapy

of anxiety states when other conventional agents were not. Still, as of this writing, there are no

adequately controlled studies of atypical antipsychotics in the monotherapy of any anxiety

disorder. Until there is, we would recommend using the atypical antipsychotics in the treatment of

anxiety only when more established agents have been exhausted.

NORADRENERGIC AGENTS

In recent years, a number of studies have pointed to the potential use of -blockers (e.g.,

propranolol) and primarily presynaptic but also postsynaptic 2 receptor agonists (e.g., clonidine)

to ameliorate symptoms of anxiety. Use of these agents stems from the observation that certain

symptoms of anxiety (e.g., palpitations, sweating) suggest involvement of the sympathetic nervous

system. Investigations were first directed toward the use of -blockers in anxious musical

performers. A number of years later, clonidine was shown by Gold et al. (1978) to be effective in

blocking physiological symptoms associated with opioid withdrawal, resulting in its eventual study

in patients with anxiety disorders and possibly nicotine withdrawal. This drug exerts 2

(presynaptic) receptor agonist effects; however, because it is also a postsynaptic 2 agonist, its

pharmacological actions are complex.

Generic and trade names of key noradrenergic agents are summarized in Table 6–3 later in this

chapter.Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

20 of 35

31/01/2009 06:51

Clinical Indications

-Blockers (e.g., propranolol) are indicated for hypertension and for prophylaxis against angina,

arrhythmias, migraine headaches, and hypertrophic subaortic stenosis. They are often quite useful

in relieving akathisia in patients taking neuroleptics, although they are not approved by the FDA for

this use (see Chapter 4: “Antipsychotic Drugs”). They are also not approved by the FDA for use in

anxiety, although several studies have suggested that propranolol may be useful. These studies,

originally conducted in Great Britain, pointed to -blockers’ having particularly potent effects on

the somatic manifestations of anxiety (e.g., palpitations, tremors) and less dramatic effects on the

psychic component of anxiety. The antitremor properties of these drugs have resulted in their being

commonly used in treating patients whose hand tremors developed secondary to lithium carbonate

use (see Chapter 5: “Mood Stabilizers”).

A number of reports suggested that although -blockers had some use in generalized anxiety, they

were not particularly effective in blocking panic attacks. Indeed, Gorman et al. (1983) reported

that propranolol failed to block lactate-induced panic attacks. However, some investigators noted

that propranolol might block panic anxiety resulting from isoproterenol (an adrenergic agonist)

infusions and thus could still be effective in treating some patients with panic attacks. Pindolol, a

mixed -adrenergic receptor agonist and antagonist with serotonergic properties, has been

reported to augment antidepressant response to SSRIs (see Chapter 9: “Augmentation Strategies

for Treatment-Resistant Disorders”).

-Blockers are widely believed, by medical students, house staff, musicians, and performers, to be

useful in allaying anxiety during speaking or performing in public. Such use means that either the

several weakly positive controlled studies of the efficacy of -blockers in such situations are

somehow flawed or placebos that have convincing rationales for efficacy can have a powerful

effect.

Clonidine has an FDA-approved indication for the treatment of hypertension. As noted earlier,

clonidine has been widely studied as a means of blocking physiological symptoms of opioid

withdrawal (e.g., palpitations, sweating). The drug has also been studied in anxiety and in panic

disorder and has been shown to be effective in both, although tolerance to the antianxiety effects

frequently develops. It is conceivable that the drug’s mixed, partial pre- and postsynaptic receptor

agonist properties may enter into the development of tolerance. Clonidine has also been used to

test various aspects of the catecholamine hypotheses of affective and anxiety disorders. Studies on

nicotine withdrawal have yielded mixed results (Franks et al. 1989; Glassman et al. 1988).

Clonidine has also been used to block the tachycardia and excessive salivation seen with clozapine

use (see Chapter 4: “Antipsychotic Drugs”).

Dosage and Administration

Using propranolol as a model, clinicians should begin patients with peripheral symptoms of anxiety

or patients with lithium-induced tremor or with familial tremor at 10 mg bid and increase the

dosage incrementally to approximately 30–120 mg/day (see Table 6–3). Although the usual

maintenance dosage of the drug in patients with hypertension is as high as 240 mg/day, such a

dosage is rarely needed for anxious or tremulous patients. Generally, the use of these agents in

patients with anxiety disorders should parallel that of the benzodiazepines; trials should be made

of having patients stop the drug after a few weeks of treatment.

Many patients with tremors secondary to lithium carbonate treated with a -blocker show a

reemergence of their tremors after discontinuation of the -blocker, resulting in their continuing to

take a -blocker for a prolonged period. We know of no major untoward effects; however, some

patients may become lethargic and even depressed while taking -blockers, so clinicians need to

keep this in mind when treating patients with a major affective disorder (see “Side Effects” section

below). This potential effect is a matter of some debate. We have also used propranolol for

TCA-induced tremor without affecting the depression in the vast majority of patients.

If propranolol or another -blocker is to be used as prn medication to decrease the physiologicalPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

21 of 35

31/01/2009 06:51

and, perhaps, psychological effects of stage fright or other circumscribed, predictable, socially

phobic situations, the patient should try the proposed dose (usually 10 or 20 mg) once or twice

before using it before a performance to block stage fright to make sure he or she can tolerate the

drug comfortably at that dose (Jefferson 1995). Clonidine should be started at a dosage of 0.1 mg

bid, and the dosage should be increased by 0.1 mg every 1–2 days to a total daily dose of 0.4–0.6

mg (Table 6–3). Because some studies have indicated that tolerance to this drug develops,

clinicians should attempt to limit the duration of exposure whenever possible.

Side Effects

Side effects of the -blockers include bradycardia, hypotension, weakness, fatigue, clouded

sensorium, impotence, gastrointestinal upset, and bronchospasm. For the psychiatrist, a few

caveats appear warranted. Clinicians need to remember that these drugs are contraindicated in

asthmatic patients because they may produce bronchospasm and in patients with Raynaud’s

disease because of the risk of increased peripheral vasoconstriction. Pindolol, which acts as a

mixed -adrenergic receptor agonist and antagonist, has less effect on receptors that control

bronchial constriction and has been argued to be potentially safe in patients with asthma. However,

its marked agonist effects can result in unpleasant stimulation, and we have not found it

particularly useful in treating anxious patients. As for the capacity of -blockers to cause

depression, we have not seen patients who have developed true depressive disorders. Rather, we

have noted that some patients may feel “washed out” or lethargic. However, clinicians at other

institutions have reported cases of propranolol-induced depression with endogenous features that

remitted on discontinuation of the drug. One strategy is to switch to a -blocker that is less

lipophilic and that exerts fewer central nervous system (CNS) effects (e.g., atenolol). This strategy

may be particularly useful in men who have experienced decreased sexual potency while taking

propranolol. In stopping -blockers, it is wisest to taper the dose to avoid any rebound phenomena

that could result in untoward cardiac or blood pressure effects.

Clonidine has a mixed side-effect profile. Its major side effects include dry mouth, sedation or

fatigue, and hypotension. These effects are often found unacceptable by anxious patients. In

hypertensive patients, scheduling bid dosing (with two-thirds of the dose given at bedtime) has

been advocated to deal with its sedating effects. Discontinuation should be gradual to avoid

rebound autonomic symptoms or the hypertensive crises that have been reported in hypertensive

patients who were suddenly withdrawn from the drug.

ANTIHISTAMINES

The antihistamine hydroxyzine has indications for the treatment of anxiety and tension associated

with psychoneurotic conditions or physical disease states. It is also indicated in the treatment of

pruritus due to allergic conditions and for pre- and postoperative sedation. In psychiatric practice,

antihistamines are less commonly used in treating anxious patients, reflecting their less potent

anxiolytic effect (Table 6–3). Hydroxyzine’s major side effects are drowsiness and dry mouth. It

does not produce physical dependence; it may produce CNS depression when added to alcohol,

narcotic analgesics, CNS depressants, or TCAs. Another antihistamine, diphenhydramine, is

commonly used in medicine and psychiatry as a sedative-hypnotic (see Chapter 7: “Hypnotics”).

Table 6–3. Other antianxiety/daytime sedative agents

Generic name Brand name Formulations and strengths Dosage range

(mg/day)

Barbiturates

amobarbital Amytal Powder for injection: 250-mg, 500-mg vials 60–150

butabarbital Butisol Tablets: 15, 30, 50, 100 mg 45–120

Elixir: 30 mg/5 mL (480-mL)

mephobarbital Mebaral Tablets: 32, 50, 100 mg 150–200Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

22 of 35

31/01/2009 06:51

Generic name Brand name Formulations and strengths Dosage range

(mg/day)

pentobarbital Nembutal Injection: 50 mg/mL (20-mL, 50-mL vials) 150–200 once (up to

500 mg)

phenobarbital Various

generics

Tablets: 15, 30, 60, 100 mg 30–120

Tablets: 16.2, 32.4, 64.8, 97.2 mga

Elixir: 20 mg/5 mL (480-mL)

Injection: 65 mg/mL, 130 mg/mL (1-mL

prefilled syringes, single-dose vials)

Carbamate

meprobamate Miltown Tablets: 200, 400 mg 1,200–1,600

Noradrenergic

agents

clonidine Catapres Tablets: 0.1, 0.2, 0.3 mg 0.2–0.6

Catapres-TTS Transdermal patch: 0.1, 0.2, 0.3 mg/day

Duraclon Injection: 0.1 mg/mL, 0.5 mg/mL (10-mL vial)

propranolol Inderal Tablets: 10, 20, 40, 60, 80, 90 mg 60–160

Concentrate: 80 mg/mL (30-mL)

Oral solution: 20 mg/5 mL, 40 mg/5 mL

(500-mL)

Injection: 1 mg/mL (1-mL ampule)

Inderal LAb

Capsules: 60, 80, 120, 160 mg

Antihistamines

hydroxyzine HCl Atarax Tablets: 10, 25, 50, 100 mg 200–400

Syrup: 10 mg/5 mL (120-mL, 480-mL)

Vistaril Injection: 25 mg/mL, 50 mg/mL (1-mL

single-dose vial, 10-mL multidose vial)

50–100

hydroxyzine

pamoate

Vistaril Capsules: 25, 50, 100 mg 200–400

Suspension: 25 mg/5 mL (120-mL, 480-mL)

Other

buspirone BuSpar Tablets: 5, 7.5, 10 mg; Dividose tablets: 15, 30

mg

15–60

Note. For information regarding antidepressants as anxiolytics, see Chapter 3: “Antidepressants.”

a16.2 mg equivalent to 1/4 grain, 32.4 to 1/2 grain, and so forth.

bSustained release.

BUSPIRONE

The development of buspirone—a nonbenzodiazepine, generally nonsedating anxiolytic—stirred

considerable excitement in psychopharmacological circles a decade or more ago. It represented the

first prominent anxiolytic to be introduced since the benzodiazepines. The drug was originally

developed as a potential antipsychotic agent. Although found in early clinical trials to have little

antipsychotic potency, buspirone was eventually shown to have antiaggression effects in primatesPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

23 of 35

31/01/2009 06:51

and antianxiety effects in humans. The structure of buspirone is shown in Figure 6–2. Buspirone

went off patent in 2001 and is now available generically in the United States.

Figure 6–2.

Chemical structure of buspirone.

The drug does not bind with high affinity to benzodiazepine and GABA receptors, although it may

have an effect on the chloride channel coupled to the benzodiazepine-GABA receptor complex.

Buspirone has little antiseizure effect. Its anxiolytic effects were originally postulated to occur via

dopaminergic properties, although the drug’s central dopaminergic effects were not entirely clear.

Later, buspirone was shown to exert its antianxiety effects by acting as a partial agonist of the

5-HT1A receptor. This action is shared by gepirone and ipsapirone, related antianxiety drugs that do

not have effects on the dopaminergic system.

Buspirone is an interesting, frustrating drug (Cole and Yonkers 1995) with properties that should

make it the drug of choice in treating GAD and related anxiety disorders (e.g., social phobia, mixed

anxiety and depression, anxiety in patients with a history of substance abuse). Buspirone is as

effective as diazepam and superior to placebo in double-blind trials involving anxious outpatients.

The available data from such studies do not show diazepam to be faster acting than buspirone,

although most physicians assume that benzodiazepines are faster acting. One small subanalysis

showed that patients with a history of benzodiazepine treatment did less well while taking

buspirone than patients who had never taken a benzodiazepine. Although the difference was

statistically significant, half the patients with a history of benzodiazepine use did improve while

taking buspirone.

Most psychiatrists and many physicians assume that buspirone is weaker and slower in onset than

are benzodiazepines, and many believe it is never effective in patients who have had a

benzodiazepine in the past. These assumptions are simply not true. Buspirone is lacking in

benzodiazepine-like effects and will not relieve benzodiazepine withdrawal symptoms. If patients

like the sedation they feel after a single dose of benzodiazepine (most do not like it), they will not

get that “pause that refreshes” from buspirone. Unfortunately, both buspirone and benzodiazepines

take 2–4 weeks to cause a full antianxiety effect. Psychiatrists rarely see anxiety patients who

have never taken a benzodiazepine and therefore never see buspirone-appropriate patients and do

not believe that buspirone is effective. Primary care physicians tend to learn from psychiatrists

about drugs with complex dosage requirements that take weeks to act. Buspirone therefore is

rarely used in adequate dosages—up to 30 mg/day or higher for 4–6 weeks—and is consequently

widely believed not to work. Some, but not all, more recent studies show that buspirone is more

effective than placebo in patients with depression, with social phobia, and with combined anxiety

and alcoholism, at dosages of 30–60 mg/day, which patients do not tolerate as initial dosages. The

best part of the buspirone story is therefore seldom reached.

Patients receiving drug therapy for anxiety disorders often do not need years of maintenance

medication. Benzodiazepines can pose problems when they are tapered (particularly abruptly) and

stopped. Anxious patients find that benzodiazepine withdrawal symptoms resemble the anxiety

symptoms that first brought them to the doctor, become upset, and often end up taking diazepam

or alprazolam again. If the patient had been taking buspirone instead, no withdrawal symptoms

would have occurred; in fact, patients tend to improve a bit more in the first 2 weeks afterPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

24 of 35

31/01/2009 06:51

buspirone is stopped. Buspirone is therefore a much more flexible drug to use in treating anxious

patients: it can be tapered and withdrawn easily, and it enables the treating doctor to know readily

whether it is still needed—without the patient’s having to struggle with physical symptoms of drug

withdrawal.

Buspirone should be initiated at a dosage of 5 mg bid, and the dosage should be increased

gradually to 30–60 mg/day. Buspirone is not helpful in patients who are just stopping a

benzodiazepine. However, if buspirone is added to the benzodiazepine for 2–6 weeks, the patient

may feel “more better” because the two antianxiety drugs work by different mechanisms. Even if

this effect does not occur, the benzodiazepine can be slowly tapered, often quite smoothly

(Udelman and Udelman 1990), and the patient can be stabilized while taking buspirone.

Buspirone is well tolerated by medically ill elderly patients, does not depress respiration in patients

with lung disease, and has some utility in patients with organic impulse disorders and in AIDS

patients with anxiety. It does not adversely affect coordination or cognition. In short, it ought to be

a major improvement over the benzodiazepines. However, its use in psychiatry has been limited to

mainly adjunctive treatment.

Side effects of buspirone include headache, nausea, dizziness, and tension, which generally are not

major problems. Indeed, the drug appears to have a more desirable side-effect profile than do the

benzodiazepines. It does not appear to impair motor coordination, and it shows little untoward

interaction with alcohol. According to an early report, the drug may exacerbate psychosis in

patients with schizoaffective disorder, an effect that reflects complex prodopaminergic properties.

This has not been a problem in clinical usage in the United States. High doses have, on the other

hand, been reported to improve dyskinesia in patients with severe tardive dyskinesia. As noted in

Chapter 11 (“Pharmacotherapy for Substance Use Disorders”), buspirone use also can result in

reduced drinking in alcoholic outpatients.

The relative merits and side effects of buspirone versus the SSRIs in treating GAD have, to our

knowledge, not been studied, so no clear statement can be made about this issue. It seems safe to

say that buspirone has even fewer discontinuation effects than do SSRIs. There is one controlled

trial in which venlafaxine XR at 75 and 150 mg/day appeared to be more effective than both

30 mg/day of buspirone and placebo in GAD patients (Davidson et al. 1999).

BARBITURATES

Fifty to sixty years ago, the only medications widely used in psychiatric patients for the treatment

of anxiety or agitation were barbiturates. Although barbiturates are still widely used in anesthesia,

use of barbiturates in psychiatry is now exceedingly uncommon. Longer-acting barbiturates, such

as phenobarbital or barbital, were widely used for daytime sedation, and shorter-acting

barbiturates with presumably more rapid onset of action, such as secobarbital, amobarbital, or

pentobarbital, were widely used as hypnotics. (For further discussion of hypnotics, see Chapter 7:

“Hypnotics.”) Amobarbital in particular was also widely used as a daytime sedative and, in

combination with D-amphetamine, as a widely used mixed sedative and stimulant called Dexamyl,

no longer available in the United States. Phenobarbital is the only barbiturate that is widely used in

general clinical practice currently, and it is used essentially only in the treatment of epilepsy. It has

some efficacy as a daytime sedative and possibly as an antianxiety agent in doses of 15–30 mg

three or four times a day. It was also used as a long-acting sedative in some detoxification

programs (as the “methadone” of the barbiturate group) in withdrawing shorter-acting sedatives

or, occasionally, alcohol. However, the small therapeutic index and high abuse potential relative to

the benzodiazepines have rendered the oral barbiturates all but obsolete. Still, the barbiturates are

widely used in anesthesia and are still occasionally used in hospital settings.

In double-blind, controlled clinical trials comparing phenobarbital with placebo as well as with a

benzodiazepine or meprobamate, phenobarbital was generally slightly more effective than placebo

and inferior to the newer antianxiety agents. In many such studies, the dose of phenobarbital used

was fixed and low, and its efficacy may well have been reduced by the conditions of the trial.Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

25 of 35

31/01/2009 06:51

However, many patients find the sedative effect of phenobarbital rather dysphoric and unpleasant,

and its utility as an antianxiety drug is therefore limited. In patients taking phenobarbital for

epilepsy and, more strikingly, in children or even adults with a history of

attention-deficit/hyperactivity disorder (ADHD), phenobarbital can in fact aggravate hyperactivity

and disorganized behavior. Occasionally, increased hyperactivity and disturbed behavior in some

children and adolescents can be traced to antiasthmatic drugs that contain phenobarbital.

Meprobamate and benzodiazepines can also exacerbate symptoms of ADHD.

It is quite possible that amobarbital or other relatively shorter-acting barbiturates, unlike

phenobarbital, might be as effective as benzodiazepines in the daytime treatment of anxiety.

Several have indications for daytime sedation, and of these, butabarbital (Butisol) is occasionally

used in such a way. (See Table 6–3 for formulations and strengths and Figure 6–3 for chemical

structures of the barbiturates.) No good controlled clinical trials have been done comparing such

shorter-acting and possibly less dysphoric and more euphoric barbiturates with benzodiazepines.

There is, however, little point in carrying out such trials, because it is reasonably clear that the

addiction and abuse liability of barbiturates is substantially greater than that of most

benzodiazepines, including even diazepam. Barbiturates like amobarbital also have the

disadvantage of a relatively low lethal dose, perhaps 1,000 mg taken in a single overdose. In

addition, barbiturates induce enzymes that metabolize other important medications. It is also

possible, but not proved, that tolerance to barbiturates develops somewhat more rapidly when they

are taken in escalating doses. In a classic study by Isbell et al. (1950), performed at the Addiction

Research Center in Lexington, Kentucky, in the 1940s, patients taking very large daily doses of

barbiturates managed to complain of going into withdrawal while being so ataxic and

malcoordinated that they were falling down when they tried to walk and were slurring their speech.

This suggests that tolerance to the antianxiety and/or euphoriant effects of the barbiturates

develops more rapidly than does tolerance to the drugs’ effects on psychomotor functions.

Figure 6–3.

General chemical formulas of the barbiturates.

Amobarbital Sodium (Amytal) as a Parenteral Solution

Amobarbital sodium (Amytal) as a parenteral solution is mostly of historical significance but may

still be of some limited value in psychiatry. When the drug is given intramuscularly to quietPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

26 of 35

31/01/2009 06:51

agitated behavior in disturbed or psychotic patients, a dose of 100 mg is common, with a range

between 50 and 250 mg, depending on the weight of the patient and the degree of excitement.

Parenteral amobarbital sodium has one major advantage over parenteral antipsychotics such as

chlorpromazine, haloperidol, or loxapine in that it acts more rapidly, perhaps in 10–20 minutes, and

when it acts, it tends to produce sleep rather than tranquilization. We know of no systematic study

of the relative utility of amobarbital compared with either lorazepam, chlorpromazine, or

haloperidol in treating acutely disturbed behavior requiring only a single injection. The advantage

of amobarbital, as noted, is that it produces sleep relatively rapidly. The disadvantage is that when

the patient awakens, there is no residual antipsychotic activity to modulate the patient’s

subsequent behavior. Although in the past amobarbital was administered intramuscularly several

times a day through prolonged psychotic excitements, there is no real reason to believe that it is in

fact regularly antipsychotic or has any prolonged benefit. There is some suggestion that too

frequent use could result in either tolerance or, occasionally, delirium. Local experience suggests

that lorazepam, given intramuscularly in 1- to 2-mg doses, is as useful as, and is safer than,

amobarbital.

Intravenous amobarbital has been used in extreme emergency conditions in psychiatry to produce

an anesthetic type of sleep within a few minutes. Intravenous amobarbital is usually given in a

solution of 50 mg/mL at the rate of 1 mL every 60 seconds up to a total dosage of 350 mg, with the

dose titrated according to the patient’s response. If such “anesthetic” use of amobarbital is being

prescribed, the physician should carefully watch the patient’s breathing and vital signs and give the

medication slowly to make sure that respiration is not suppressed. The major danger in addition to

suppression of the respiratory center in such treatment is the occasional production of a potentially

lethal laryngospasm in patients with irritation of the larynx and upper respiratory system.

Barbiturates can, of course, also produce crises in patients with acute intermittent porphyria.

Intravenous amobarbital in doses of 100–300 mg, and sometimes higher, has been used also to

conduct Amytal interviews with patients with psychiatric conditions. When the amobarbital is

injected slowly over 5–10 minutes, in most psychiatric patients a state of relaxation and mild

intoxication with slurred speech can be achieved, during which patients will often talk more easily

and more volubly about their problems and past experiences. Sometimes patients under these

conditions will reveal material not previously told to the psychiatrist. Although amobarbital has

been called “truth serum,” it is by no means certain that narratives told by patients under the

influence of amobarbital are likely to be any more truthful than those told in the fully conscious

state.

The Amytal interview was developed during World War II by Grinker and Spiegel (1945) as a

treatment for severe combat fatigue. In the typical situation, soldiers emerged from combat

essentially mute, shaking, paralyzed with fear, and seeming peculiar, blocked, and dysphoric. They

were either mute or unable to answer questions in more than monosyllables and appeared unable

to cope emotionally with the traumatic events they had recently experienced. Under the influence

of intravenous amobarbital, such soldiers often were able to give vivid and emotionally charged

accounts of their horrifying experiences, and this form of catharsis often discharged their inner

tensions and enabled them to function thereafter in a more normal and organized fashion, with a

substantially reduced level of anxiety. It is reasonable to believe that amobarbital might be of use

in similar conditions that resemble some kind of acute traumatic stress syndrome encountered in

clinical practice.

Amytal interviews have also been conducted, often with some success, with patients with

hysterical amnesia. Such patients can often, but by no means always, retrieve repressed memories

of past events and give reasonable accounts of relevant portions of their past. The interview can be

used both for patients with isolated episodes of amnesia—for example, for episodes of rape,

assault, or murder—and for patients who profess total amnesia for their entire life.

Amobarbital given intravenously is also occasionally effective in resolving hysterical paralyses and

other conversion symptoms. In cases of acute PTSD and after combat or severe trauma, the effectPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

27 of 35

31/01/2009 06:51

is so dramatic that it seems unlikely placebo would have the same effect. The same applies to use

in fugue or amnestic states or hysterical paralyses. A benzodiazepine, with lorazepam available as

a solution (mg/mL) being the most appropriate, may well be equally effective and less likely to

cause respiratory suppression. No stimulants are legally available for parenteral use, and reports of

their value in facilitating talking therapy, with the stimulant usually mixed with amobarbital, are

purely anecdotal.

It is interesting to note that during the early 1960s, intravenous amobarbital was used to establish

a “sedation threshold”—the dose at which slurred speech or electroencephalogram (EEG) changes

or other endpoints were reached. Patients with DSM-II (American Psychiatric Association 1968)

psychotic depression (e.g., severe, psychomotorically retarded, maybe delusional) reached the

sedation threshold at far lower dosages than did patients with “neurotic” depressions (e.g.,

anxious, less severe) (Shagass et al. 1956).

Since the Amytal interview was introduced, a variety of other drugs have been given to facilitate

talking in psychotherapy. These drugs included methamphetamine, once marketed in ampules for

intravenous use for exactly that purpose, and thiopental sodium, a faster shorter-acting

barbiturate. These were followed by lysergic acid diethylamide (LSD) in oral doses of 50–300 g

(see below), other hallucinogens such as psilocybin, and, most recently, Ecstasy (MDMA;

methylenedioxymethamphetamine). All these agents are very likely to make patients talk more. It

is more difficult to show convincingly that they produce better or faster change in psychotherapy.

Hallucinogens have been studied, with mixed results; none is currently available for medical use.

The one comparative study with follow-up, by Shagass and Bittle (1967), of use of LSD in treating

sociopathic inpatients suggested that the insight and improved behavior dissipated after about 6

months.

Catatonia

Catatonia, a stepchild of psychiatry, is generally only recognized when the patient becomes mute

and immobilized, though other signs can make the syndromal diagnosis. Catatonia used to be

considered a form of schizophrenia, but it is now clear that in most cases an affective disorder

diagnosis ends up being made, although it can occur as part of schizophrenia, delirious mania, or

toxic disorders. It is given special mention here, in a chapter on antianxiety agents, because it

responds very well, usually, to sedative drugs—amobarbital sodium in the 1950s and

benzodiazepines in the last few decades—and because catatonia does not respond to neuroleptics.

Perhaps neuroleptic malignant syndrome (NMS) is, in fact, a variant of catatonia produced by

neuroleptics. If sedatives in systematic, high dosages (see below) do not relieve the symptoms

rapidly (in minutes or hours), then ECT almost always works early in the condition and can even be

used as a maintenance treatment to prevent recurrence (Fink 2001b).

Giving either amobarbital sodium intravenously at a slow rate (50 mg/mL, 1 mL/minute) or 1–2

mg of lorazepam intravenously in the same way slowly, the physician has a chance to carefully

titrate the dose of sedation, to watch for the patient’s awakening from stupor, and to avoid

respiratory depression. Even when the medication is given this way, the procedure should be done

only in a medical facility with oxygen and a crash cart and where facilities for intubation and

tracheotomy are available. These facilities are almost never needed, but laryngospasm can be a

lethal side effect of sodium amobarbital if the patient has an irritated larynx and pharynx. No one

knows whether lorazepam is totally free of this risk.

Having given this warning, it is noteworthy that many patients in catatonic stupor will relax,

awake, talk, feed, and hydrate themselves and act quite normal for 1–4 hours after an amobarbital

injection—one of the most remarkable therapeutic events in psychiatry. In the experience of one of

the authors (J.O.C.), most patients in acute stupor when awakened were nonpsychotic and had no

idea why they had been unable to move or talk. Some, of course, were delusional and were

presumed to have schizophrenia. Repeated amobarbital infusions, if given twice a day, enabled

staff to feed and hydrate catatonic patients but did not seem to produce any lasting change in thePrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

28 of 35

31/01/2009 06:51

recurring stupor.

Lorazepam, as a single therapy, often seems to treat the stupor effectively. Fink and associates at

Stony Brook begin with lorazepam 3 mg/day in divided doses but increase the dosage in daily or

every-other-day steps to as high as 16 mg/day. Patients presenting with catatonia should

immediately receive a workup for ECT and, if the trial of lorazepam fails, be given at least six

suprathreshold seizures, even if the patient appears much improved after, say, three treatments. A

regression after early improvement following two or three seizures is likely if ECT is stopped

(Petrides and Fink 1996).

Excited catatonia and delirious mania respond equally well to benzodiazepines followed by ECT if

necessary.

Neuroleptic malignant syndrome is claimed to also respond to benzodiazepines. These medications

may be less toxic than dantrolene or bromocriptine. Initial cooling of febrile NMS patients is the

first order of business, and ECT can be effective if drug therapies fail.

Maintenance treatment for 6 months using whatever treatment worked in the acute treatment of

the catatonia is indicated. Lorazepam at a dosage of 2–6 mg/day, as well as augmentation with

maintenance ECT (biweekly, monthly, or bimonthly), may be needed. One choice is to pick a regular

schedule; an alternative is to restart ECT within 48 hours of any sign of recurrence.

One interdisciplinary caution: Neurologists have been known to use the abnormal interictal EEG

seen with patients administered ECT to diagnose encephalitis and stop the needed ECT regimen

(Petrides et al. 2001).

Obviously, the Amytal interview can occasionally be useful in those rare cases when a mute

catatonic patient has been admitted to an emergency room or a psychiatric unit with no history and

no identification and the clinician is unable to discover even what the patient’s name and address

are, much less what the patient’s history or probable psychiatric diagnosis is. In such situations,

amobarbital given intravenously can be quite helpful in clarifying the situation. However, it should

be administered only after all reasonable medical or pharmacological causes for such a mute or

unresponsive state have been ruled out. Lorazepam, given intramuscularly, may well be as useful

as amobarbital sodium in relieving such stupor.

MEPROBAMATE

Meprobamate is another agent of historical significance that is rarely used for any condition at this

time. Meprobamate occupies an intermediate position between the benzodiazepines and the

barbiturates, both pharmacologically and historically. It was first marketed about 1956, having

evolved from a chemically related muscle relaxant, mephenesin. Meprobamate, whose structure is

shown in Figure 6–4, has muscle relaxant and sedative properties but was initially evaluated as an

antianxiety agent. On the basis of a small number of enthusiastic findings from uncontrolled studies

on anxiety, it was released to the market at a time when the FDA required only evidence of safety

rather than evidence of efficacy as well. It became an instant national success and had widespread

publicity in the lay media. Meprobamate has anticonvulsant activity, mainly against absence

attacks. Its antianxiety effects in laboratory animals can be blocked by naloxone. However,

although not as toxic as barbiturates, meprobamate has proved less safe than the benzodiazepines.

Given the number of alternatives currently available, it is increasingly difficult to justify a

meprobamate prescription.

Figure 6–4.Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

29 of 35

31/01/2009 06:51

Chemical structure of meprobamate.

After an oral dose, peak blood level occurs in 1–3 hours. It is hydroxylated by the liver, then

changed to a glucuronide and excreted through the kidney. Its half-life is 6–17 hours in acute

administration and longer after chronic administration. In suicidal overdose, meprobamate tablets

form a lump (bezoar) in the patient’s stomach, which, if not removed by gastroscopy, can break up

and reinduce coma after the patient has appeared to recover.

Meprobamate does not affect benzodiazepine or GABA receptors. It seems to act by potentiating

the action of endogenously released adenosine; it blocks reuptake of adenosine, which is itself a

sedative. Caffeine is an adenosine antagonist.

Now that the dust has settled, some 50 years later, it is clear that meprobamate is effective as an

antianxiety agent in the same sense that diazepam or chlordiazepoxide is effective (Cole and

Yonkers 1995), although controlled studies directly comparing the efficacy of meprobamate with

that of benzodiazepines are almost nonexistent.

The clinical dosage of meprobamate is on the order of 400 mg three or four times a day, being

approximately equivalent to 5 mg of diazepam three or four times a day (Table 6–3).

The major side effects are sedation and malcoordination. The drug is relatively safe in

overdose—less lethal than intermediate-acting barbiturates like pentobarbital but a good deal less

safe than diazepam. The drug produces physical dependence and tolerance in much the same way

as do the barbiturates and the benzodiazepines. Significant withdrawal effects, such as

convulsions, agitation, and delirium, occur after clinically relatively lower doses of

meprobamate—for example, after 3,200 mg, or eight 400-mg tablets a day.

Currently, it is hard to identify any unique advantages of meprobamate as an antianxiety drug (Cole

and Yonkers 1995). It is a reasonably effective and satisfactory hypnotic at a dosage of 400–800

mg/day at bedtime. Clinically, we have occasionally seen anxious patients who have a marked

subjective intolerance to benzodiazepines, becoming agitated, dysphoric, and restless while taking

any of several of them. Some of these patients tolerate meprobamate quite well, perhaps because

its action is extremely different from that of benzodiazepines. To our knowledge, no one has

evaluated meprobamate in the treatment of conditions such as akathisia or panic disorder, for

which some benzodiazepines appear to be effective. The introduction of triplicate prescriptions for

benzodiazepines in New York State (essentially placing them in Schedule II) reportedly led to an

acute rise in prescriptions for meprobamate, which remained in Schedule IV—an unfortunate

consequence of legislative action. The drug is less safe than are the benzodiazepines.

Some older patients have been receiving maintenance meprobamate therapy for many years; it

may be clinically sensible for them to continue taking that drug. One study showed that a large

number of elderly patients in Florida judged their meprobamate regimen to be quite helpful (Hale

et al. 1988).

Deprol, a discontinued meprobamate-containing combination medication, deserves mention here.

This combination of benactyzine and meprobamate was marketed for use in depression. Each tablet

contained 400 mg of meprobamate and 1 mg of benactyzine hydrochloride. There is essentially no

evidence that benactyzine by itself is an effective antidepressant, although the compound is

anticholinergic and might conceivably have some antidepressant properties. One trial many years

ago involving patients with schizophrenia showed that benactyzine alone increased hallucinatoryPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

30 of 35

31/01/2009 06:51

and psychotic behavior. However, a handful of studies suggested that Deprol was effective in

treating some depressions and may perhaps be more effective than either of its ingredients alone.

The question remains moot, as the drug is no longer available.

BIBLIOGRAPHY

American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 2nd

Edition. Washington, DC, American Psychiatric Association, 1968

American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 3rd

Edition, Revised. Washington, DC, American Psychiatric Association, 1987

American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th

Edition. Washington, DC, American Psychiatric Association, 1994

American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th

Edition, Text Revision. Washington, DC, American Psychiatric Association, 2000

Baer L, Greist JH: An interactive computer administered self-assessment and self-help program for

behavior therapy. J Clin Psychiatry 58 (12, suppl): 23–28, 1997

Ballenger JC, Burrows GD, DuPont RL Jr, et al: Alprazolam in panic disorder and agoraphobia:

results from a multicenter trial, I: efficacy in short-term treatment. Arch Gen Psychiatry

45:413–422, 1988 [PubMed]

Ballenger JC, Pecknold J, Rickels K, et al: Medication discontinuation in panic disorder. J Clin

Psychiatry 54 (10, suppl):15–21, 1993

Barlow DH, Gorman JM, Shear MK, et al: Cognitive-behavioral therapy, imipramine, or their

combination for panic disorder: a randomized controlled trial. JAMA 238:2529–2536, 2000

Baxter LR, Thompson JM, Schwartz JM, et al: Trazodone treatment response in

obsessive-compulsive disorders correlated with shifts in glucose metabolism in the caudate nuclei.

Psychopathology 20:114–122, 1987 [PubMed]

Benzodiazepine seizures: an update. Int Drug Ther Newsl 24:5–7, 1989

Bloch MH, Landeros-Weisenberger A, Kelmendi B, et al: A systematic review: antipsychotic

augmentation with treatment refractory obsessive-compulsive disorder. Mol Psychiatry

11(7):622–632, 2006 [erratum: Mol Psychiatry 11(8):795, 2006]

Bogan AM, Koran LM, Chuong HW, et al: Quetiapine augmentation in obsessive-compulsive disorder

resistant to serotonin reuptake inhibitors: an open-label study. J Clin Psychiatry 66(1):73–79, 2005

[PubMed]

Braestrup C, Squires RF: Brain specific benzodiazepine receptors. Br J Psychiatry 133: 249–260,

1978 [PubMed]

Bystritsky A, Ackerman DL, Rosen RM, et al: Augmentation of serotonin reuptake inhibitors in

refractory obsessive-compulsive disorder using adjunctive olanzapine: a placebo-controlled trial. J

Clin Psychiatry 65(4):565–568, 2004 [PubMed]

Chao IL: Olanzapine augmentation in panic disorder: a case report. Pharmacopsychiatry

37(5):239–240, 2004 [PubMed]

Cole JO, Yonkers KA: Non-benzodiazepine anxiolytics, in American Psychiatric Press Textbook of

Psychopharmacology. Edited by Schatzberg AF, Nemeroff CB. Washington, DC, American Psychiatric

Press, 1995, pp 231–244

Coupland NJ, Lillywhite A, Bell CE, et al: A pilot controlled study of the effects of flumazenil in

post-traumatic stress disorder. Biol Psychiatry 41: 988–990, 1997 [PubMed]

Davidson JRT: Pharmacotherapy of post-traumatic stress disorder: treatment options, long-term

follow-up, and predictors of outcome. J Clin Psychiatry 61 (5, suppl): 52–56, 2000Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

31 of 35

31/01/2009 06:51

Davidson JR, Dupont RL, Hedges D, et al: Efficacy, safety, and tolerability of venlafaxine

extended-release and buspirone in outpatients with generalized anxiety disorder. J Clin Psychiatry

60:528–535, 1999 [PubMed]

deBeurs E, van Balkom AJ, Lange A, et al: Treatment of panic disorder with agoraphobia:

comparison of fluvoxamine, placebo, and psychological panic management combined with exposure

and of exposure in vivo alone. Am J Psychiatry 152:683–691, 1995

DeVeaugh-Geiss J, Landau P, Katz R: Preliminary results from a multicenter trial of clomipramine in

obsessive-compulsive disorder. Psychopharmacol Bull 25:36–40, 1989 [PubMed]

Fink M: Catatonia: syndrome or schizophrenia subtype? Recognition and treatment. J Neural

Transm 108:637–644, 2001a

Fink M: Treating neuroleptic malignant syndrome as catatonia. J Clin Psychopharmacol 21:121–122,

2001b

Fluvoxamine for obsessive-compulsive disorder. Med Lett Drugs Ther 37(942): 13–14, 1995

Foa EB: Psychosocial treatment of post-traumatic stress disorder. J Clin Psychiatry 61 (5,

suppl):43–48, 2000

Frank JB, Kosten TR, Giller EL, et al: A randomized clinical trial of phenelzine and imipramine for

PTSD. Am J Psychiatry 145:1289–1291, 1988 [PubMed]

Franks P, Harp J, Bell B: Randomized, controlled trial of clonidine for smoking cessation in a

primary care setting. JAMA 262:3011–3013, 1989 [PubMed]

Freeman CP, Trimble MR, Deakin JF, et al: Fluvoxamine vs clomipramine in the treatment of

obsessive-compulsive disorder: a multicenter, randomized, double-blind, parallel group

comparison. J Clin Psychiatry 55:301–305, 1994 [PubMed]

Friedman MJ: Toward a rational pharmacotherapy for post-traumatic stress disorder: an interim

report. Am J Psychiatry 145:281–285, 1988 [PubMed]

Friedman MJ: What might the psychobiology of post-traumatic stress disorder teach us about future

approaches to pharmacotherapy? J Clin Psychiatry 61 (7, suppl): 44–51, 2000

Fyer AJ, Liebowitz MR, Gorman JM, et al: Discontinuation of alprazolam treatment in panic patients.

Am J Psychiatry 144:303–308, 1987 [PubMed]

Gao K, Muzina D, Gajwani P, et al: Efficacy of typical and atypical antipsychotics for primary and

comorbid anxiety symptoms or disorders: a review. J Clin Psychiatry 67(9):1327–1340, 2006

[PubMed]

Gelpin E, Bonne O, Peri T, et al: Treatment of recent trauma survivors with benzodiazepines: a

prospective study. J Clin Psychiatry 57:390–394, 1996 [PubMed]

Gladsjo JA, Rapaport MH, McKinney R, et al: Absence of neuropsychological deficits in patients

receiving long-term treatment with alprazolam-XR for panic disorder. J Clin Psychopharmacol

21:131–138, 2001 [PubMed]

Glassman AH, Stetner F, Walsh BT, et al: Heavy smokers, smoking cessation, and clonidine: results

of a double-blind, randomized trial. JAMA 259: 2863–2866, 1988 [PubMed]

Glover H: A preliminary trial of nalmefene for the treatment of emotional numbing in combat

veterans with post-traumatic stress disorder. Isr J Psychiatry Relat Sci 30:255–263, 1993

[PubMed]

Glue P, Fang A, Gandelman K, et al: Pharmacokinetics of an extended release formulation of

alprazolam (Xanax XR) in healthy normal adolescent and adult volunteers. Am J Ther

13(5):418–422, 2006 [PubMed]Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

32 of 35

31/01/2009 06:51

Goddard AW, Brouette T, Almai A, et al: Early coadministration of clonazepam with sertraline for

panic disorder. Arch Gen Psychiatry 58:681–686, 2001 [PubMed]

Gold MS, Redmond DE Jr, Kleber HD: Clonidine in opiate withdrawal. Lancet 1:929–930, 1978

[PubMed]

Goldberg HL: Buspirone hydrochloride: a unique new anxiolytic agent: pharmacokinetics, clinical

pharmacology, abuse potential, and clinical efficacy. Pharmacotherapy 4:315–324, 1984 [PubMed]

Gorman JM, Levy GF, Liebowitz MR, et al: Effect of acute beta-adrenergic blockade on lactate

induced panic. Arch Gen Psychiatry 40:1079–1082, 1983 [PubMed]

Granville-Grossman KL, Turner P: The effect of propranolol on anxiety. Lancet 1:788–790, 1966

[PubMed]

Greenblatt DJ, Shader RI, Abernethy DR: Current status of benzodiazepines [first of two parts]. N

Engl J Med 309:354–359, 1983a

Greenblatt DJ, Shader RI, Abernethy DR: Current status of benzodiazepines [second of two parts].

N Engl J Med 309:410–415, 1983b

Greenblatt DJ, von Moltke LL, Harmatz JS, et al: Alprazolam pharmacokinetics, metabolism, and

plasma level: clinical implications. J Clin Psychiatry 54 (10, suppl): 4–11, 1993

Grinker R, Spiegel J: Men Under Stress. Philadelphia, PA, Blakiston, 1945

Hale WE, May F, Moore M, et al: Meprobamate use in the elderly: a report from the Dunedin

Program. J Am Geriatr Soc 30:1003–1005, 1988

Heimberg RG: Current status of psychotherapeutic interventions for social phobia. J Clin Psychiatry

62 (1, suppl):36–42, 2001

Hembree EA, Foa EB: Posttraumatic stress disorder: psychological factors and psychosocial

interventions. J Clin Psychiatry 61 (7, suppl):33–39, 2000

Herman JB, Rosenbaum JF, Brotman AW: The alprazolam to clonazepam switch for the treatment of

panic disorder. J Clin Psychopharmacol 7: 175–178, 1987 [PubMed]

Hirschfeld RM, Weisler RH, Raines SR, et al, for the BOLDER Study Group: Quetiapine in the

treatment of anxiety in patients with bipolar I or II depression: a secondary analysis from a

randomized, double-blind, placebo-controlled study. J Clin Psychiatry 67:355-362, 2006

Insel TR (ed): New Findings in Obsessive-Compulsive Disorder. Washington, DC, American

Psychiatric Press, 1984

Insel TR, Murphy DL, Cohen RM, et al: Obsessive-compulsive disorder: a double-blind trial of

clomipramine and clorgyline. Arch Gen Psychiatry 40: 605–612, 1983 [PubMed]

Isbell H, Altschul S, Kornetsky C, et al: Chronic barbiturate intoxication: an experimental study.

AMA Archives of Neurology and Psychiatry 64:1–28, 1950 [PubMed]

Jefferson JW: Social phobia: a pharmacological treatment overview. J Clin Psychiatry 56 (5,

suppl):18–24, 1995

Jenike MA, Baer L, Ballantine T, et al: Cingulotomy for refractory obsessive-compulsive disorder.

Arch Gen Psychiatry 52:384–392, 1995 [PubMed]

Kahn R, McNair D, Covi L, et al: Effects of psychotropic agents in high anxiety subjects.

Psychopharmacol Bull 17:97–100, 1981

Kavirajan H: The amobarbital interview revisited: a review of the literature since 1966. Harv Rev

Psychiatry 7:153–165, 1999 [PubMed]

Keck PE Jr, Strawn JR, McElroy SL: Pharmacologic treatment considerations in co-occurring bipolar

and anxiety disorders. J Clin Psychiatry 67(suppl 1):8–15, 2006Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

33 of 35

31/01/2009 06:51

Klein DF: Importance of psychiatric diagnosis in prediction of clinical drug effects. Arch Gen

Psychiatry 16:118–126, 1967 [PubMed]

Klein DF: False suffocation alarms, spontaneous panics, and related conditions: an integrative

hypothesis. Arch Gen Psychiatry 50:306–317, 1993 [PubMed]

Lader M: Summary and commentary, in Pharmacology of Benzodiazepines. Edited by Usdin E,

Skolnick P, Tallman JF, et al. New York, Macmillan, 1982, pp 53–60

Liebowitz MR, Fyer AJ, McGrath P, et al: Clonidine treatment of panic disorder. Psychopharmacol

Bull 17:122–123, 1981

Liebowitz MR, Gorman JM, Fyer AJ, et al: Social phobia: review of a neglected anxiety disorder.

Arch Gen Psychiatry 42:729–736, 1985 [PubMed]

Liebowitz MR, Fyer AJ, Gorman JM, et al: Phenelzine in social phobia. J Clin Psychopharmacol

6:93–98, 1986 [PubMed]

Lucki I, Rickels K, Geller AM: Chronic use of benzodiazepines and psychomotor and cognitive test

performance. Psychopharmacology (Berl) 88: 426–433, 1986 [PubMed]

Lydiard RB, Falsetti SA: Treatment options for social phobia. Psychiatr Ann 17: 409–423, 1994

Marazziti D, Pfanner C, Dell’Osso B, et al: Augmentation strategy with olanzapine in resistant

obsessive compulsive disorder: an Italian long-term open-label study. J Psychopharmacol

19(4):392–324, 2005 [PubMed]

McDougle CJ: Update on pharmacologic management of obsessive-compulsive disorder: agents and

augmentation. J Clin Psychiatry 58 (12, suppl): 11–22, 1997

McDougle CJ, Naylor ST, Cohen DT, et al: A double-blind placebo-controlled study of fluvoxamine in

adults with autistic disorder. Arch Gen Psychiatry 53:1001–1008, 1996 [PubMed]

Meltzer HY, Fleming R, Robertson A: The effect of buspirone on prolactin and growth hormone

secretion in man. Arch Gen Psychiatry 40:1099–1102, 1983 [PubMed]

Menza M, Harris D: Benzodiazepines and catatonia: an overview. Biol Psychiatry 26: 842–846, 1989

[PubMed]

Mindus P, Jenike MA: Neurosurgical treatment of malignant obsessive-compulsive disorder.

Psychiatr Clin North Am 15:921–938, 1992 [PubMed]

Montgomery SA, Tobias K, Zomberg GL, et al: Efficacy and safety of pregabalin in the treatment of

generalized anxiety disorder: a 6-week, multicenter, randomized, double-blind, placebo-controlled

comparison of pregabalin and venlafaxine. J Clin Psychiatry 67:771–782, 2006 [PubMed]

Mooney JJ, Schatzberg AF, Cole JO, et al: Enhanced signal transduction by adenylate cyclase in

platelet membranes of patients showing antidepressant responses to alprazolam: preliminary data.

J Psychiatr Res 19:65–75, 1985 [PubMed]

Nemeroff CB: Use of atypical antipsychotics in refractory depression and anxiety. J Clin Psychiatry

66 (suppl 8):13–21, 2005

Noyes R, Anderson DJ, Clancy J, et al: Diazepam and propranolol in panic disorder and agoraphobia.

Arch Gen Psychiatry 41:287–292, 1984 [PubMed]

Noyes R Jr, DuPont RL Jr, Pecknold JC, et al: Alprazolam in panic disorder and agoraphobia: results

from a multicenter trial, II: patient acceptance, side effects, and safety. Arch Gen Psychiatry

45:423–428, 1988 [PubMed]

Nutt DJ: The psychobiology of postraumatic stress disorder. J Clin Psychiatry 61 (5, suppl):24–29,

2000

Pande AC, Davidson JR, Jefferson JJ, et al: Treatment of social phobia with gabapentin: aPrint: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

34 of 35

31/01/2009 06:51

placebo-controlled study. J Clin Psychopharmacol 19:341–348, 1999 [PubMed]

Pearlstein T: Antidepressant treatment of post-traumatic stress disorder. J Clin Psychiatry 61 (7,

suppl):40–43, 2000

Pecknold JC, Swinson RP, Kuch K, et al: Alprazolam in panic disorder and agoraphobia: results from

a multicenter trial, III: discontinuation effects. Arch Gen Psychiatry 45:429–436, 1988 [PubMed]

Petrides G, Fink M: Choosing a dosing strategy for electrical stimulation in ECT. J Clin Psychiatry

57:487–488, 1996 [PubMed]

Petrides G, Fink M, Husain MM, et al: ECT readmission rates in psychotic vs nonpsychotic depressed

patients: a report from CORE. J ECT 17:244–253, 2001 [PubMed]

Phillips KA, McElroy SL, Keck PE, et al: A comparison of delusional and nondelusional body

dysmorphic disorder in 100 cases. Psychopharmacol Bull 30:179–186, 1994 [PubMed]

Phillips KA, McElroy SL, Dwight MM, et al: Delusionality and response to open-label fluoxetine in

body dysmorphic disorder. J Clin Psychiatry 62: 87–91, 2001 [PubMed]

Phillips KA, Albertini RS, Rasmussen SA: A randomized placebo-controlled trial of fluoxetine in body

dysmorphic disorder. Arch Gen Psychiatry 59: 381–388, 2002 [PubMed]

Pivac N, Kozaric-Kovacic D, Muck-Seler D: Olanzapine versus fluphenazine in an open trial in

patients with psychotic combat-related post-traumatic stress disorder. Psychopharmacology (Berl)

175(4):451–456, 2004 [PubMed]

Pollack MH, Tesar GE, Rosenbaum JF, et al: Clonazepam in the treatment of panic disorder and

agoraphobias: a one-year follow-up. J Clin Psychopharmacol 6:302–304, 1986 [PubMed]

Pollack MH, Simon NM, Worthington JJ, et al: Combined paroxetine and clonazepam treatment

strategies compared to paroxetine monotherapy for panic disorder. J Psychopharmacol

17(3):276–282, 2003 [PubMed]

Pollack MH, Simon NM, Zalta AK, et al: Olanzapine augmentation of fluoxetine for refractory

generalized anxiety disorder: a placebo controlled study. Biol Psychiatry 59(3):211–215, 2006

[PubMed]

Problems associated with alprazolam therapy. Int Drug Ther Newsl 23:29–31, 1988

Riba J, Rodriguez-Fornells A, Strassman RJ, et al: Psychometric assessment of the Hallucinogen

Rating Scale. Drug Alcohol Depend 62:215–223, 2001 [PubMed]

Rickels K: Alprazolam extended-release in panic disorder. Expert Opin Pharmacother

5(7):1599–1611, 2004 [PubMed]

Rickels K, Schweizer E: Maintenance treatment studies in anxiety disorders: some methodological

notes. Psychopharmacol Bull 31:115–123, 1995 [PubMed]

Rickels K, Chase WG, Downing RW: Long-term diazepam therapy and clinical outcome. JAMA

250:767–771, 1983 [PubMed]

Rickels K, Schweizer E, Csanalosi I, et al: Long-term treatment of anxiety and risk of withdrawal:

prospective comparison of clorazepate and buspirone. Arch Gen Psychiatry 45:444–450, 1988

[PubMed]

Rickels K, DeMartinis N, Rynn M, et al: Pharmacologic strategies in discontinuing benzodiazepine

treatment. J Clin Psychopharmacol 19 (suppl 2): 12S–16S, 1999

Ries RK, Roy-Byrne PP, Ward NG, et al: Carbamazepine treatment for benzodiazepine withdrawal.

Am J Psychiatry 146:536–537, 1989 [PubMed]

Sathananthan GL, Sanghavi I, Phillips N, et al: MJ 9022: correlation between neuroleptic potential

and stereotypy. Curr Ther Res Clin Exp 18:701–705, 1975 [PubMed]Print: Chapter 6. Antianxiety Agents http://www.psychiatryonline.com/popup.aspx?aID=236742&print=yes…

35 of 35

31/01/2009 06:51

Schneier FR: Treatment of social phobia with antidepressants. J Clin Psychiatry 62

(1, suppl):43–48, 2001

Schweizer E, Rickels K, Lucki I: Resistance to the antianxiety effect of buspirone in patients with a

history of benzodiazepine use (letter). N Engl J Med 314:719–720, 1986 [PubMed]

Shagass C, Bittle RM: Therapeutic effect of LSD: a follow-up study. J Nerv Ment Dis 144:471–478,

1967 [PubMed]

Shagass C, Naiman J, Milhalik J: An objective test which differentiates between neurotic and

psychotic depression. Arch Neurol Psychiatry 75:461–471, 1956 [PubMed]

Sheehan DV, Ballenger J, Jacobsen G: Treatment of endogenous anxiety with phobic, hysterical, and

hypochondriacal symptoms. Arch Gen Psychiatry 37:51–59, 1980 [PubMed]

Smith WT, Londborg PD, Glaudin V, et al: Short-term augmentation of fluoxetine with clonazepam

in the treatment of depression: a double-blind study. Am J Psychiatry 155:1339–1345, 1998 [Full

Text] [PubMed]

Solomon SD, Gerrity ET, Muff AM: Efficacy of treatments for posttraumatic stress disorder: an

empirical review. JAMA 268:633–638, 1992 [PubMed]

Southwick SM, Krystal JH, Bremmer JD, et al: Noradrenergic and serotonergic function in

posttraumatic stress disorder. Arch Gen Psychiatry 54:749–758, 1997 [PubMed]

Spiegel DA: Psychological strategies for discontinuing benzodiazepine treatment. J Clin

Psychopharmacol 19 (suppl 2):17S-23S, 1999

Strassman RJ: Hallucinogenic drugs in psychiatric research and treatment: perspectives and

prospects. J Nerv Ment Dis 183:127–138, 1995 [PubMed]

Sutherland SM, Davidson JRT: Pharmacotherapy for posttraumatic stress disorder. Psychiatr Clin

North Am 17:409–423, 1994 [PubMed]

Swedo SE, Leonard HL, Rapoport JL, et al: A double-blind comparison of clomipramine and

desipramine in the treatment of trichotillomania (hair pulling). N Engl J Med 321:497–501, 1989

[PubMed]

Tesar GE, Rosenbaum JF: Successful use of clonazepam in patients with treatment-resistant panic

disorder. J Nerv Ment Dis 174:447–482, 1986

Tyrer PJ, Lader MH: Clinical response to propranolol and diazepam in somatic and psychic anxiety.

BMJ 2:14–16, 1974 [PubMed]

Tyrer P, Shawcross C: Monoamine oxidase inhibitors in anxiety disorders. J Psychiatr Res 22 (suppl

1):87–98, 1988

Udelman HD, Udelman DL: Concurrent use of buspirone in anxious patients during withdrawal from

alprazolam therapy. J Clin Psychiatry 51 (9, suppl): 46–50, 1990

Uhlenhuth EH (ed): Benzodiazepine dependence and withdrawal: myths and management. J Clin

Psychopharmacol 19 (suppl 2):1S–29S, 1999

van der Kolk BA: The drug treatment of post-traumatic stress disorder. J Affect Disord 13:203–213,

1987

van der Kolk BA, Dreyfuss D, Michaels M, et al: Fluoxetine in posttraumatic stress disorder. J Clin

Psychiatry 55:517–522, 1994

Zitrin CM, Klein DF, Woerner MG: Treatment of phobias, I: comparison of imipramine hydrochloride

and placebo. Arch Gen Psychiatry 40:125–138, 1983 [PubMed]

Copyright © 2009 American Psychiatric Publishing, Inc. All Rights Reserved.