Chapter 7 Hypnotics

Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

1 of 20

31/01/2009 06:52

Print Close Window

DOI: 10.1176/appi.books.9781585622825.237638

Manual of Clinical Psychopharmacology >

Chapter 7. Hypnotics

INSOMNIA

Insomnia is a common problem. Various population surveys find self-reported rates of symptoms of

insomnia in about a third of adults. About a tenth of the adults in the United States describe their

sleeping problem as occurring on a regular, chronic basis, and about the same proportion have

sleeping difficulty that qualifies for an insomnia disorder diagnosis.

Most of the medical writing on the subject advises physicians and the public to avoid the regular

use of hypnotics and assumes that the sleepless public will go to physicians or sleep clinics, where

staff will find types of insomnia that respond to behavioral or other nonsedative medical therapies.

Certainly conditions such as sleep apnea, narcolepsy, and restless legs syndrome do occur, and

they can be treated in a variety of ways. Sleep apnea can be life-threatening. Some insomnia may

respond to treatment of the underlying illness (e.g., hyperthyroidism, depression, mania). One U.S.

National Institutes of Health (NIH) Consensus Conference spent a great deal of time worrying

about the possible adverse effects of hypnotics and suggested that, at most, hypnotics be

prescribed for brief periods or be used only every third night. Lasagna (1996), in a cogent editorial

documenting the adverse effects of untreated insomnia on auto accidents, job performance, and

quality of life, noted that in the United States individuals with insomnia rarely consult physicians

and, when they do, rarely mention the insomnia. About 3% of the population regularly takes

over-the-counter hypnotics containing sedative antihistamines or sedative antihistamines plus mild

analgesics. Sales of both types of sleep aids have been escalating, essentially unstudied, for many

years.

We therefore have a situation in which a growing number of prescription hypnotics are available for

prescription use but are rather unpopular with some polysomnographers and sleep experts in

particular and with physicians in general. One gets the odd impression that many individuals with

chronic insomnia make do with diphenhydramine or with wakefulness while more effective

hypnotics are withheld from them. A large-scale study is needed in order to reexamine the

long-term effects of over-the-counter sleeping medications and untreated insomnia as well as of

the chronic use of zolpidem, eszopiclone, or temazepam, possibly the three better of the available

prescription hypnotics.

The availability of ramelteon has altered the discourse on hypnotics somewhat. Ramelteon, a

melatonin agonist, does not have the habit-forming potential of traditional hypnotics and is not

sedating. On the other hand, ramelteon may not be as effective in addressing the spectrum of sleep

difficulties as benzodiazepine and nonbenzodiazepine hypnotics.

When insomnia does lead patients to seek help from a psychiatrist or a sleep disorder specialist,

the patients complain, often despairingly, of very poor sleep or total insomnia, but they

nevertheless appear, on all-night sleep recordings or from nurses’ or relatives’ observations, to be

sleeping most of the night. Attempts to objectify insomnia in terms of the time it takes for the

person to fall asleep or the length of time actually slept have often not proved useful. Further, the

psychiatrist is not uncommonly faced with patients who are actually awake all night, particularly

those with marked depression and anxiety, whose sleep is so disturbed that they feel exhausted

the next day and claim that they “can’t function” or, in reality, cannot function.

Primary care physicians often see patients with untreated and undiagnosed insomnia. The presence

of two of the treatable sleep disorders can be guessed at by the physician’s asking questions of the

patient and his or her bed partner. Sleep apnea is likely present when the partner notes the patient

stops breathing in the night then snorts and gasps; snoring is also likely. The patient may notePrint: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

2 of 20

31/01/2009 06:52

daytime sleepiness instead of wakefulness at night. Restless legs syndrome is subjectively obvious

to the patient as a discomfort in his or her limbs that is relieved by movement, much like akathisia.

The restlessness is worse later in the day and can disturb sleep or can disturb the bed partner,

since jerky leg movements at night feel like kicks. Periodic leg movement disorder is diagnosed

when the leg movements occur mainly during sleep and may cause awakenings.

Falling asleep during the day can be a symptom of narcolepsy and responds well to older stimulants

and to modafinil (Provigil), the new French drug now approved for use in the United States in

treating narcolepsy.

However, most patients coming to psychiatrists or psychiatric clinics with insomnia will already be

taking medication. In these cases, the clinical problems are as follows:

1. Is a diagnosable, special sleep disorder being missed? Is referral to a sleep clinic indicated?

Are any of the patient’s prescribed or self-administered medications (selective serotonin reuptake

inhibitors [SSRIs], bronchodilators, caffeine, cocaine) causing the problem, or is the problem due to a

medical condition (i.e., chronic pain, urinary frequency)?

2.  
3. Is the insomnia due to an undertreated or recurrent psychiatric condition?

If the patient is already taking a hypnotic and it is (or is not) working, should the medication be

continued or stopped, and if the medication is stopped, should a new medication be started?

4.  

Among psychiatric patients, troublesome insomnia is often one feature of a wider symptom

complex, which may also include diurnal rhythm disturbances of activity, mood, and so forth.

Depressed patients classically experience early-morning awakening and diurnal variation. Many,

however, have marked difficulty falling asleep and staying asleep as well. Patients with pronounced

initial insomnia may be separated into those who sleep fitfully and poorly even after falling asleep

and those who sleep well even after several hours of initial insomnia. Such patients can sleep from

4:00 A.M. until noon. A few patients with very severe depression complain of almost complete

insomnia, reporting that they lie in bed all night without sleeping, often experiencing dysphoric,

miserable ruminations.

Other psychiatric conditions are also manifested by sleep disturbances. Patients with anxiety

disorders are more likely to have trouble falling asleep. Patients with mania or hypomania may stay

up all night, either happily overactive or dysphorically agitated, as may patients with schizophrenia

or schizophreniform psychoses. Classically, manic patients do not necessarily complain of

experiencing sleeplessness; rather, they may claim that they are not sleeping much and do not

seem to miss the usual sleep they get. Patients with dementia may become more confused and

agitated toward evening (sundowning) and may also be agitated at night after sleeping all day.

Fatigue is often a factor in sundowning. Patients very upset by major stressors—for example,

bereavement, rejections, or physical trauma—may have insomnia as part of an acute stress

response. Patients with posttraumatic stress disorder (PTSD) are often afraid of falling asleep and

thus have initial insomnia.

Insomnia in all the relatively acute psychiatric conditions mentioned above may be fairly

straightforward to treat. Insomnia as part of a depressive disorder often responds to a standard

antidepressant; in fact, some of the more sedating antidepressants (mirtazapine, amitriptyline,

doxepin, trazodone) may be useful hypnotics even in the absence of depression (see “Sedative

Antihistamines and Other Nonbenzodiazepine Psychoactive Drugs With Hypnotic Properties”

section later in this chapter). For depression with poor sleep, it is sensible to begin with a sedative

antidepressant (see Chapter 3: “Antidepressants”), but antidepressants with more stimulant

effects, such as bupropion and fluoxetine, usually improve sleep as the depressive syndrome

improves. Protriptyline was once reported to be a good treatment for sleep apnea. In a study at

McLean Hospital of more than 100 depressed patients taking fluoxetine, we found that insomnia at

intake was not associated with a poorer clinical response to fluoxetine, even though fluoxetine can

cause insomnia as a side effect. In cases of manic and schizophrenic excitement and in some cases

of organic agitation, the insomnia responds well to antipsychotic drugs of any class, although onePrint: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

3 of 20

31/01/2009 06:52

of the newer atypical antipsychotics, chlorpromazine, or loxapine may initially be more effective as

a hypnotic before the general syndrome ameliorates.

Thus, the best general approach is to treat the psychiatric condition that underlies the insomnia

with drugs that are appropriate to that condition rather than first prescribing a benzodiazepine or

other hypnotic drug for the insomnia and then treating the depression or psychosis. If a patient

should be taking a drug that turns out to aggravate or cause insomnia, short-term use of a hypnotic

or a drug like trazodone or mirtazapine may be useful.

In principle, a psychiatrist can run an inpatient ward or treat outpatients without ever prescribing a

hypnotic. In practice, however, life often is (or appears to be) more complicated. Other patients,

families, and nursing staff members become very upset if patients do not retire to bed and sleep

without disturbing them. Often, hypnotics end up being prescribed to reduce the patient’s distress

and distress in his or her milieu. Such prescribing is often justifiable but can pose problems. First,

the hypnotic may not put the patient to sleep, but instead may leave the patient (unpredictably)

groggy, confused, and even more agitated. Second, a long-acting hypnotic, such as flurazepam,

may leave the patient groggy the next day. Third, once the patient is used to getting a hypnotic

(and the doctor is used to prescribing it), the practice may continue for weeks or even months, long

after the initial phase of the illness or the stress of hospitalization has passed. When the sleeping

pill is finally stopped, rebound insomnia is likely to occur, and hypnotic use is likely to be resumed

to control it. (Rebound insomnia is a transient worsening of sleep disturbance occurring soon after

discontinuation of the hypnotic.) In fact, with short-acting hypnotics such as triazolam (or alcohol),

rebound insomnia can occur even 4 hours after the medication is taken. Such an effect can

precipitate overuse of the compound: some patients will take additional doses to help them fall

back to sleep.

Another problematic situation involves the new patient being admitted to the hospital (or visiting a

new psychiatrist) who is thoroughly accustomed to taking 10 mg of zolpidem, 60 mg of flurazepam,

or 3 mg of eszopiclone at bedtime (having taken the medication for months or years) but who still

complains of poor sleep. As might be expected, the quality of sleep is reportedly much worse if the

hypnotic is stopped. Here, two obvious options are available: to taper the hypnotic gradually while

treating the major psychiatric disorder or to continue it while dealing with the disorder. The

difficulty with the second, interpersonally easier, option is that the hypnotic may never be

discontinued. A more complicated version of this scenario is when the patient, at the time he or she

comes to the psychiatrist for treatment, is taking several psychoactive drugs of different classes, all

of which need to be withdrawn or changed. In this situation, it is common to leave the hypnotic at a

stable dose and taper and stop the other drugs first. This method avoids combining, and thus

confusing, the effects of stopping the hypnotic with the already potentially complex effects of

discontinuing an antidepressant or an antipsychotic.

At present, the only widely approved indications for hypnotics in treating insomnia are 1) brief (3-

to 7-day) use or 2) occasional use for transitory insomnia caused by either acute life stressors or

major shifts in diurnal rhythm (as in jet lag or in changing from one work shift to another). One

hypnotic, eszopiclone (Lunesta), has been approved for more chronic use.

Although it is wise to avoid prolonged hypnotic use, prolonged use of a benzodiazepine hypnotic

may be less harmful than is often said and may, in fact, provide some benefit. Sleep loss is widely

believed to sometimes precipitate manic or depressive episodes. The available evidence from sleep

laboratories suggests that hypnotics improve sleep measurably for only a few weeks, though

placebo-controlled studies have shown no diminution in the efficacy of hypnotics for up to 24

weeks, and single-blind studies show efficacy up to a year. So far, no studies have clearly

established a treatment duration after which benzodiazepine hypnotics fail to “work.” Many

confirmed users of hypnotics swear that the hypnotics are always helpful and even vitally

necessary for years. Perhaps the initial sedation at peak blood levels provides a familiar,

conditioned cue conducive to falling asleep. There is also evidence that benzodiazepines cause the

patient to forget episodes of wakefulness; when the patient is no longer taking benzodiazepines,Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

4 of 20

31/01/2009 06:52

sleep may objectively be no worse, but he or she perceives (recalls) it as being worse. Certainly,

some patients who have taken a hypnotic every night for several years continue to complain of

poor sleep and daytime fatigue and dysphoria for months after the hypnotic was stopped.

The current evidence is that triazolam, zolpidem, zaleplon, and eszopiclone continue to be effective

(i.e., improve sleep as measured by polysomnographic criteria) for 4–5 weeks in placebo-controlled

studies and for as long as 6 months in single-blind studies. Studies using other, less formal criteria

(subject or observer ratings of sleep) have shown benefit for up to a year. Data are not available

beyond that point. Since earlier editions of this manual, sleep experts appear to have become more

permissive about the possible benefits of longer-term use of benzodiazepine-like hypnotics (Buysse

and Dorsey 2002). However, package inserts still limit the use of such hypnotics to durations of use

validated by placebo-controlled studies. The exception to this admonition is eszopiclone, which is

approved for both acute and maintenance treatment of insomnia. Patients sometimes interpret

such restrictions to mean that serious toxicity has been shown to occur if the hypnotic is taken for

longer than specified in the Physicians’ Desk Reference (PDR). Such an interpretation does not

seem to be accurate. Oversedation, poor coordination, and memory problems are more likely to

occur very early in treatment. The evidence for serious behavioral or other toxicity (e.g., impaired

psychomotor performance early in treatment, auto accidents, falls in the elderly) is mixed and

inconclusive (Buysse and Dorsey 2002).

Behavioral Approaches

Cognitive-behavioral therapy is one of the most consistently effective approaches to the long-term

management of primary insomnia (Edinger and Means 2005). The more clinically and scientifically

validated (nonpharmacological) approaches to insomnia are as follows:

Stimulus control therapy (sleep hygiene). The patient learns to associate being in bed only with sleep

(or sex) by getting out of bed when he or she does not fall asleep rapidly in order to avoid linking bed

and bedroom with miserable sleepless nights.

1.  

Sleep restriction therapy. The patient reduces time in bed progressively (with daytime naps forbidden)

until he or she is asleep for 80%–90% of the time he or she is allowed to be in bed.

2.  

Continuous positive airway pressure (for patients with sleep apnea). The patient learns to wear a mask

that delivers light, constant positive air pressure to the pharynx and thereby prevents the choking spells

from occurring.

3.  

Bright light therapy. A patient who sleeps an adequate amount but at the wrong time of the night or day

can have his or her sleep-wake rhythm reset through exposure to a light box.

4.  

Obviously, behavioral approaches require patient involvement and cooperation and are best

managed by experts in sleep problems.

Pharmacological Approaches

For the past 30 years, the major efforts to develop better hypnotics have focused entirely on

chemicals that bind to at least some form of benzodiazepine receptor. These agents are hoped to

turn out to be better than, say, temazepam by being faster acting, being less likely to cause a

morning hangover, not being “abusable” or “addictive,” and not causing any change in

polysomnographic sleep records, or, perhaps, causing only an increase in stage 3 and rapid eye

movement (REM) sleep.

So far, this semisystematic approach has yielded a plethora of benzodiazepines, one of which,

quazepam, may be selective for benzodiazepine1 receptors, and three agents (zolpidem, zaleplon,

and eszopiclone) that are active at benzodiazepine receptors but are not chemically

benzodiazepines. So far, none of these drugs have been shown to be remarkably different in effect

from the older benzodiazepine hypnotics, though zolpidem, eszopiclone, and zaleplon may be a bit

better than triazolam in that they are less likely to cause rebound insomnia, are less likely to

exacerbate obstructive airway problems through muscle relaxation, and may show continued

efficacy, when given nightly, for much longer.Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

5 of 20

31/01/2009 06:52

In the interim, in fact for decades, drugs developed for other purposes end up being used for

insomnia, at least by psychiatrists. We suspect that trazodone in doses of 50–200 mg hs is used

more frequently for insomnia than for its original, U.S. Food and Drug Administration

(FDA)–approved indication, depression. Unfortunately, trazodone’s use for insomnia began only as

the company’s patent was expiring, so no systematic or extensive studies of its effects on sleep and

insomnia have been done. However, most clinicians agree that it works. Its problems are morning

drowsiness and, conceivably though very rarely, priapism. If tolerance to its hypnotic effect occurs,

it develops only after months or years. The drug has no known abuse liability and has no clear

withdrawal symptoms. Nefazodone, a related agent, also may improve sleep but, in our clinical

judgment, is a much weaker hypnotic than trazodone and carries the risk of hepatotoxicity. For this

reason, nefazodone is rarely used now as an antidepressant and is almost never used as a hypnotic.

Among the older tricyclic antidepressants (TCAs), amitriptyline, imipramine, doxepin, and

trimipramine, and probably nortriptyline, are good hypnotics, though they are rarely studied as

such. Katz et al. (1991), in a report from the National Institute of Mental Health (NIMH)

Collaborative Depression Study, noted that sleep ratings improved strikingly in the first 2 weeks

with amitriptyline but that this hypnotic effect did not predict an overall improvement in the

depressions being treated. Trimipramine is the best studied and appears to be the only TCA to

facilitate sleep but not to affect REM latency or nocturnal penile tumescence. If it did not

sometimes cause morning drowsiness, dry mouth, constipation, and weight gain, trimipramine

might be the ideal hypnotic at doses of 25–100 mg at bedtime. It is not safe for use by suicidal

patients.

Since gabapentin was released for use in seizures, its use has widely spread into treating pain,

bipolar disorder, anxiety, and insomnia. It seems quite safe in doses from 100 mg to 900 mg at

bedtime and maybe higher. It is not metabolized by the body and seems quite safe in overdose,

since the intestine appears not to be able to absorb more than about 900 mg at any one time. It

appears to be free of abuse liability. For this reason, gabapentin is sometimes used in patients with

alcohol dependence as an alternative to benzodiazepines. One open study found that in alchoholic

patients with sleep disturbance, gabapentin appeared more effective than trazodone as a hypnotic

(Karam-Hage et al. 2003). Again, although commonly prescribed by neurologists and psychiatrists

in treating insomnia, gabapentin is not approved by the FDA for this indication.

All these nonbenzodiazepines (as well as diphenhydramine and maybe melatonin and valerian) may

be nonabusable, relatively safe and effective ways to treat insomnia, but prescribing clinicians

would sleep better if someone—perhaps the NIH or, in the case of gabapentin, the drug

company—would undertake the studies necessary to document their potential safety and efficacy, if

such proves to be the case.

BENZODIAZEPINE HYPNOTICS

The benzodiazepines are the most widely prescribed sedative-hypnotics in the United States today.

Although most benzodiazepines have hypnotic properties, as of this writing, only five

benzodiazepines have an FDA-approved indication as a hypnotic: flurazepam, temazepam,

estazolam, quazepam, and triazolam. Four nonbenzodiazepines, zolpidem, zaleplon, eszopiclone,

and ramelteon, are also FDA approved as hypnotics.

The principles for discriminating among these drugs are similar to those described in Chapter 6

(“Antianxiety Agents”) for the anxiolytic benzodiazepines: structure, pharmacokinetics, absorption,

and distribution (Table 7–1). For example, most of the benzodiazepine hypnotics belong to

separate structural subclasses: 2-keto (flurazepam), 3-hydroxy (temazepam), and triazolo

(triazolam and estazolam) (Figure 7–1).

Table 7–1. Benzodiazepine hypnotics

Generic

name

Brand

name

Formulation

and

strengths

Dosage

(mg/day)

Absorption Major active

metabolites

Approximate

half-life (hrs)Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

6 of 20

31/01/2009 06:52

Generic

name

Brand

name

Formulation

and

strengths

Dosage

(mg/day)

Absorption Major active

metabolites

Approximate

half-life (hrs)

flurazepama

Dalmane Capsules: 15,

30 mg

15–30 Intermediate hydroxyethylflurazepam 1

desalkylflurazepam 100

temazepama Restoril Capsules: 7.5,

15, 22.5, 30

mg

15–30 Intermediate None 8

quazepam Doral Tablets: 7.5,

15 mg

7.5–15 Intermediate oxoquazepam 39

desalkyloxoquazepam 73

triazolama

Halcion Tablets:

0.125, 0.25

mg

0.125–0.5 Intermediate — 3

estazolama

ProSom Tablets: 1, 2

mg

1–4 Intermediate — 16

aAvailable in generic form.

Figure 7–1.Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

7 of 20

31/01/2009 06:52

Chemical structures of benzodiazepine hypnotics.

The metabolism and half-lives of these subclass members parallel those of their anxiolytic

counterparts. Flurazepam is oxidized in the liver and, like diazepam, has a relatively long half-life

(40 hours). It also forms a long-acting (100 hours) metabolite, desalkylflurazepam, a metabolite

shared with quazepam. Temazepam is conjugated with a glucuronide radical in the liver and has a

much shorter half-life (8 hours) and no active metabolites. Triazolam and estazolam are oxidized

but with no clearly active metabolites; triazolam has an extremely short half-life (3–6 hours)

(Table 7–1), while estazolam is more like temazepam. Flurazepam and triazolam have more rapid

absorption (peak blood levels at 30 and 20 minutes, respectively) than temazepam, which may not

be absorbed for 45–60 minutes. The slower absorption accounts for patients’ not falling asleep

rapidly after taking this medication. Clinicians should advise patients to take temazepam

approximately 1 hour before bedtime to avoid anticipatory discomfort and their resorting to

premature repeat dosing.

A benzodiazepine hypnotic, quazepam, was released in the early 1990s. Chemically, it does not fall

into any of the three classes described in Chapter 6 (“Antianxiety Agents”). However, the drug’s

pharmacokinetics are in the same range as those of flurazepam, and it is metabolized to

desalkylflurazepam in the body; the drug and its metabolite have half-lives of about 40 hours.Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

8 of 20

31/01/2009 06:52

Metabolism and elimination are slowed in elderly patients. This pattern suggests that sedation the

next morning should be more of a problem than early rebound insomnia. Quazepam’s only special

feature is that it binds selectively to type 1 benzodiazepine receptors, whereas flurazepam and

other older benzodiazepines bind equally to both type 1 and type 2 sites in the brain. There is no

evidence that this special binding property is of particular clinical value. Quazepam’s active

metabolite does bind to both receptor types. Quazepam is a long-acting hypnotic, probably similar

to flurazepam in its clinical utility. It is available in 7.5-mg and 15-mg tablets.

The distribution of the sedative benzodiazepines is relatively rapid. Indeed, the acute formation of

high peak blood levels is thought to account for patients’ falling asleep relatively quickly with these

drugs. The decline from peak levels may account for why patients can awaken often, even when

significant drug plasma levels are still apparent. Moreover, the decline from peak levels and the

need for peak levels to induce sleep also explain why patients treated with longer-acting

benzodiazepines like flurazepam require nightly doses.

Some investigators and clinicians have argued that a non-long-acting compound (e.g., temazepam

or zolpidem) offers great advantages over longer-acting agents, because it is largely excreted

before the next morning. Although at least one study has pointed to the safe use of triazolam for

inducing sleep on transatlantic flights and preventing jet lag, there may be disadvantages

associated with very-short-acting hypnotics, particularly rebound insomnia and anterograde

amnesia. With some short-acting hypnotics, rebound insomnia occurs within the first 2 nights after

discontinuation and can be troublesome for patients. Although acutely apparent with shorter-acting

compounds, it may occur 5–7 days after discontinuation of benzodiazepines with longer half-lives

and may be misinterpreted as a reemergence of the underlying insomnia rather than as a rebound

phenomenon. It seems less likely to occur with zolpidem and zaleplon. When rebound insomnia

occurs, clinicians should avoid resumption of the drug. Rather, reassurance and the prescription of

antihistamine sedatives (e.g., 50 mg of diphenhydramine), sedating antidepressants (e.g., 50–100

mg of trazodone), or gabapentin (100 or 300 mg hs) for a few days may prove beneficial. When this

does not work, reinstituting the benzodiazepine and gradually tapering the dose is an alternative

strategy.

The side effects of the sedative-hypnotic benzodiazepines are similar to those of their anxiolytic

counterparts. They include sedation, ataxia, anterograde amnesia, slurred speech, and nausea.

These side effects are not particularly dangerous, although sedation can be a problem if the

individual attempts to drive, to operate heavy machinery, and so forth. There is weak evidence to

suggest that the longer-half-life compounds cause clinically important cognitive problems the

following day; however, this area has not been well studied. Refined and detailed psychological

testing can usually detect impairment of cognitive or psychomotor function the morning after a

benzodiazepine hypnotic has been used to induce sleep. The real question is whether this

impairment is clinically important; the vast majority of users of hypnotics do not note behavioral

impairment the next day unless they note persistent sedation as well. However, as noted earlier,

there have been several reports of pronounced anterograde amnesia (“blackouts”) in patients

treated with triazolam. Indeed, some healthy individuals who have used triazolam to induce sleep

during a transatlantic flight have described not remembering arriving in Europe or what they did

the next day. In the few personal accounts from patients and their friends and colleagues of

blackouts after taking a benzodiazepine, the individuals appear to have done their usual activities

blamelessly during the period they cannot recall. Similar reactions have been observed with other

benzodiazepines and with zolpidem and zaleplon. By far, the greater number of published or FDA

reports of anterograde amnesia concern triazolam, but it is difficult to tell whether this difference is

real or due to the extensive media publicity about triazolam (Bunney et al. 1999).

One possible problem with longer-acting benzodiazepines is illustrated by the case of a patient who

had been treated with flurazepam nightly for years but who was shifted to an equivalent dose of

temazepam. He developed a clear benzodiazepine withdrawal syndrome; the shorter-acting drug

could not adequately replace the longer-acting one for the full 24 hours between doses.Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

9 of 20

31/01/2009 06:52

In addition to pharmacologically explainable differences in shifting from one hypnotic to another,

some patients claim (vigorously) that, say, lorazepam works beautifully, whereas temazepam in its

new formulation does nothing. Beyond this, zolpidem or zaleplon is often rejected or criticized by

patients who are used to triazolam or flurazepam; in these cases, these agents may be just

different enough in their subjective effects from benzodiazepines to leave the patient worried that

the hypnotic “isn’t working,” even though these same agents do fine as hypnotics in patients with

insomnia who have never taken a benzodiazepine. The newer drugs affect mainly sleep latency;

they may get the patient to sleep but may not keep him or her sleeping for long enough. Zolpidem,

zaleplon, and eszopiclone appear to be different enough from benzodiazepines to be free of

rebound insomnia but may not have less danger of abuse than older benzodiazepines (Rush and

Griffiths 1996; Rush et al. 1999).

Time (and further studies) will tell, but these newer, not-quite-a-benzodiazepine hypnotics seem

effective and well tolerated by both young and old. They deserve the benefit of the doubt as being

safer than chloral hydrate or pentobarbital, probably better than the benzodiazepine hypnotics, and

probably more effective than the antihistamine over-the-counter Benadryl types of preparations.

Melatonin, the newly advertised, naturally occurring sleep-inducing hormone, and valerian, an old

herbal preparation, are discussed at the end of this chapter.

NONBENZODIAZEPINE HYPNOTICS

Drugs such as zolpidem, zaleplon, and eszopiclone lack the chemical structure of the

benzodiazepines (Figure 7–2). However, these drugs still act on type 1 benzodiazepine receptors.

The nonbenzodiazepine hypnotics lack activity at type 2 benzodiazepine and peripheral type 3

benzodiazepine receptors. Thus, the nonbenzodiazepine hypnotics lack the muscle-relaxing and

anticonvulsant properties of the benzodiazepines. The specific hypnotic effects of these drugs have

distinct advantages in some patients. For example, the muscle-relaxing properties of

benzodiazepines may further compromise the airways of patients with sleep apnea. Likewise, we

prefer that patients undergoing electroconvulsive therapy (ECT) use a nonbenzodiazepine hypnotic

rather than a benzodiazepine hypnotic in the evening prior to ECT because a nonbenzodiazepine is

somewhat less likely to raise the seizure threshold.

Figure 7–2.Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

10 of 20

31/01/2009 06:52

Chemical structures of the nonbenzodiazepine hypnotics zolpidem, zaleplon, and eszopiclone.

The nonbenzodiazepines have a reputation of being less addictive than benzodiazepines. Even

though high doses of the nonbenzodiazepines may induce euphoria (Gelenberg 2000), very few

cases of dependence have been described in the world literature. In a survey of the scientific

literature, only 36 cases of zolpidem dependence and 22 cases of zopiclone (related to eszopiclone)

dependence had been described by 2003 (Hajak et al. 2003). While this type of survey significantly

underreports the true incidence of dependence, the nonbenzodiazepines appear to be less

problematic in this regard than the benzodiazepines.

At the time of this writing, zolpidem accounts for most of the dollars spent on hypnotics in the

United States. There are probably more prescriptions written for benzodiazepine hypnotics, but

these are available in generic form and less expensive. While all nonbenzodiazepines have short

half-lives in healthy subjects (zaleplon, about 1–2 hours; zolpidem, 2–4 hours; eszopiclone, 4–5

hours), many patients taking these medications, particularly zolpidem and eszopiclone, are able to

sleep through the night. Zaleplon is somewhat lighter, and some patients complain of not being

able to fall or stay asleep as well as they do with zolpidem. However, zaleplon is the only hypnotic

that one could effectively take in the middle of the night and not feel too groggy in the morning.

The most common side effects of the nonbenzodiazepines are drowsiness and dizziness. Other

benzodiazepine side effects, such as ataxia, slurred speech, and amnestic symptoms, are less

common with the nonbenzodiazepine hypnotics. However, withdrawal symptoms, including

insomnia, muscle cramps, and even seizures, have been (rarely) reported.

The nonbenzodiazepine hypnotics should be taken on an empty stomach to induce more rapid onset

of sleep. Zolpidem is usually dosed at 5–10 mg at night. Some patients seem to need 20 mg,

although there is little evidence that this dose is much more effective than 10 mg. But there is little

evidence it presents more of a problem than smaller doses. Zaleplon can be taken in doses of 10–20

mg at night. If the patient wakes up in the middle of the night, a 10-mg dose may be taken.

Eszopiclone has usually been initiated at a dose of 1–2 mg to facilitate falling asleep and 3 mg to

sustain sleep if 2 mg is inadequate. In the elderly, the recommended dosage is 1–2 mg/day. Unlike

most hypnotics, eszopiclone has been studied in the management of insomnia for periods of 12

months with no substantial evidence of tolerance.

Melatonin Receptor Agonist: Ramelteon

While melatonin has been available over the counter for many years as a potential treatment of

insomnia and phase-shift disorders, melatonin supplements are not subject to careful and

consistent manufacturing. In addition, the efficacy of most melatonin supplements appears to

range from modest to minimal. Ramelteon (Rozerem) is a specific melatonin MT1 and MT2 receptor

agonist that was approved in 2005 as the first FDA-approved hypnotic that is not a controlled

substance. Ramelteon appears to act on melatonin receptors in the suprachiasmatic nucleus

responsible for sleep-wake cycles. Since it has no effects on the benzodiazepine receptor, it is not

associated with abuse potential, rebound insomnia, motor deficits, or exacerbation of problems

such as chronic obstructive pulmonary disease or obstructive sleep apnea. In addition, it is not

likely to contribute to confusion or memory in geriatric patients in a way that benzodiazepine or

nonbenzodiazepine agents can. On the other hand, the primary utility of ramelteon is in reducing

sleep latency or the time it takes to fall asleep; it may be of less utility than standard hypnotics in

keeping patients asleep. Single-dose and repeated-dose studies have indicated that 8 mg of

ramelteon decreases the time to persistent sleep relative to placebo and might increase total sleep

time (Erman et al. 2006). Preliminary studies also indicate that ramelteon might be useful in

shifting disorders associated with phase delay of the sleep cycle, such as in jet lag, insomnia in the

elderly, and insomnia associated with other neuropsychiatric disorders (McGechan and Wellington

2005).

The primary side effects of ramelteon have included headache, somnolence, fatigue, dizziness, and

nausea, but these have generally been mild and may improve after several weeks of treatment.Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

11 of 20

31/01/2009 06:52

High-fat meals can reduce absorption, and thus patients should be informed not to take ramelteon

shortly after a rich meal. Ramelteon is metabolized via cytochrome P450 3A4, and serum levels

may be increased by fluvoxamine and decreased by carbamazepine.

It is not an unreasonable question to ask what the benefits of ramelteon are over less-expensive

melatonin supplements. Certainly, the consistency of the manufacturing of ramelteon over the

much more variable manufacturing of over-the-counter food supplements such as melatonin is an

advantage. The potency of ramelteon might be expected to also be greater than that of

over-the-counter melatonin, but this has not been established, and no direct comparisons of

ramelteon and over-the-counter melatonin have been reported. Nonetheless, ramelteon has a

unique niche among currently available hypnotics as a well-tolerated agent with no abuse

potential. In patients with a history of substance abuse, geriatric patients, and patients with initial

insomnia, ramelteon might be worth trying before other agents. Our general sense is that

ramelteon may not be as consistently effective as more traditional hypnotics but is far less likely to

cause mischief as well.

SEDATIVE ANTIHISTAMINES AND OTHER NONBENZODIAZEPINE

PSYCHOACTIVE DRUGS WITH HYPNOTIC PROPERTIES

Hydroxyzine compounds are the only antihistamines with some documented efficacy in the

treatment of anxiety disorders. They also have an indication for preoperative and postanesthesia

sedation. These drugs are available in capsules or tablets ranging from 10 to 100 mg each.

However, our clinical experience with psychiatric patients suggests that the drugs are neither much

appreciated by patients nor particularly effective on the few occasions when we have tried them.

They are reasonably free of side effects, although hydroxyzine and the other antihistamines exert

anticholinergic effects. When taken with other anticholinergic agents, hydroxyzine and related

compounds can pose problems, particularly at high doses. Their main value may be as a delaying

action for patients who are inclined to abuse sedative-hypnotic or benzodiazepine drugs, because

they do not produce either physical or psychological dependence.

Diphenhydramine (Benadryl) is another antihistamine sometimes used for its sedative or alleged

hypnotic effects. It has not been well studied in either capacity but has some sedative properties

and is occasionally judged by patients to be acceptable. There is a well-controlled study showing

some efficacy for diphenhydramine in treating insomnia (Rickels et al. 1983). The dose for sleep is

50–100 mg. Diphenhydramine is now available over the counter in 25-mg formulations. The drug is

anticholinergic and can be used for acute dystonic reactions to antipsychotics (see Chapter 4:

“Antipsychotic Drugs”). Promethazine (Phenergan) is a phenothiazine without antipsychotic

properties that is marketed as an antihistamine with sedative properties. Again, it is occasionally

found useful as a mild sedative at doses of 25–100 mg, but it is not a major psychiatric drug.

Doxylamine, another antihistamine, is the ingredient used in most over-the-counter preparations

(e.g., Unisom) that are used for either tranquilization or hypnosis.

Several TCAs exert marked antihistaminergic effects and are excellent hypnotics. So is trazodone, a

heterocyclic antidepressant. Doxepin, amitriptyline, and trimipramine can be used in doses of

25–75 mg at bedtime. Trazodone is best begun, as a hypnotic, at a dose of 50 mg hs, with 50-mg

increments up to 200 mg hs. At some dose in that range improved sleep should occur, perhaps with

undesirable sedation on the following morning. Some patients lose this side effect if the dose is

held steady for 3–5 days; others require a lower dose that, it is hoped, will be effective without

inducing a hangover. Sometimes adult patients taking trazodone on an empty stomach will faint

from postural hypotension if they do not go to bed soon after taking it as a hypnotic. Occasionally,

elderly patients develop orthostatic hypotension the next day after taking this agent at bedtime.

Some studies of trazodone’s effects on sleep and of its utility in other unexpected areas (agitation

in the elderly, if given in 25-mg doses; PTSD [Hertzberg et al. 1996]) have emerged. It retains, so

far, its record with us of mixing innocuously with all other psychiatric drugs, including SSRIs and

monoamine oxidase inhibitors (MAOIs). There is even a study showing that vis-à-vis triazolam or

zolpidem, trazodone has no euphoriant effects in individuals with a history of substance abuse

(Rush et al. 1999). Mirtazapine can also be an effective hypnotic at dosages of 7.5–15 mg/day. ThePrint: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

12 of 20

31/01/2009 06:52

drug is a potent serotonin2 (5-HT2) receptor antagonist and has antihistaminic properties as well.

Gabapentin and pregabalin are probably also as safe and useful as trazodone in treating insomnia.

They have the added potential benefit of having both analgesic and hypnotic effects, as shown in

large placebo-controlled studies involving patients with postherpetic or diabetic neuralgia, and are

reported to be potentially useful in the treatment of other painful conditions that interfere with

sleep. The dose to be used is more of a problem, with the dosage ranging between 300 mg and

1,200 mg taken at bedtime for gabapentin and between 150 and 600 mg/day for pregabalin.

-Aminobutyric acid (GABA) analogs such as gabapentin and pregabalin appear to be more useful for

insomnia or anxiety symptoms in psychiatric patients than as a primary mood stabilizer, although

occasional patients with bipolar disorder appear to do uniquely well while taking gabapentin (see

Chapter 5: “Mood Stabilizers”) (Ketter et al. 1999). There is one disturbing report describing

withdrawal symptoms (anxiety, agitation, and dysphoria) that were worse than those occurring at

baseline (Cora-Locatelli et al. 1998), but no one we know of locally has observed or heard of such a

phenomenon. Again, gabapentin has been mixed with a wide variety of other psychoactive drugs,

mainly in treating bipolar patients, without evidence of adverse interactions (Chouinard et al.

1998).

TCAs may also be effective as hypnotics (Gursky and Krahn 2000) and, like gabapentin and

pregabalin, may have analgesic properties as well if given a prolonged trial in patients with chronic

pain and insomnia. Most TCAs, as well as most other antidepressants, suppress or delay REM sleep.

Nefazodone and mirtazapine do not appear to do so, and the importance of suppressing (or

encouraging) various sleep phases in depressed patients is still unclear despite many studies.

Trimipramine, among the TCAs, may be the least likely to affect sleep architecture, but it may cause

dry mouth, constipation, and weight gain. It seems less likely to cause orthostatic hypotension or

to affect penile tumescence than other TCAs in doses of 25–100 mg hs. Independent evidence that

TCAs (at least amitriptyline and imipramine) can improve sleep without relieving depression comes

from the NIMH Collaborative Depression Study (Katz et al. 1991).

It is hard to make clear recommendations about the use of benzodiazepine or related hypnotics

versus trazodone, mirtazapine, gabapentin, or pregabalin. Clearly, patients with an abuse potential

may be at greater risk for dependence while taking benzodiazepines and related agents. Those

patients with significant anxiety or depression might do better when an agent such as pregabalin or

mirtazapine is used to also help sleep. The TCAs are less safe in overdose than other potential

hypnotics and should generally be avoided for hypnotic use.

BARBITURATES

A number of barbiturates have or had an FDA-approved indication for use as sedative-hypnotics

(Table 7–2). Their use as oral hypnotics now is rarely justified. However, the barbiturates continue

to be used as intravenous agents of choice in anesthesia and occasionally in critical care settings.

The structures of these compounds are shown in Figure 6–3 (see Chapter 6: “Antianxiety Agents”).

Pentobarbital, secobarbital, amobarbital, and a combination of secobarbital and amobarbital are

most commonly used for nighttime sedation. In addition, several of these have indications for both

daytime and nighttime sedation. (The use of amobarbital, butabarbital, and other barbiturates for

daytime sedation is discussed in Chapter 6: “Antianxiety Agents.”) Dosage ranges of these

compounds and routes of administration are listed in Table 7–2.

Table 7–2. Barbiturates for insomnia

Generic name Brand

name

Formulations and strengths Nighttime dose

(mg)a

butabarbital sodium Butisol

Sodium

Tablets: 16.2, 30, 32.4, 50 mg 50–100

Elixir: 30 mg/5 mL (480-mL bottle)

pentobarbital Nembutal Injection: 50 mg/mL (20-mL, 50-mL vials) 150–200Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

13 of 20

31/01/2009 06:52

Generic name Brand

name

Formulations and strengths Nighttime dose

(mg)a

Capsule: 100 mg

phenobarbital Various

generics

Tablets: 15, 30, 60, 100 mg

Tablets: 16.2, 32.4, 64.8, 97.2 mgb

Concentrate: 20 mg/5 mL (480-mL)

Injection: 65 mg/mL (1-mL prefilled syringec and

single-dose vial), 130 mg/mL (1-mL prefilled syringe)

secobarbital Seconal Capsule: 100 mg

100

secobarbital and

amobarbital

Tuinal Capsule: 50, 50 mg each 50–200

aAdult dosages. For child dosages, consult the latest edition of Physicians’ Desk Reference (PDR) or a similar

reference.

b16.2 mg equivalent to 1/4 grain, 32.4 to 1/2 grain, and so forth.

cTubex, Carpuject.

For many years barbiturates were widely used for their hypnotic effects, but their use has dwindled

with the introduction of the safer benzodiazepines. The half-lives of barbiturates determine their

clinical applications. The extremely short-acting barbiturates (e.g., hexobarbital) are used for

preanesthesia or anesthesia. Intermediate-duration barbiturates (e.g., pentobarbital) are used for

induction and maintenance of sleep. Longer-acting compounds may be used for agitation or anxiety

(see Chapter 6: “Antianxiety Agents”). Barbiturate preparations are variations of a barbituric acid

structure with substitutions at one or more key positions (see Figure 6–3), resulting in differences

in lipophilia and half-lives.

The barbiturates have become less widely used because of their limited safety margin in overdose,

their potential for dependence, and the degree of central nervous system (CNS) depression they

induce. These drugs are extremely potent hypnotics, particularly for patients who have not

previously taken barbiturates. Indeed, some patients may demonstrate considerable somnolence

24 hours after ingestion. Moreover, these agents pose potential problems when mixed with alcohol

or other CNS depressants and are contraindicated in patients with acute intermittent porphyria or

attention-deficit/hyperactivity disorder.

In depressed patients, the administration of barbiturates often resulted in a marked reduction in

TCA plasma levels and diminution of antidepressant effects because of their induction of liver

microsomal enzymes and accelerated degradation of the TCA. Clinicians should keep this in mind

when considering adjunctive use of hypnotics in depressed patients. Indeed, we saw depressed

patients who were chronically treated with barbiturates for sleep and failed to respond to

amitriptyline and other TCAs but who did respond after discontinuing their barbiturates (see

Chapter 9: “Augmentation Strategies for Treatment-Resistant Disorders”).

Currently, the main remaining use in outpatient psychiatry for intermediate-acting

sedative-hypnotic barbiturates, such as pentobarbital, is probably limited to the older patient who

occasionally needs a hypnotic for insomnia and, because of prolonged past experience with

barbiturates, finds a barbiturate sleeping pill preferable to any of the newer benzodiazepines. Such

prescribing is, of course, a matter of giving in to a patient’s experience and preference rather than

a mandatory or necessary clinical maneuver. (For discussion of various uses of amobarbital,

including its use in treating catatonia and in analytically oriented interviews, see Chapter 6:

“Antianxiety Agents.”)

For patients who have become dependent on barbiturates, detoxification is necessary. Abrupt

withdrawal should be avoided because it can result in seizures, delirium, and even death. When thePrint: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

14 of 20

31/01/2009 06:52

daily dose of barbiturate is not known, physicians can give test doses of a barbiturate (watching for

the emergence of nystagmus, slurred speech, ataxia, and sedation) to determine the currently

needed dosage. The test dose can then act as a barometer of the dose at which to begin a

detoxification program (see Chapter 11: “Pharmacotherapy for Substance Use Disorders”).

NONBARBITURATE HYPNOTICS

In the late 1940s and the 1950s, several nonbarbiturate hypnotic drugs were developed in the hope

that they would be safer and better than the barbiturates. Unfortunately, this hope was not

realized. They proved, generally, to have as many limitations as the barbiturates.

Only one of these drugs currently available for use—chloral hydrate (Figure 7–3)—is generally

viewed as effective. The rest are no longer available. Chloral hydrate is a somewhat effective and

reasonably safe sleeping medication in doses between 500 and 1,500 mg at bedtime (Table 7–3).

Early prescription levels of 500 mg at bedtime often proved inadequate to produce sedation, and

most prescribers have come to favor 1,000 mg, particularly in younger adult patients. The drug is

often used in double-blind trials of other psychiatric medications as an adjunct medication because

of its presumed safety and “cleanliness.” It is unlikely that this reputation is fully deserved,

particularly because chloral hydrate itself was once widely abused in England in the early 1900s. It

also has a very low margin of safety, the lethal dose being about 10 times the hypnotic dose.

Chloral hydrate is irritating to the stomach. It was also used with alcohol as a “Mickey Finn” to

knock out victims who were being shanghaied. Alcohol and trichlorethanol, the metabolite and

active principal of chloral hydrate, are presumed to interact to increase the combination potency.

Trichlorethanol slows the metabolism of ethanol, while ethanol speeds the conversion of chloral

hydrate to trichlorethanol (Rall 1993).

Figure 7–3.

Chemical structures of nonbarbiturate hypnotics.

Table 7–3. Other nighttime hypnotic agents

Generic name Brand name Formulations and strengths Dosage (mg/day)a

zolpidem

Ambienb

Tablets: 5, 10 mg 5–10

zaleplon

Sonatab

Capsules: 5, 10 mg 5–10

eszopiclone

Lunestab

Tablets: 1, 2, 3 mg 1–3

ramelteon Rozerem Tablet: 8 mg 8

chloral hydrate Noctec Capsule: 500 mg 500–1,500

Syrup: 500 mg/5 mL (480-mL)

aAdult dosages. Patients may require slightly higher dosages of chloral hydrate or ethchlorvynol. For child

dosages, consult the latest edition of Goodman & Gilman’s.

b Lower doses may be indicated when combined with potent cytochrome P450 3A4 inhibitors (e.g., fluoxetine).

Another hypnotic, paraldehyde, is a colorless liquid with a very pungent odor and a burning,

disagreeable taste. The oral hypnotic dose is 5 or 10 mL. However, the medication has no

advantages and may be quite toxic if the material is old and deteriorated. It is too irritating to be

given parenterally, although intramuscular and rectal administration were once common.

Paraldehyde’s former use in the detoxification of patients with chronic alcoholism has been

supplanted by use of the longer-acting and safer benzodiazepines.

Other nonbarbiturate hypnotics, including glutethimide, ethinamate, methyprylon, and

ethchlorvynol, were developed because of the known dangers of addiction and lethality on suicidal

ingestion of barbiturates. Unfortunately, none of these drugs proved to be either safer or less

capable of producing physical or psychological dependence than the drugs that they were supposed

to replace. There are currently no logical reasons for using them in preference to the safer

benzodiazepines.Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

15 of 20

31/01/2009 06:52

METHAQUALONE

Methaqualone (Quaalude), a drug that still deserves passing mention (Figure 7–3), was a

nonbarbiturate hypnotic marketed as a safe and effective replacement for the more dangerous

older sleeping pills. It became widely abused in the illicit street market and achieved a good deal of

publicity as the “love” drug in the 1960s. Methaqualone was alleged to have aphrodisiac properties,

a proposition that has never been specifically tested in any scientific way. The drug, however, can

cause paresthesias, which may somehow increase sexual feelings.

Methaqualone has acquired a remarkably tainted reputation and is now no longer available for

prescription use in the United States. Whether the drug was really more likely to be abused or was

more euphoriant than the barbiturates is quite unclear. The one available study relevant to the

point suggested that it was really no different from pentobarbital in its subjective effects. In our

local clinical experience, we found occasional patients with severe insomnia who claimed that

methaqualone was far superior as a hypnotic—in terms of both speed of onset and lack of

hangover—than other available hypnotics, even comparable barbiturates. In any event,

methaqualone, for good or for bad, is no longer available.

OVER-THE-COUNTER SLEEPING AGENTS

L-Tryptophan

A hypnotic drug that seems to be free of dependence or abuse liability is the amino acid

L-tryptophan. It has not been formally approved for medical use as a hypnotic drug in the United

States, but it was available in both drugstores and health food stores as a dietary supplement,

either in the pure form or combined with other nutriments. There was reasonable evidence that, at

doses of 1–6 g at bedtime, L-tryptophan had some efficacy as a sleeping pill in some patients with

insomnia. It was neither as powerful as the major hypnotics nor, presumably, as dangerous. It

would be, however, a rather expensive way to get a night’s sleep, because the drug used to cost (in

the 1980s) almost $1 per gram.

In 1989, a small series of seriously ill patients with fever, myalgia, and massive eosinophilia were

identified. All had had exposure to L-tryptophan, and most had persistent illness requiring steroids

for relief. There were at least 37 deaths and 1,500 cases of eosinophilia-myalgia syndrome

reported. The cause was thought to be a foreign manufacturer’s producing a toxic variant of the

compound. On this basis, the FDA withdrew all L-tryptophan from the market. It is currently

available only as an investigational new drug or by order from Canadian pharmacies.

The risk of this serious disease is unknown, but L-tryptophan was used for years without evidence

of massive eosinophilia. If it were to be considered safe to use again, the starting dose should be

1–2 g at bedtime, with an increase to 3–4 g after two to three nights. The amount of available

L-tryptophan in milk—particularly when warmed—rivals a 2-g dose, and some clinicians have

advocated this time-honored approach for insomnia. Some workers have further noted that

carbohydrates facilitate the absorption of L-tryptophan, suggesting that milk and cookies may be

better than milk alone. We know of no double-blind studies of this approach, however.

The ill-considered use of L-tryptophan to treat insomnia in patients who are being treated with

fluoxetine or MAOIs should be avoided. The serotonin syndrome, manifested by myoclonic jerks,

hyperpyrexia, confusion, and coma, can be precipitated if doses of 1–4 g are given initially to

patients taking fluoxetine or MAOIs. The manufacturer of fluoxetine notes that the addition of

L-tryptophan to the regimen of patients already being treated with fluoxetine is contraindicated. It

is unclear whether beginning with 500 mg of L-tryptophan at bedtime and titrating upward is

clinically useful or safe in patients taking MAOIs. We have tried L-tryptophan for MAOI-related

insomnia in doses of 1–2 g at bedtime with little success but without adverse effects.

In recent years, a substitute for tryptophan, 5-hydroxytryptophan (or 5-HTP), has been marketed

in the United States for the treatment of depression and anxiety. Over the years, there have been a

number of small open and controlled trials supporting the use of 5-HTP. Most of these trialsPrint: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

16 of 20

31/01/2009 06:52

involved fewer than 20 patients, and definitive studies are lacking.

Melatonin

Melatonin, a naturally occurring hormone, is often used over the counter to treat insomnia and jet

lag. The prescription drug ramelteon is a melatonin agonist that became available in 2006 as a

hypnotic with no abuse potential. In humans, at least, melatonin appears to be excreted by the

pineal gland early during the sleep cycle and may well contribute to our “natural” circadian

rhythms. Blind subjects do not have a regular 24-hour rhythm as sighted people do. The melatonin

excretion patterns of blind subjects are also irregular. Elderly subjects with insomnia have low

levels of melatonin.

Melatonin has been shown to act as a hypnotic in small, brief studies of insomnia due to factors

such as jet lag or bad diurnal rhythm, but the correct dose and the effects of chronic (i.e., over

7 days) administration are unknown. Doses between 0.2 mg and 5 mg seem to be effective; higher

doses may be less effective (Zhdanova et al. 1997).

The substance is available over the counter in drugstores in at least (locally) 0.2-mg and 3-mg dose

units. It is sold as a dietary supplement. It is said that oral ingestion of melatonin has no effect on

the pineal gland’s daily production of the material—an odd state of affairs for a hormone. The

Medical Letter on Drugs and Therapeutics (“Melatonin” 1995) says that the commercial melatonin

can be synthetic or can be extracted from hog pineal glands. Some preparations contain

unidentified impurities. Experts on melatonin claim that it is a remarkably efficient antioxidant,

more potent than any other currently available. There is considerable variation in melatonin levels

produced by specific preparations.

Informal reports from friends and patients who have tried over-the-counter melatonin for mild

insomnia or jet lag have been favorable; it does feel subjectively as if it were a mild sedative. It

also appears to help overcome insomnia in elderly persons. However, so far, no single reliable

producer’s melatonin has been adequately tested for either efficacy or safety. This lack of adequate

study points to the discouraging object lesson of the disastrous effects of badly made L-tryptophan

on a few health food store patrons. Melatonin can be tried as a hypnotic, but unanticipated adverse

effects might occur (“Melatonin” 1995).

Valerian Extract

Valerian has come to be popularly assumed to be a useful herbal preparation for improving sleep or

decreasing anxiety. It binds to GABA receptors, perhaps because the extract actually contains

GABA. Valerian’s innate GABA cannot cross the blood-brain barrier and should not, therefore, help

with insomnia or anxiety. Valerian extract is also believed to bind to serotonin and adenosine

receptors (Houghton 1999; Wong et al. 1998). The adenosine receptor may be responsible for

meprobamate’s sedative-hypnotic actions.

Several small studies suggest that valerian extract may affect sleep positively, but the studies were

small and the findings were inconsistent (Buysse and Dorsey 2002). One can only repeat the

warning from L-tryptophan: “dietary supplements,” under which category valerian falls, are not

evaluated for safety and efficacy, and the user runs the risk of unpredictable adverse effects. On

the other hand, if one were to see a patient who had used valerian extract for months or years for

periodic insomnia with satisfaction, such a patient may be better off continuing to use valerian than

shifting to a benzodiazepine-like hypnotic.

If valerian proves to be effective for insomnia, it will be a very delayed lesson for Dr. Harry Gold, an

early clinical pharmacologist who pioneered double-blind studies in angina and pain at Cornell

Medical College in the 1940s. He used to assert that valerian extract was the ideal placebo—it had

an interesting color and taste, it had multiple uses (the pharmacopoeia of the time listed about 50

uses for it), it could be given either in drops or in cupfuls, and it was perfectly benign and without

pharmacological effects. Maybe he was wrong.

GENERAL CONCLUSIONSPrint: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

17 of 20

31/01/2009 06:52

For the physician seeing a patient with recent-onset insomnia without evident primary medical,

psychiatric, or sleep diagnoses (e.g., restless legs, sleep apnea) that would account for it,

prescribing a small amount of a shorter-acting hypnotic (zolpidem, temazepam, lorazepam) in the

lowest dose size taken every 3 days for up to 4 weeks seems reasonable, assuming the patient has

no history of substance abuse.

For short-term relief of insomnia, zolpidem 5–10 mg hs has been generally satisfactory, although it

would be hard to prove that it is the best of the available benzodiazepine-like drugs. Zaleplon

probably produces a briefer hypnotic effect than zolpidem, and it has not been as well adopted as

has zolpidem. Eszopiclone may turn out to be a major addition—effective acutely as well as over

extended periods. Patients who awaken too early after taking zolpidem may do better taking

temazepam as a longer-acting sedative. However, any of the benzodiazepines may be effective

when the right dose for the patient is determined, though flurazepam, quazepam, and possibly

clonazepam may have long enough half-lives to cause morning sedation and, perhaps, psychomotor

impairment.

For patients with long-standing insomnia who are already accustomed to a nightly hypnotic of the

benzodiazepine class, physical dependence probably exists and a slow but purposeful taper of

dosage over months, not days or weeks, may be needed. If shifting to a nonbenzodiazepine

hypnotic (e.g., trazodone), it would seem sensible to first add the new hypnotic and make sure it

improves sleep for at least 2 weeks before beginning to taper the benzodiazepine.

Trazodone 50 mg hs, mirtazapine 7.5–15 mg hs, trimipramine 25–50 mg hs, or gabapentin 300–500

mg hs may do as well as a benzodiazepine and can be taken nightly with little concern about

physical dependence; but because these are “off-label” (non–FDA-approved) uses, written

justification for their use should be made in the medical record. Over-the-counter Benadryl

(diphenhydramine), at two 25-mg capsules, may be almost as effective. The rare occurrence of

priapism with trazodone deserves mention to male patients, with a warning to seek emergency

room help if an erection persists beyond 4 hours (Thompson et al. 1999).

All the above except the benzodiazepine-like drugs could be used in patients with a history of

substance abuse problems, although such patients can sometimes escalate the dose of any drug

rapidly and unwisely. In very unstable, impulsive individuals, gabapentin or trazodone are safest.

On the other hand, if an alcoholic individual in prolonged remission comes to a physician’s office

with either a credible history of intractable, severe, untreated insomnia or a history of safe,

cautious, appropriate use of a hypnotic like clonazepam for months or years, continuing or starting

a benzodiazepine-like hypnotic with careful monitoring is probably safe.

For patients with long-term insomnia being treated with hypnotics and without diagnosable sleep

problems that require treatment with nonhypnotics, having the patient continue to take the

hypnotic while evaluating and treating any concurrent Axis I psychiatric disorders is sensible. Once

the patient’s condition and the patient-physician relationship are stable, alternatives can be

considered. These include shifting to nonbenzodiazepine sleep-inducing drugs, tapering the original

hypnotic very slowly, referring the patient to a sleep program for more specialized diagnosis and

possible behavioral or other nonsedative treatments, or making and documenting a conscious,

responsible treatment plan in which the patient continues to take a stable benzodiazepine dose for

a defined period, say a year, for sensible reasons (e.g., waiting for a divorce to become final,

finishing a major job, recovering fully from a manic episode).

At the other extreme, one sees patients with a desperate, frantic need for a full night’s sleep who

have been taking large doses of several drugs (e.g., temazepam 90 mg plus trazodone 200 mg plus

gabapentin 600 mg per night) with little benefit. In our experience such patients need to be

referred to a sleep program. If the program fails or a program is unavailable, they need to be told

that hypnotic medications are not the answer, that even 90 mg of temazepam either will not work

or will harm them, and that they need to have their bedtime medications gradually but firmly

tapered while they engage in psychotherapy and a search is made for other, more treatable

conditions.Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

18 of 20

31/01/2009 06:52

BIBLIOGRAPHY

Backonja M, Beydoun A, Edwards KR, et al: Gabapentin for the symptomatic treatment of painful neuropathy

in patients with diabetes mellitus: a randomized controlled trial. JAMA 280:1831–1836, 1998 [PubMed]

Bunney WE, Azarnoff DL, Brown BW, et al: Report of the Institute of Medicine Committee on the Efficacy and

Safety of Halcion. Arch Gen Psychiatry 56: 349–352, 1999 [PubMed]

Buysse DJ, Dorsey CM: Current and experimental therapeutics of insomnia, in American College of

Neuro-Psychopharmacology: A Fifth Generation of Progress. Edited by Davis JM, Charney D, Coyle J, et al.

Philadelphia, PA, Lippincott Williams & Wilkins, 2002, pp 1931–1943

Chouinard G, Beauclair L, Belanger M: Gabapentin: long-term antianxiety and hypnotic effects in psychiatric

patients with comorbid anxiety-related disorders (letter). Can J Psychiatry 43:305, 1998 [PubMed]

Cora-Locatelli G, Greenberg BD, Martin JD, et al: Rebound psychiatric and physical symptoms after gabapentin

discontinuation (letter). J Clin Psychiatry 59:131, 1998 [PubMed]

Dement W, Seidel W, Carskadon M, et al: Changes in daytime sleepiness/alertness with nighttime

benzodiazepines, in Pharmacology of Benzodiazepines. Edited by Usdin E, Skolnick P, Tallman JF, et al. New

York, Macmillan, 1982, pp 219–228

Earley CJ, Allen RP: Pergolide and carbidopa/levodopa treatment of restless legs syndrome and periodic leg

movements in sleep in a consecutive series of patients. Sleep 19:801–810, 1996 [PubMed]

Edinger JD, Means MK: Cognitive-behavioral therapy for primary insomnia. Clin Psychol Rev 25(5):539–558,

2005 [PubMed]

Ehrenberg BL, Eisensher I, Corbett KE, et al: Valproate for sleep consolidation in periodic limb movement

disorder. J Clin Psychopharmacol 20:574–578, 2000 [PubMed]

Elie R, Ruther E, Farr I, et al: Sleep latency is shortened during 4 weeks of treatment with zaleplon, a novel

nonbenzodiazepine hypnotic. J Clin Psychiatry 60:536–544, 1999 [PubMed]

Erman M, Seiden D, Zammit G, et al: An efficacy, safety, and dose-response study of ramelteon in patients

with chronic primary insomnia. Sleep Med 7(1):17–24, 2006 [PubMed]

Falco M: Methaqualone misuse: foreign experience and US drug control policy. Int J Addict 11:597–610, 1976

[PubMed]

Garfinkel D, Laudon M, Nof D, et al: Improvement in sleep quality in elderly people by controlled-release

melatonin. Lancet 346:541–544, 1995 [PubMed]

Gelenberg A (ed): Zaleplon: a new nonbenzodiazepine hypnotic. Biological Therapies in Psychiatry 23:5–6,

2000

Gillin JC, Byerley WF: Drug therapy: the diagnosis and management of insomnia. N Engl J Med 322:239–248,

1990 [PubMed]

Gillin JC, Reynolds CF, Shipley JE: Sleep studies in selected adult neuropsychiatric disorders, in Psychiatry

Update: American Psychiatric Association Annual Review, Vol 4. Edited by Hales RE, Frances AJ. Washington,

DC, American Psychiatric Press, 1985, pp 352–360

Greenblatt DJ, Shader RI, Abernethy DR: Current status of benzodiazepines, I. N Engl J Med 309:354–359,

1983a

Greenblatt DJ, Shader RI, Abernethy DR: Current status of benzodiazepines, II. N Engl J Med 309:410–415,

1983b

Gursky JT, Krahn LE: The effects of antidepressants on sleep: a review. Harv Rev Psychiatry 8:298–306, 2000

[PubMed]

Hajak G, Muller WE, Wittchen H-U, et al: Abuse and dependence potential for the nonbenzodiazepine hypnotics

zolpidem and zopiclone: a review of case reports and epidemiological data. Addiction 98:1371–1378, 2003Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

19 of 20

31/01/2009 06:52

[PubMed]

Hauri PJ, Sateia MJ: Nonpharmacological treatment of sleep disorders, in Psychiatry Update: American

Psychiatric Association Annual Review, Vol 4. Edited by Hales RE, Frances AJ. Washington, DC, American

Psychiatric Press, 1985, pp 361–378

Hertzberg MA, Feldman ME, Beckham JC, et al: Trial of trazodone for posttraumatic stress disorder using a

multiple baseline group design. J Clin Psychopharmacol 16:294–298, 1996 [PubMed]

Hertzman PA, Blevins WL, Mayer J, et al: Association of the eosinophilia-myalgia syndrome with the ingestion

of tryptophan. N Engl J Med 322:869–873, 1990 [PubMed]

Houghton PJ: The scientific basis for the reputed activity of valerian. J Pharm Pharmacol 51:505–512, 1999

[PubMed]

Hughes RJ, Sack RL, Lewy AJ: The role of melatonin and circadian phase in age-related sleep-maintenance

insomnia: assessment in a clinical trial of melatonin replacement. Sleep 21:52–68, 1998 [PubMed]

Johnson MW, Suess PE, Griffiths RR: Ramelteon: a novel hypnotic lacking abuse liability and sedative adverse

effects. Arch Gen Psychiatry 63(10):1149–1157, 2006 [PubMed]

Kales A: Benzodiazepines in the treatment of insomnia, in Pharmacology of Benzodiazepines. Edited by Usdin

E, Skolnick P, Tallman JF, et al. New York, Macmillan, 1982, pp 199–217

Kales A: Quazepam: hypnotic efficacy and side effects. Pharmacotherapy 10:1–12, 1990 [PubMed]

Karam-Hage M, Brower KJ: Open pilot study of gabapentin versus trazodone to treat insomnia in alcoholic

outpatients. Psychiatry Clin Neurosci 57(5):542–544, 2003 [PubMed]

Katz MM, Koslow SH, Maas JW, et al: The timing, specificity and clinical prediction of tricyclic drug effects in

depression. Psychol Med 17:297–309, 1987 [PubMed]

Katz MM, Koslow SH, Maas JW, et al: Identifying the specific clinical actions of amitriptyline: interrelationships

of behavior, affect and plasma level in depression. Psychol Med 21:599–611, 1991 [PubMed]

Ketter TA, Post RM, Theodore WH: Positive and negative psychiatric effects of antiepileptic drugs in patients

with seizure disorders. Neurology 53 (5, suppl 2):S53–S67, 1999

Lasagna L: Over-the-counter hypnotics and chronic insomnia in the elderly. J Clin Psychopharmacol

15:383–386, 1996

McGechan A, Wellington K: Ramelteon. CNS Drugs 19(12):1057–1065 [discussion: 1066–1067], 2005

Melatonin. Med Lett Drugs Ther 37:111–112, 1995

Mendelson WB (ed): Current strategies in the treatment of insomnia. J Clin Psychiatry 53 (12, suppl):1–45,

1992

National Center on Sleep Disorders Research: Insomnia: Assessment and Management in Primary Care (NIH

Publ No 98-4088). Washington, DC, National Center on Sleep Disorders Research, 1998

Rall TW: Hypnotics and sedatives, ethanol, in Goodman and Gilman’s The Pharmacologic Basis of Therapeutics,

8th Edition. Edited by Gilman AG, Rall TW, Nies AS, et al. New York, McGraw-Hill, 1993, pp 345–382

Regenstein Q: Specific effects of sedative/hypnotic drugs in the treatment of incapacitating insomnia. Am J

Med 83:909–916, 1987

Reiter RJ, Melchiorri D, Sewerynek E, et al: A review of the evidence supporting melatonin’s role as an

antioxidant. J Pineal Res 18:1–11, 1995 [PubMed]

Reynolds CF, Buysse DJ, Kupfer DJ: Disordered sleep: developmental and biopsychosocial perspectives on the

diagnosis and treatment of persistent insomnia, in Psychopharmacology: The Fourth Generation of Progress.

Edited by Bloom FE, Kupfer DJ. New York, Raven, 1995, pp 1617–1629

Rickels K, Morris RJ, Newman H, et al: Diphenhydramine in insomnia. J Clin Psychiatry 23:235–242, 1983Print: Chapter 7. Hypnotics

http://www.psychiatryonline.com/popup.aspx?aID=237641&print=yes…

20 of 20

31/01/2009 06:52

Rickels K, Ginsberg J, Morris RJ, et al: Doxylamine succinate in insomniac family practice patients: a

double-blind study. Current Therapeutic Research 35:532–540, 1984

Roffwarg H, Erman M: Evaluation and diagnosis of the sleep disorders: implications for psychiatry and other

clinical specialties, in Psychiatry Update: American Psychiatric Association Annual Review, Vol 4. Edited by

Hales RE, Frances AJ. Washington, DC, American Psychiatric Press, 1985, pp 294–328

Roth T, Seiden D, Sainati S, et al: Effects of ramelteon on patient-reported sleep latency in older adults with

chronic insomnia. Sleep Med 7(4):312–318, 2006 [PubMed]

Rowbotham M, Harden N, Stacey B, et al: Gabapentin for the treatment of postherpetic neuralgia: A

randomized controlled trial. JAMA 280:1837–1842, 1998 [PubMed]

Rush C, Griffiths R: Zolpidem, triazolam, and temazepam: behavioral and subject-rated effects in normal

volunteers. J Clin Psychopharmacol 16: 146–157, 1996 [PubMed]

Rush CR, Baker RW, Wright K: Acute behavioral effects and abuse potential of trazodone, zolpidem and

triazolam in humans. Psychopharmacology (Berl) 144:220–233, 1999 [PubMed]

Scharf MB, Mayleben DW, Kaffeman M, et al: Dose response effects of zolpidem in normal geriatric subjects. J

Clin Psychiatry 52:77–83, 1991 [PubMed]

Schneider-Helmert D: Why low-dose benzodiazepine dependent insomniacs can’t escape their sleeping pills.

Acta Psychiatr Scand 78:706–711, 1988 [PubMed]

Seidel WF, Roth T, Roehrs T, et al: Treatment of a 12-hour shift of sleep schedule with benzodiazepines.

Science 224:1262–1264, 1985

Shamir E, Landon M, Barak Y, et al: Melatonin improves sleep quality in patients with chronic schizophrenia. J

Clin Psychiatry 61:373–377, 2000 [PubMed]

Thompson JW, Ware M, Blashfield R: Psychotropic medications and priapism: a comprehensive review. J Clin

Psychiatry 51:430–433, 1999

Wang MY, Wang SY, Tsai PS: Cognitive behavioural therapy for primary insomnia: a systematic review. J Adv

Nurs 50(5):553–564, 2005 [PubMed]

Wong AH, Smith M, Boon HS: Herbal remedies in psychiatric practice. Arch Gen Psychiatry 55:1033–1044,

1998 [PubMed]

Zhdanova IV, Lynch JH, Wurtman RJ: Melatonin: a sleep-promoting hormone. Sleep 20:899–907, 1997

[PubMed]

Copyright © 2009 American Psychiatric Publishing, Inc. All Rights Reserved.