Numerous classes of medications have been used as sedative-hypnotic drugs. In current practice, benzodiazepine receptor agonists and antidepressants are the two classes with the most widespread applications (Walsh and Schweitzer 1999). Stimulants, dopaminergic drugs, and anticonvulsants also play a role in the treatment of some sleep disorders. Details regarding the pharmacology of these agents can be found in other chapters in this volume. The relevant properties of these drugs for the treatment of sleep disorders are discussed briefly here (

Table 89-2).

Benzodiazepine receptor agonists Benzodiazepines, imidazopyridines, and cyclopyrrolones all facilitate the activity of the inhibitory neurotransmitter γ-aminobutyric acid (GABA). They bind at specific benzodiazepine receptors, which form part of a macromolecular complex with GABA receptors and a chloride ion channel (Greenblatt 1992). In general, benzodiazepines bind at each of three subtypes of Ω receptors; some benzodiazepines (e.g., quazepam) bind more specifically at the benzodiazepine receptor subtype in low doses, and the imidazopyridine zolpidem and pyrazolopyrimidine zaleplon also bind selectively at these receptors. All benzodiazepines produce hypnotic, anxiolytic, myorelaxant, and anticonvulsant effects. The relative specificity for these actions, as well as affinity for benzodiazepine receptors, varies considerably among specific agents. Zolpidem and zaleplon have very little anxiolytic, anticonvulsant, or myorelaxant effect at clinical doses.

Benzodiazepine receptor agonists and related compounds reduce sleep latency and intermittent wakefulness during sleep. Their effects on specific sleep stages are more variable. Most benzodiazepines suppress Stage 3/4 non-rapid eye movement (NREM) sleep, increase Stage 2 sleep, increase sleep spindles, and mildly suppress REM sleep. Zolpidem and zaleplon have fewer effects on specific sleep stages. Benzodiazepine receptor agonists are also weak respiratory suppressants and at high doses can worsen sleep apnea (Guilleminault et al. 1984). Clinically, this effect is small in patients with mild to moderate apnea or chronic obstructive pulmonary disease who take short-acting drugs (Mendelson 1991; Steens et al. 1993).

The adverse effects of benzodiazepines and related compounds include daytime sedation, ataxia, and impairment of anterograde memory and psychomotor performance. In addition, benzodiazepines are a risk factor for serious injurious falls and hip fractures in the elderly (Herings et al. 1995). Such studies have not consistently evaluated the risk of insomnia itself, which also may be related to falls (Koski et al. 1998). In addition, benzodiazepines have the potential for producing tolerance, dependence, and withdrawal phenomena. Some authors assert that these problems are more prevalent and more significant than previously thought (N. S. Miller and Gold 1991). Others point out, however, that even well-controlled studies are intrinsically biased to show dependence effects and that such studies overestimate the actual occurrence of dependence in representative users of benzodiazepines (Shader and Greenblatt 1993).

Antidepressants Antidepressants include tricyclic antidepressants (TCAs) and related compounds such as trazodone. Selective serotonin reuptake inhibitors (SSRIs) have similar but much more specific pharmacological actions. Other antidepressants that inhibit neurotransmitter reuptake or have other synaptic effects include venlafaxine, bupropion, and mirtazapine. Monoamine oxidase inhibitors (MAOIs) have little utility for the treatment of most sleep disorders. These different classes of antidepressants differ considerably in pharmacological actions, effects on sleep, and side-effect profiles. Only specific sleep effects are outlined here.

Specific tricyclic drugs have a range of sedative potential, which is related in part to cholinergic and α-adrenergic receptor blockade and serotonergic reuptake blockade. As a result of these properties, these drugs vary from quite sedating (e.g., amitriptyline, doxepin, trimipramine) to activating (protriptyline, desipramine). Effects on objective measures of sleep continuity generally parallel subjective sedative effects. TCAs have little overall effect on Stage 3/4 NREM sleep, although some drugs, such as amitriptyline, can increase Stage 3/4 sleep. SSRIs are also generally experienced as somewhat activating, although individual patients may complain of notable sedation. For instance, fluoxetine produces dose-related activation in about 30% of patients and an apparently non-dose-related sedation in about 15% (Beasley et al. 1992). Objectively, SSRIs, particularly fluoxetine, impair sleep continuity in a dose-related fashion. They have little effect on Stage 3/4 NREM sleep. Venlafaxine’s effects are similar to those of the SSRIs, including REM suppression and reduced sleep continuity.

MAOIs, particularly tranylcypromine, are generally experienced as somewhat activating and may interrupt sleep continuity at night. They have little specific effect on Stage 3/4 NREM sleep. Some patients experience sedation approximately 6 hours after taking a dose of these drugs.

Almost all classes of antidepressant drugs, given in sufficient doses, suppress REM sleep. Individual drugs vary widely in the potency of this effect. For instance, some tricyclic agents, such as clomipramine, and MAOIs, including both tranylcypromine and phenelzine, suppress REM sleep very strongly. Other tricyclic and related drugs, including trimipramine and trazodone, as well as SSRIs, suppress REM sleep weakly and inconsistently, and only at higher doses. Bupropion, an atypical antidepressant agent with no clear effects on noradrenergic or serotonergic neurotransmission, and nefazodone, a serotonin type 2 (5-HT2) receptor antagonist, may actually increase REM sleep (Nofzinger et al. 1995; Ware et al. 1994), making them unique among antidepressant drugs.

Antidepressant drugs also increase neuromuscular tone and as a result can increase restless legs symptoms and periodic limb movements during sleep (Ware et al. 1984), although this effect is not consistent (Buysse et al. 1996). Amoxapine and SSRIs have been reported to cause akathisia, which may also disrupt sleep. On the other hand, TCAs and SSRIs tend to decrease by a small degree the number of apneas and hypopneas (Buysse et al. 1996; Hanzel et al. 1991; Kopelman et al. 1992), although this effect is of questionable clinical utility.

Antihistamines Antihistamine drugs vary in their degree of penetration into the central nervous system. Histamine1 receptor blocking agents that penetrate the central nervous system decrease vigilance and wakefulness and improve subjective sleep continuity in patients with insomnia (Rickels et al. 1983). Their effects on nocturnal sleep, however, are less consistent. Tolerance and “hangover” the following morning limit their utility. Furthermore, these drugs often have anticholinergic activity, which can lead to confusion, particularly in the elderly. A call-in survey of patients with insomnia found that over-the-counter hypnotic drugs (consisting mainly of antihistamine and anticholinergic agents) were less effective and produced more side effects than benzodiazepines (Balter and Uhlenhuth 1991). Histamine2 receptor-blocking agents, such as cimetidine, may cause nightmares but have little sedative potential.

Barbiturates Barbiturates bind at a distinct site near the benzodiazepine-GABA-chloride channel macromolecular complex. Like benzodiazepines, they facilitate GABA activity and membrane hyperpolarization. They have clear sedative effects in humans, which is verified by a decrease in wakefulness during sleep. Barbiturates suppress both Stage 3/4 NREM sleep and REM sleep and increase sleep spindling activity. The utility of barbiturates is limited by several factors, including the development of tolerance (particularly with short-acting agents, such as pentobarbital), a low therapeutic index with the potential for lethal overdose, and the equally dangerous possibility of life-threatening withdrawal seizures. For these reasons, barbiturates play essentially no role in the current treatment of insomnia disorders.

Nonbarbiturate, nonbenzodiazepine drugs Like barbiturates, the older nonbarbiturate, nonbenzodiazepine drugs are limited by their substantial toxicity. These drugs include agents such as glutethimide, methyprylon, and meprobamate. Tolerance and low therapeutic index, systemic side effects (including hepatotoxicity), and withdrawal seizures make these unsuitable agents for the current treatment of sleep disorders.

Alcohol Alcohol has both subjective and objective sedative properties. Doses of alcohol given before sleep diminish sleep latency and improve sleep continuity during the first part of the sleep period. Stage 3/4 NREM sleep increases, and REM sleep is suppressed. Alcohol has a short duration of action, however, and wakefulness and REM sleep tend to “rebound” during the second half of the night, leading to worse sleep overall. Sleep apnea is increased by alcohol (Scrima et al. 1982). Given its potential for abuse, tolerance, dependence, serious systemic side effects, and withdrawal syndrome, alcohol has no place in the clinical management of sleep disorders.

Stimulants Amphetamines and other stimulant drugs are indicated for the treatment of some disorders of excessive daytime sleepiness. Specific agents include traditional stimulants (dextroamphetamine, methylphenidate, methylamphetamine), pemoline, and modafinil. Traditional stimulant drugs act by releasing neuronal stores and blocking reuptake of dopamine and norepinephrine. The exact mechanism of action for pemoline and modafinil are uncertain, but the latter may act on receptors for the hypothalamic peptide orexin (Chemelli et al. 1999). Each of these drugs increases wakefulness and decreases sleepiness during the day by objective measures. Traditional stimulants also can impair nocturnal sleep continuity and suppress Stage 3/4 NREM sleep and REM sleep. These effects are dose related and depend on the timing of the dose relative to sleep.

Adverse effects of traditional stimulants include agitation, irritability, insomnia, appetite suppression and weight loss, hypertension, and cardiac arrhythmias. Hypersomnia often increases on withdrawal from regular use. Pemoline generally produces fewer side effects, but it can lead to hyponatremia, and several cases of fulminant hepatic failure have been reported. Regular monitoring of liver function tests is recommended. Modafinil is generally well tolerated, with headaches and gastrointestinal upset being the most common side effects. Withdrawal hypersomnia does not appear to occur after withdrawal of modafinil.

Caffeine Caffeine and other methylxanthines, such as theophylline, also have stimulant activity. These drugs produce dose-related increases in nocturnal wakefulness, with no clear or specific effects on particular sleep stages. Side effects of nervousness, agitation, and cardiac arrhythmias, as well as the relatively weak stimulant effect of these drugs, limit their clinical utility.

A wide variety of other medications and drugs of abuse have effects on sleep. Common examples include antihypertensive agents, which act on the sympathetic nervous system; anticonvulsant medications; and decongestants. Drugs of abuse such as cocaine and heroin also can affect sleep. However, these agents are only rarely prescribed for the specific treatment of sleep disorders.