ABC of Sleep Medicine (ABC Series)

Chapter 11

Drugs Used in Sleep Medicine


·        Many classes of drug, whether prescribed or recreational, are taken with the specific aim of influencing the sleep-wake cycle

·        The formal evidence base for the pharmacological treatment of the majority of sleep disorders is very limited and many drugs are used empirically

·        Caffeine, modafinil and amphetamine-like agents are the most commonly used agents for promoting wakefulness, often in combination

·        Clear protocols for the drug treatment of chronic insomnia are lacking and long-term courses of hypnotic agents are generally discouraged through fears of dependence and tolerance

·        If an agent is used to aid sleep, its pharmacokinetic profile should be appropriately matched for the type of insomnia

·        Many drugs such as benzodiazepines and the majority of anti-depressants will produce drowsiness but often fail to improve the quality of any sleep obtained

·        It is relatively rare for hypnotic or sedative drugs to increase or enhance the deeper stages of non-REM (slow wave) sleep

Whether obtained by a prescription, over the counter or illegally, one of the commonest reasons for a person to take a drug is to influence the sleep–wake cycle. Typically, the elderly seek agents to improve their nocturnal sleep whereas younger age groups want to stay more awake.

Unfortunately, the evidence base for pharmacological treatment of the vast majority of sleep disorders remains sparse. As a result, drugs are often chosen either empirically or from anecdotal evidence and personal experience. Furthermore, given that most trial data are generated from short-term studies, little guidance exists on the appropriate length of treatment courses.

Only a minority of drugs are formally licensed for a sleep-related indication. In addition, because sleep medicine remains a relatively new discipline, familiarity with many of these agents is low. A further reason for reduced confidence in prescribing drugs for sleep disorders is that many general physicians might feel that a certain sleep-related symptom reflects lifestyle choices more than a purely medical problem.

Agents to promote wakefulness

The vast majority of subjects with a significant primary sleep disorder such as narcolepsy will benefit from drug therapy to improve alertness. In some other clinical situations where daytime somnolence remains severe despite first-line therapy, it may also be appropriate to recommend wake-promoting agents. The use of traditional psychostimulant drugs such as amphetamine has declined largely through concerns over dependency and the relatively non-specific mode of action.

The earliest treatments for narcolepsy when it was first described by French neuropsychiatrists in the late nineteenth century almost certainly would have kept subjects awake (Figure 11.1).

Figure 11.1 A ‘recipe’ outlining a treatment plan for narcolepsy described by Dr Gelineau, a Parisian neuropsychiatrist who first devised the term ‘narcolepsy’.



Caffeine is clearly widely available in a vast variety of drinks and also tablet form. However, it can often be difficult to judge precise doses consumed; these typically vary between 50 and 200 mg. The main relevant pharmacological action of caffeine is to inhibit adenosine (A1) receptors in the basal forebrain. This is thought to oppose the natural sleep drive, which arises partly from increasing adenosine levels in the brain after prolonged wakefulness. There is a wide variation between individuals both in the alerting response to caffeine and the likelihood of adverse effects such as insomnia.

Although a relatively mild wake-promoting agent for most, caffeine is probably under-used as a supplement in sleep disorders that cause significant somnolence. At the very least, it can usefully improve vigilance in sleep-deprived subjects.


Modafinil was first introduced as a unique wake-promoting drug in the 1980s after the serendipitous discovery that it kept cats awake for prolonged periods in experimental drug studies assessing depression. The new drug was noted to have a different chemical structure to traditional psychostimulant agents such as amphetamine (Figure 11.2). Its precise mode of action remains unclear although it appears to activate the ‘wake’ nuclei in the hypothalamus fairly specifically. Unlike amphetamine, it does not produce euphoria. Furthermore, tolerance after prolonged use is usually minimal. The half-life is quite long (10–15 hours) such that typical doses (100 or 200 mg) are taken shortly after awakening and at lunchtime to avoid insomnia at night.

Figure 11.2 Chemical structures of (a) modafinil and (b) amphetamine.


Although modafinil is formally licensed in many countries only for the treatment of primary sleep disorders such as narcolepsy, many clinicians recommend its use in other situations. Excessive sleepiness associated with head injuries or other neurological disorders often benefit from the drug, as do patients with severe obstructive sleep apnoea who remain sleepy despite optimal therapy with a positive airways pressure mask. Furthermore, it is occasionally appropriate to prescribe the drug before a night shift at work if other countermeasures to improve drowsiness have failed.

Some claim that modafinil is effective as a cognitive enhancer or ‘smart drug’. Although it remains unclear whether it improves attention and performance in those who are not sleepy, it is sometimes misused in student populations, for example, in attempts to boost academic achievement.

The more common side effects of modafinil include headache and gastrointestinal upset. It does not have the same stimulatory effects on the cardiovascular system as more traditional stimulants but blood pressure can rise slightly and should be monitored. Woman of child-bearing age should be told that modafinil is known to induce liver enzymes and has the potential of rendering standard oral contraceptive therapy ineffective.

Amphetamine and related drugs

Amphetamines and drugs such as methylphenidate have a very similar mode of action and are generally only used as second-line agents in primary sleep disorders causing hypersomnolence. Amphetamine is most commonly used as a supplement to modafinil or as an alternative if there are side effects. It is given in low doses (typically 5 mg up to five times a day) and usually taken flexibly. A single dose typically gives up to two hours of increased alertness at levels roughly equivalent to the wake-promoting effects of drinking around six cups of coffee.

At low doses, amphetamine and similar drugs promote the release of dopamine and noradrenaline from nerve terminals and inhibit their reuptake. As a result, the pharmacological effects are not limited to increased wakefulness. Agitation or palpitations may occur as unwanted side effects.

Amphetamine will generally suppress appetite and worsen hypertension, which may also limit its usefulness. The euphoric effects seem less pronounced when used in conditions such as narcolepsy. Indeed, frank abuse in very rarely seen in this clinical population.


Selegiline was first introduced in the 1950s as a dog food additive to give pets a ‘boost’. It is metabolised in the liver to amphetamine, making it useful as a mild psychostimulant in addition to its more familiar role as an agent to improve the motor symptoms of Parkinson's disease. Although not licensed as a wake-promoting drug, it is sometimes recommended as an alternative to amphetamine in elderly patients or those not suitable for more powerful drugs. Doses higher than those used in Parkinson's disease (more than 20 mg daily) seem to be safe in narcoleptics, for example.


Mazindol was first developed as an appetite suppressant but found early promise as a psychostimulant in narcoleptic patients. It's mode of action remains obscure although the clinical effects resemble those of amphetamine. It is sometimes used as an unlicensed alternative agent for narcolepsy with cataplexy in those resistant or intolerant to conventional therapy.

Side effects include nausea and agitation. Concerns over a possible theoretical risk of pulmonary hypertension with long-term use have necessitated specialist use only.

Agents to induce or improve nocturnal sleep

General principles

Although hypnotics are amongst the most commonly used class of drugs, prescribing guidelines and protocols are poorly developed. In the last decade there has been increasing reluctance to recommend hypnotics, largely through fears of dependence, tolerance and morning side effects. In patients with severe and chronic symptoms, the poor availability of behavioural or other non-pharmacological therapies, leads to understandable frustration and anxiety, potentially fuelling worsening insomnia and an increased demand for a pharmacological approach.

Many patients with chronic insomnia have tried numerous strategies before seeking medical help. Antihistamines available over-the-counter or herbal remedies such as lavender are rarely of major benefit. Similarly, for obvious reasons, excess alcohol should be discouraged if used primarily as an aid to achieve sleep onset.

Wherever possible, a treatable cause for insomnia should be addressed. Restless legs syndrome causing significant sleep onset problems and obstructive sleep apnoea producing unrefreshing sleep are common examples of missed diagnoses.

If hypnotic drug therapy is considered appropriate, the nature of the insomnia should be determined. A drug's absorption and elimination properties should match the profile of insomnia. Short-acting drugs should be reserved for problems actually falling asleep and longer-acting agents if sleep interruption late in the night is the major concern (Table 11.1).

Table 11.1 Properties of commonly used hypnotic agents.


Although many hypnotics may extend total sleep time, they vary in their ability to produce good quality sleep. Awakening unrefreshed from sleep is particularly associated with drugs that inhibit the deep non-REM (slow wave) stage. Finding an effective hypnotic agent for an individual may take several attempts and many subjects remain disillusioned that they cannot find an ideal drug.


In the 1970s and 1980s, benzodiazepines (BZs) largely replaced barbiturates as the most commonly used class of hypnotic drug. There are numerous types of BZ with varying pharmacokinetic properties although all modulate the inhibitory neurotransmitter gamma-aminobutyric acid (GABA). Because they indirectly enhance the effects of the brain's endogenous GABA, the risks of respiratory depression and excessive sedation are limited, making them relatively safe in overdose. However, if taken with other drugs that more directly affect the GABA-A receptor, such as alcohol, problems may well arise.

Potentially as a result of their anxiolytic properties, BZs may improve sleep subjectively with little objective change in sleep parameters measured by polysomnography.

The general inhibitory effects of BZs may produce muscle relaxation, memory impairment and gait disturbances. These effects are undesirable for those that need to get up through the night for child care purposes or micturition, for example. Snoring and sleep apnoea are also enhanced, especially in the presence of alcohol.

BZs are probably best used as intermittent therapy. It is important to instruct patients that rebound insomnia may be expected on discontinuation and that doses should ideally be tapered when withdrawing.

Benzodiazepine receptor agonist drugs (‘z-drugs’)

Shorter-acting hypnotic drugs were developed primarily to avoid the ‘hangover’ effects of BZs. The three generally available ‘z-drugs’ are zopiclone, zolpidem and zaleplon. These drugs have been shown to be generally effective in sleep-onset insomnia with little evidence for physical dependence, even in trials lasting 12 months. Their mode of action is also via indirect enhancement of GABA function.

Although some maintain that the shorter-acting drugs zolpidem and zaleplon come close to the ideal hypnotic, confusional arousals early in the night may occur and lead to potentially dangerous behaviours. Care should be taken particularly if there is a recent or past childhood history of parasomnia activity.

If sleep maintenance insomnia is the primary symptom, some subjects benefit from zolpidem or zaleplon taken through the night if they do not have to be awake for the next four hours.


Melatonin is produced naturally in the pineal gland and is best viewed as a natural hormone that facilitates sleep onset. Its release in the brain is maximal around dusk and is under tight circadian control. The robust diurnal pattern of melatonin release appears to diminish markedly with increasing age.

When used pharmacologically, melatonin can act to ‘advance’ a subject's clock at low doses (typically 0.5 mg or less) or to generally improve sleep quality at higher doses (2–10 mg). The former use can help jet lag or those with an extreme ‘night owl’ tendency. Timing is very important and the drug should be taken around 2–3 hours before desired sleep onset.

More commonly, melatonin is used as a simple hypnotic. It appears generally very safe with few adverse effects although it may lack potency. It is used particularly in very young or elderly populations although controlled evidence for efficacy is lacking. Prolonged-release preparations of melatonin have gained a licence for insomnia in the elderly in many countries although availability varies considerably. In some areas, melatonin is routinely used as a nutritional supplement bought ‘over the counter’, despite concerns over quality control and accurate dosing.

Other hypnotic drugs

In the past, a variety of agents, such as chloral hydrate and chlormethiazole, have been used as powerful hypnotic agents. Concerns over safety and dependence have dramatically reduced prescribing levels.

Sodium oxybate is a new drug licensed for use in narcolepsy (Chapter 3) which acts as an extremely effective hypnotic. It is short acting and enhances deep non-REM sleep as its main pharmacological action. It is the sodium salt of gamma hydroxybutyrate (GHB), a drug misused notoriously as a ‘date rape’ agent. Largely due to expense and concerns over abuse, sodium oxybate is unlikely to be used for simple or primary insomnia in the foreseeable future.


It is common practice to prescribe antidepressant drugs as a surrogate treatment for insomnia, even in the absence of frank depressive symptomology. The various classes of antidepressants vary in their sedative properties and effects on sleep architecture (Chapter 10).

The tendency for most antidepressants to suppress REM sleep has led to their routine use in treating REM sleep-related phenomena. Frequent episodes of REM sleep paralysis or cataplexy often respond well to tricyclic preparations or venlafaxine, for example.

Analgesics including anti-epileptic agents

Nocturnal pain due to a variety of causes is a significant factor in many subjects with insomnia. Poor quality sleep has also been shown to reduce pain thresholds and may therefore promote hyperalgesia. Improving sleep in patients with generalised pain syndromes such as fibromyalgia may indirectly help with pain control.

Although many agents used for neuropathic pain are sedative, relatively few improve sleep quality or continuity. Tricyclic antidepressants often worsen any tendency for restless legs syndrome and rarely produce refreshing sleep. Opiates are also not generally effective at improving the depth of sleep and can lead to significant breathing disorders, such as central sleep apnoea (Chapter 4).

Some of the agents initially developed as anti-epileptics are effective both in treating neuropathic pain and enhancing deep non-REM sleep. When sleep patterns have been assessed in drug trials for pain syndromes, pregabalin and gabapentin appear particularly helpful.

Potential sleep-related side effects of commonly used drugs

It is often difficult to predict how an individual drug prescribed as a treatment for a medical or psychiatric condition will affect a subject's sleep–wake cycle. Clearly, drugs that adversely affect the quantity or quality of overnight sleep might produce symptoms both of insomnia and/or daytime somnolence. Some potentially unrecognised consequences of commonly used classes of drugs are outlined in Table 11.2.

Table 11.2 Possible adverse consequences of commonly used classes of drug on the sleep–wake cycle.

Class of drug

Potential side effect on the sleep–wake cycle

Benzodiazepines and other hypnotics

Although total sleep time may be enhanced, proportionally more light sleep is usually obtained potentially leading to unrefreshing nocturnal sleep and daytime somnolence

Enhanced snoring or even sleep apnoea may be observed

Evidence from case reports suggest that short-acting hypnotics such as zolpidem may trigger unwelcome parasomnia activity in those predisposed


It is often difficult to predict whether an antidepressant drug (ADD) will be relatively alerting or sedative to an individual patient

The majority of ADDs will tend to suppress REM sleep and, therefore, reduce dreams and nightmares if excessive. However, some ADDs (notably venlafaxine and mirtazapine) have the potential for suppressing certain elements of REM sleep, such as muscle paralysis, and may, therefore, induce or exacerbate REM sleep behaviour disorder

Most ADDs have the capacity for worsening restless legs syndrome and associated periodic limb movements during sleep

Even strongly sedative ADDs rarely produce sleep that is particularly refreshing to the subject

Dopamine blocking agents

The majority of antipsychotic agents (neuroleptics) will be sedative and have the capacity of inducing weight gain (increased risk of obstructive sleep apnoea) 

Antipsychotics and many anti-emetic drugs will worsen restless legs syndrome


Some subjects appear particularly sensitive to the effects of opiates on overnight breathing control such that central sleep apnoea may be induced. This may mimic obstructive sleep apnoea (OSA) on investigations but not be controlled by standard OSA therapies such as positive ventilations masks or mandibular advancement

Anti-epilepsy drugs

Anti-epilepsy drugs (AEDs) have variable effects on sleep architecture although most are generally sedative 

Gabapentin and pregabalin tend to increase deep or slow wave sleep and may possibly improve sleep quality if it is generally disturbed or if anxiety is present

Lamotrigine is usually arousing although inter-individual differences are considerable

Further reading

Artigas, F., Nutt, D. and Shelton, R. (2002) Mechanism of action of antidepressants. Psychpharmcol Bull36, 123–132.

Brown, M.A. and Guilleminault, C. (2011) A review of sodium oxybate and baclofen in the treatment of sleep disorders. Curr Pharm Des17, 1430–1435.

Buscemi, N., Vandermeer, B., Friesen, C. et al. (2007) The efficacy and safety of drug treatments for chronic insomnia in adults: a meta-analysis of RCTs. J Gen Intern Med22, 1335–1350.

Mitler, M.M. and Hajdukovic, R. (1991) Relative efficacy of drugs for the treatment of sleepiness in narcolepsy. Sleep14, 218–220.

Srinivasan, V., Brzezinski, A., Pandi-Perumal S.R. et al. (2011) Melatonin agonists in primary insomnia and depression-associated insomnia: are they superior to sedative-hypnotics? Prog Neuropsychopharmacol Biol Psychiatry35, 913–923.

Tzellos, T.G., Toulis, K.A., Goulis, D.G. et al. (2010) Gabapentin and pregabalin in the treatment of fibromyalgia: a systematic review and a meta-analysis. J Clin Pharm Ther35, 639–656.

If you find an error or have any questions, please email us at Thank you!