Clinical manifestations

Presentation and course

Clinical features of insomnia induced by drugs are like those due to other causes. Temporal association with the use of drugs known to cause sleep disturbances is an important feature. Some of the clinical manifestations of various sleep disorders shown in Table 1 may be linked. For example, sleep apnea may manifest as insomnia. Sleep may be disturbed as a sequel of other drug-induced adverse drug reactions. Patients with movement disorders may have difficulty sleeping and patients on diuretics may have to get up frequently at night to urinate.

Prognosis and complications

The prognosis for recovery from drug-induced insomnia and daytime sleepiness is good if the responsible medication is identified and discontinued. However, outcome of insomnia or daytime drowsiness due to drug withdrawal varies according to the drug involved and underlying disorders of the patient.

Biological basis

Etiology and pathogenesis

Biological basis of drug-induced neurologic disorders is interaction of a drug with 1 or more of the neurotransmitters or receptors that are involved in sleep and wakefulness. Multiple interactions between the disease being treated, preexisting sleep disorders, genetic factors, and direct or indirect effects of drugs are possible (19). Etiology should be considered according to the type of disturbance.

Central disorder of hypersomnolence. The term "excessive sleepiness" should be differentiated from somnolence, which literally means sleepiness, because the term signifies abnormal drowsiness in a medical context. ICSD-3 refers to this group of sleep disorders as those in which “the primary complaint is daytime sleepiness not caused by disturbed nocturnal sleep or misaligned circadian rhythms.” Drug or substance abuse is considered as a cause when it is responsible for extended periods of sleepiness. The term “hypersomnolence” is rarely used in pharmacovigilance, either in clinical trials or postmarketing surveillance. However, the more frequently used terms “sleepiness” and “drowsiness” often overlap and are used interchangeably in various reports. Falling asleep suddenly during the day ("sleep attack") has also been described as an adverse reaction to some drugs and should be differentiated from narcolepsy. According to ICSD-3 criteria, narcolepsy type 1 is characterized by excessive daytime sleepiness, cataplexy, rapid eye movement sleep phenomena such as sleep paralysis, and hypnagogic hallucinations.

Drowsiness. Drug-induced drowsiness or sleepiness is usually a self-reported event, and its effect on performance is difficult to assess. Bayesian methods have been used to estimate sensitivity to drug-induced sleepiness (06). Nonhypnotic drugs that produce drowsiness are listed in Table 2.

Table 2. Nonhypnotic Drugs That May Have Excessive Drowsiness as a Side Effect

Insomnia due to drugs or substances. ICSD-3 defines insomnia as “a repeated difficulty in sleep initiation, duration, consolidation, or quality that occurs despite adequate opportunity and circumstances for sleep, and results in some form of daytime impairment.” This is a difficult symptom to evaluate. The term "rebound insomnia," not included in ICSD-3, is used in the literature to describe sleep disturbance characterized by an increase in wakefulness above the previous baseline level after the withdrawal of sedative-hypnotic drugs. It is not synonymous with "withdrawal insomnia," which implies drug dependence and is usually seen with long-term use of hypnotics. Other hypnotic drugs, such as chloral hydrate, can also lead to withdrawal problems. Withdrawal from sedative or hypnotic drugs is defined as a temporary increase in the severity of insomnia after stopping treatment. If insomnia is worse than it was before the treatment started, the term "recoil" or "overshoot insomnia” is used rather than "rebound insomnia.”

Patients with preexisting insomnia are at greater risk for developing aggravation of this symptom as an adverse effect of drugs. Drugs and substances associated with insomnia are listed in Table 3, and further details are given under various drugs along with other sleep disorders.

Table 3. Drugs and Substances Associated with Insomnia

Antidepressants. Sleep disturbances are generally more prevalent among patients with depression. These disturbances may improve with antidepressant treatment, but there may also be adverse effects due to antidepressant drugs. Some antidepressants adversely affect the physiological structure of sleep, whereas others restore it. Most antidepressants cause REM sleep reduction, generally with increased serotonin function. Intense and prolonged dreams often accompany abrupt withdrawal from antidepressant drugs as a manifestation of REM sleep rebound after drug-induced REM sleep deprivation.

Tricyclic antidepressants promote sleep in patients with insomnia due to depression and are likely to cause daytime drowsiness.

Monoamine oxidase inhibitors have also been shown to suppress REM sleep in patients with depression, even eliminating REM sleep in some patients. This class of antidepressants also appears to reduce total sleep time and may decrease sleep efficiency.

Serotonin/norepinephrine reuptake inhibitors and selective serotonin-reuptake inhibitors can derange sleep architecture and decrease restorative sleep (08). Selective serotonin reuptake inhibitors are more likely to cause insomnia. Fluoxetine significantly reduces sleep efficiency as well as REM time. The mechanism responsible for this is not known, but it may be due to a general increase in central arousal. Fluoxetine may disrupt sleep due to so called “activating effects,” whereas antidepressants with sedative properties, eg, doxepin, mirtazapine, and trazodone rapidly improve sleep but may cause sleep problems in long-term treatment due to oversedation (21).

Antiemetic drugs. Most antiemetic drugs penetrate the blood-brain barrier and produce sleepiness by their action on the dopaminergic, histaminergic, or cholinergic systems. Hyoscine is a short-acting but powerful anticholinergic agent that reduces REM sleep but increases light (stage 2) sleep. REM sleep is increased on withdrawal of the drug. Domperidone is an exception because it acts on the chemoreceptor trigger zone outside the blood-brain barrier and is unlikely to cause sleepiness.

Antiepileptic drugs. The older antiepileptic drugs are associated with a variety of sleep disturbances including marked reduction in REM sleep and insomnia. Phenobarbital decreases sleep onset latency and improves sleep continuity during short-term administration to patients with epilepsy, but risk evaluation and mitigation strategies (REMS) is suppressed. It is associated with daytime sleepiness. Newer antiepileptic drugs have less of these disturbances. In a cross-sectional analysis of veterans attending an epilepsy clinic, 40% reported insomnia, which was significantly associated with post-traumatic seizure etiology and lamotrigine, whereas levetiracetam treatment was associated with lower odds for insomnia (13).

Antihistaminics. Histamine, as a neurotransmitter, takes part in the regulation of sleep. Therefore, sleep is affected by antihistaminics. The older antihistaminics such as triprolidine cause daytime sleepiness. The newer H1-antagonists, such as terfenadine, either do not penetrate the blood-brain barrier or enter the brain slowly, and they do not induce sleep, except in high doses. Another explanation is that H2 blockers have anticholinergic and gamma-aminobutyric acid-like properties that could contribute to central nervous system reactions. Among H2-antagonists, cimetidine increases the duration of slow wave sleep, but ranitidine does not. Ranitidine, however, has been reported to cause sleep disturbances, which resolve on discontinuation of the drug and recur on resumption of therapy (20).

Antiobesity drugs. Insomnia is the most common adverse effect with drugs targeting monoamine systems. These drugs include sibutramine, bupropion, and tesofensine, and they have some positive effects on mood and anxiety with added therapeutic benefits in obese patients (15).

Antiparkinsonian drugs. Daytime sleepiness occurs frequently in Parkinson disease as the condition progresses due to pathological changes in areas of the brain involved in sleep-wake regulation, and dopamine agonists may exacerbate sleepiness in some patients. Parkinson disease patients treated with dopaminergic agents have frequent abnormal sleep patterns and hallucinations. Dopaminergic drugs have a biphasic influence on sleep, ie, they can cause both insomnia and excessive sleepiness. The exact mechanism of this paradoxical effect is not known. Lower doses of these drugs may downregulate dopaminergic input to the reticular activating system. For example, small doses of levodopa appear to improve sleep, although higher doses cause insomnia. Different dopaminergic receptor types or changes in receptor sensitivity may explain these phenomena.

There are case reports of sudden episodes of daytime sleepiness in patients taking ropinirole or pramipexole (motor vehicle accidents have resulted from some patients falling asleep when driving), and daytime sleepiness ceased when the drugs were discontinued. All dopamine agonists have the potential to produce sleep attacks during the day. Isolated cases have also been reported with bromocriptine, lisuride, pergolide, or piribedil. Excessive daytime sleepiness has been reported in patients treated by use of pramipexole, cabergoline, and levodopa as monotherapy with no significant differences between the 3 drugs. Important predictive factors are high doses, older age, and advanced disease. Sedative effect varies among different dopamine agonists. A study has shown that switching from pramipexole or ropinirole to piribedil maintains the same therapeutic motor effect but reduces daytime sleepiness to a clinically relevant degree in patients with excessive daytime sleepiness (04). Higher prevalence of sleepiness in patients with Parkinson disease, as compared with those suffering from other neurologic diseases, is related to the stage of the disease and medications, but the causes of variability in sedation levels in patients with Parkinson disease are mostly unknown. The possibility that individual susceptibility to a specific antiparkinsonian drug may play a role in the genesis of sleepiness should be kept in mind.

Antipsychotic drugs. Antipsychotics have been classified as high somnolence (clozapine), moderate somnolence (olanzapine, perphenazine, quetiapine, risperidone, ziprasidone), and low somnolence (aripiprazole, asenapine, haloperidol, lurasidone, paliperidone, cariprazine), but other factors such as dose and condition of the patient also influence the frequency and degree of somnolence (05). A study indicates that excessive sleepiness in patients with psychiatric disorders may not be solely related to medication but also to low levels of activity and other sleep disorders (16). The mechanisms of antipsychotic-induced somnolence are multifactorial, although the blockade of histamine 1 receptors and alpha 1 receptors may play a major role. Further research should sort out interactions of sleepiness, mood, activity, and psychotic symptoms for strategies to intervene for excessive sleepiness in psychosis.

Cardiovascular drugs. In this category of drugs antihypertensives are most likely to cause insomnia after prolong use. Offending medications are beta blockers alone or in combination with angiotensinogen receptor II blockers or angiotensin converting enzyme inhibitors.

Beta blockers. Sleep disruption and excessive daytime sleepiness caused by beta blockers depend on the lipid solubility of the drugs. Propranolol, which is the most lipid-soluble drug of this group, is most liable to cause sleep disruption, and atenolol, the least lipid-soluble drug, is the least likely to cause sleep disruption.

Calcium channel blockers. The exact pathomechanism of flunarizine-induced insomnia is not known, but it may be related to interference with the dopaminergic system.

Cholesterol-lowering drugs. The proposed pathomechanism is related to the differential penetration of these agents across the blood-brain barrier, a property related to lipophilicity. Simvastatin, one drug of this class, has been reported to produce less sleep disturbances, indicating that lipophilicity alone is not responsible for these side effects.

CNS-stimulant drugs. Wake-promoting drugs such as modafinil are used therapeutically for the treatment of narcolepsy and have been shown to increase extracellular dopamine. Dopamine transporters are necessary for the specific wake-promoting action of modafinil, as experimental animals with deletions of dopamine transporter gene are nonresponsive to modafinil but hypersensitive to the wake-promoting effects of caffeine.

CNS-stimulating drugs may lead to disruption of sleep and subsequent excessive daytime sleepiness. At high levels of arousal, caffeine in moderate doses may induce overarousal, leading to prolonged wakefulness and impaired sleep. In clinical trials, insomnia has been reported to be increased in children with attention-deficit hyperactivity disorder who are receiving the CNS stimulant methylphenidate as compared to children in the placebo group. However, a randomized controlled study of atomoxetine in children with attention deficit hyperactivity disorder and autism spectrum disorder found no significant differences between treatment and placebo groups, including Children's Sleep Habits Questionnaire (CSHQ) 33-item total score, total hours of sleep per day, and total minutes awake after sleep onset at the study endpoint (07). This trial concluded that atomoxetine is “sleep neutral” and suggested its use in preference to other psychostimulants when sleep disturbance is already an issue in such patients.

Nicotine administered as a transdermal patch in nonsmokers can cause disruption of the sleep architecture by increased catecholamine release due to stimulation of the central nicotinic cholinergic pathways.

Corticosteroids. High doses of prednisone are associated with insomnia. Polysomnographic studies have shown that oral dexamethasone has a disrupting effect on sleep characterized by increased wake time and stage 1 sleep and reduced time spent in REMS.

Hypnotic drugs. Excessive sleepiness may be an extension of the effect of hypnotic drugs, either due to excess dosage or an idiosyncratic response of the patient. Excessive daytime sleepiness may be due to the disruption of the sleep-wake cycle. The FDA has requested that all manufacturers of sedative-hypnotic drug products should strengthen their product labeling to include stronger language concerning potential risks. These risks include complex sleep-related behaviors like sleep-driving, defined as driving when not fully awake after ingestion of a sedative-hypnotic product, with no memory of the event. Revised labeling includes the following medications:

Pathomechanism of rebound insomnia. This is not well understood. Psychosocial factors and behavioral problems may play a role in development of rebound insomnia. Knockout studies, together with genetic studies in mouse models, have shown that withdrawal from zolpidem is influenced by a chromosome 11 locus, which points to the Gabrg2 gene, a GABA receptor subtype gene, as a promising candidate to underlie phenotypic differences in sedative-hypnotic physiological dependence and associated withdrawal episodes (09).

It is generally believed that rebound insomnia is related to rapid elimination of benzodiazepines, which results in a CNS deficiency of inhibitor mechanisms. Hypnotics suppress REM sleep movements, subsequently causing a compensatory excess of REM sleep after withdrawal.

Benzodiazepines are known to facilitate GABA receptor function, which is a major inhibitory system in the CNS, and withdrawal symptoms can be regarded as that of GABA deficiency. It may be presumed that during the period of drug administration, the production of endogenous benzodiazepine is suppressed by exogenous diazepam. This concept is compatible with a lack of rebound effect with long-acting benzodiazepines (slow elimination) or gradual reduction of the dose of short-acting benzodiazepine, which attenuates the rebound effect.

A randomized placebo-controlled study has investigated the likelihood of primary insomniacs experiencing rebound insomnia and a withdrawal syndrome on repeated placebo substitutions over 12 months of nightly use of zolpidem, which is a nonbenzodiazepine hypnotic but binds to GABA(A) receptors at the same location as benzodiazepine (17). No clinically significant manifestations of withdrawal or differences in rebound insomnia rates between the placebo group and zolpidem group were observed on the discontinuation nights over the 12 months of nightly use. It was concluded that chronic nightly hypnotic use at therapeutic doses for primary insomnia does not lead to rebound insomnia or withdrawal symptoms.

Another hypothesis is that rebound insomnia is related to tolerance and decreased receptor density. The point against this hypothesis is that tolerance is related to the duration of administration, which does not alter the intensity of rebound insomnia. Another explanation is that less sleep is needed following the discontinuation of hypnotics once the patient has reached sleep satiation (ie, after the sleep need, induced by hypnotics, has been satisfied). Observations support this hypothesis, which show that subjects with insomnia who receive hypnotics show reduced daytime sleepiness.

Nevertheless, residual next-day effects of sleep-promoting drugs are common and an important safety issue, which may require discontinuation of the drug. A review of clinical trials of lemborexant, a dual orexin receptor antagonist approved in the United States and Japan for treatment of insomnia in adults, showed that there was no significant impairment in next-day functioning among healthy subjects as well as subjects with insomnia (14). Although somnolence was the most common adverse event reported with lemborexant treatment, it was typically mild to moderate in severity and rarely caused discontinuation of the study drug.

Epidemiology

Sleep disorders are common, but there are no epidemiological studies to determine the percentage of these caused by drugs. A study of a nationally representative sample of U.S. adults from the 1999 to 2016 National Health and Nutrition Examination Survey shows that the concurrent use of 1 or more medications with insomnia as a side effect has increased by 66% and 164% (03). Compared to nonusers, those who took 2 or more of these medications were more likely to report insomnia symptoms, daytime sleepiness, and difficulty with at least 2 daytime activities due to sleepiness or tiredness. These findings also emphasize the increased risks of insomnia associated with polypharmacy. Some examples with specific drugs are:

Monoamine oxidase inhibitors. Isocarboxazid-associated insomnia has been reported in patients treated with this drug for atypical depression. Moclobemide, a reversible monoamine oxidase inhibitor, is also associated with insomnia.

Lorcainide. This antiarrhythmic agent is associated with sleep disturbances (difficulty in falling asleep, nightmares, and vivid dreams).

Flunarizine. This is a calcium channel blocker used for the treatment of migraine and has been reported to be associated with insomnia.

Pramipexole. In clinical trials, somnolence has been reported in early Parkinson disease patients treated with pramipexole, a synthetic dopamine agonist.

Prevention

Observance of generally accepted methods of sleep hygiene reduces the probability of development of drug-induced sleep disorders.

The important preventive measure is avoidance of the use of drugs known to produce disturbances of sleep. Long-term use of sedative-hypnotic drugs should be avoided. Although most withdrawal reactions have been reported with long-acting benzodiazepines, rebound insomnia occurs with greater frequency and severity with short-acting agents. Benzodiazepine therapy should be stopped as early as possible with tapering after moderate dose, prolonged-use therapy, or both. In a multicenter, open-label study of Japanese patients with insomnia, patients were given a common clinical dose of eszopiclone (2 mg) for 24 weeks, after which the treatment was abruptly discontinued and withdrawal symptoms were evaluated using the Benzodiazepine Hypnotics Withdrawal Symptom Scale (11). Results indicate that dependence and poor compliance may contribute to withdrawal symptoms with long-term benzodiazepine use. Measures to ensure proper compliance is possibly the best approach to reduce withdrawal symptoms.

CNS depressants should be avoided in infants less than one year of age, as they may cause sudden death in apnea-prone infants.

The FDA, based on its review of the Adverse Event Reporting System database, found 62 cases of serious injuries from complex sleep behaviors such as sleepwalking, sleep-driving, and sleep-using-a-stove after taking eszopiclone, zaleplon, or zolpidem. A review of the medical literature revealed 4 more cases reported from 1992 to 2018. Deaths occurred in 20 of these 66 cases. The FDA issued the following statement:

"As a result, we are requiring a Boxed Warning, our most prominent warning, to be added to the prescribing information and the patient Medication Guides for these medicines. We are also requiring a Contraindication, our strongest warning, to avoid use in patients who have previously experienced an episode of complex sleep behavior with eszopiclone, zaleplon, and zolpidem."

Differential diagnosis

Poor sleep hygiene. Persons with poor sleep hygiene are more susceptible to drug-induced insomnia. For example, those who watch TV late in the evening and drink stimulant beverages or alcohol are more liable to manifest drug-induced insomnia.

Nondrug-induced sleep disorders. Several neurologic disorders as well as those involving other systems may render the patient more susceptible to drug-induced insomnia or daytime sleepiness, eg, psychiatric disorders.

Rebound or withdrawal insomnia. This may occur due to discontinuation of a hypnotic drug.

Drug-induced parasomnias. Several parasomnias such as sleepwalking or night terrors may awaken the patients and interfere with sleep.

Confusing conditions

Drug-induced sleep disorders need to be differentiated from the disorders known to occur as a manifestation of diseases being treated with drugs. Discontinuation of the suspected drug with resulting improvement may help in the diagnosis.

It is important to differentiate common types of insomnia from drug-induced insomnia and pathological sleep because medications commonly used for insomnia, such as benzodiazepines, may aggravate pathological sleep.

Patients with advanced Parkinson disease have sleep disturbances that need to be differentiated from those induced by dopaminergic agents.

Sleep attacks induced by drugs need to be differentiated from narcolepsy.

Sleep disturbances are a feature of delirium, which is often drug-induced. The sleep-wake cycle is disrupted in delirium, leading to drowsiness during the day and insomnia at night.

Diagnostic workup

Because there are no characteristic clinical features of sleep disorders associated with drugs, the first (and most important) step is elucidation of a complete drug history. In drug-induced drowsiness, blood levels of a suspected drug should be accurately measured. For example, in an elderly patient with phenytoin toxicity who presented with drowsiness and gait disturbance, free phenytoin level was calculated to be 27 ng/dL (normal 10 to 20 ng/dL) after taking his albumin level into account, whereas total serum phenytoin level was reported to be normal (10). Phenytoin was discontinued with resolution of symptoms.

The diagnosis is based on findings of sleep laboratory examinations to document the disturbances related to drug use. A proper assessment of sleep disorders requires an overnight polysomnographic study that is usually done in a sleep laboratory. These studies document changes in sleep stages, number of arousals, sleep-related breathing disorders, and movements during sleep. This is not only important for the evaluation of sleep disorders, but also for investigation of the action of drugs on sleep. Activity monitoring by wrist actigraphs has proven its usefulness as an efficient method to assess the rest-activity cycle in sleep research. This technique is capable of measuring drug-induced changes in nocturnal and diurnal behavior.

Management

Drug-induced insomnia. Management usually involves discontinuation of the offending medication. Another possibility to consider is changing the time of administration of the suspected drug. For example, in case of insomnia associated with donepezil, the drug may be administered in the morning rather than in the evening. In situations where the responsible medication cannot be discontinued, drugs for management of insomnia may be considered, taking care to avoid drug interactions. Some examples of this approach are:

Drug-induced sleep disturbances in Parkinson disease. Currently available data are insufficient to provide effective guidelines for prevention and treatment of sleep events in patients taking dopamine agonists for Parkinson disease. Parkinson disease patients at risk of sleep attacks during the day can be managed by dose reduction of dopamine agonists and a change to alternative medications. Modafinil may be effective in reducing excessive daytime sleepiness in Parkinson disease patients who are treated with dopaminergic drugs. Deep brain stimulation of the subthalamic nucleus in patients with Parkinson disease may improve sleep quality.

Future. Drugs frequently cause sleep fragmentation either as a part of intended action or an adverse drug reaction. A better understanding of the sleep mechanisms may lead to development of drugs without the unintended side effects. One of the approaches is study of the circadian sleep patterns by EEG following drug exposure. Use of the Makarov model in the telemetered rat has shown that methylphenidate was a 5-fold more potent inducer of sleep fragmentation as compared to a new chemical entity (02).

Special considerations

Anesthesia

Anesthetic agents, like other CNS depressants, decrease neural output to upper airway muscles more than they depress the phrenic nerve activity. It is a common observation that normal persons develop upper airway obstruction during light general anesthesia if proper position of the head and neck is not maintained. Severe central apnea can develop following general anesthesia.

The use of opioids should be avoided in patients with preexisting sleep apnea syndrome, and special care is required during intubation and in the postanesthetic period. They should not be extubated until fully conscious.