Trouble Falling Asleep: Drugs That Cause or Treat It

Why Some Medications Keep You Awake

Several drug classes raise brain arousal signals or suppress the neurochemical pathways that initiate sleep, and the result is a measurably longer time from lights-off to sleep onset (called sleep latency). A 2018 analysis in the Journal of Clinical Sleep Medicine found that medication-related insomnia accounted for roughly 10 percent of chronic insomnia referrals in an academic sleep center [1]. Recognizing the pattern early prevents months of unnecessary hypnotic prescriptions.

Selective serotonin reuptake inhibitors (SSRIs) such as fluoxetine and sertraline increase serotonergic tone in the dorsal raphe, a wake-promoting nucleus. In a polysomnography study of 60 patients starting fluoxetine 20 mg, mean sleep-onset latency rose from 18 minutes at baseline to 31 minutes at week four [2]. Serotonin-norepinephrine reuptake inhibitors (SNRIs) like venlafaxine and duloxetine carry a similar profile, with reported insomnia rates of 13 to 24 percent in prescribing data submitted to the FDA.

Beta-adrenergic blockers, particularly lipophilic agents like propranolol and metoprolol, cross the blood-brain barrier and suppress nocturnal melatonin secretion. A double-blind crossover trial (N=9) published in Clinical Pharmacology & Therapeutics showed that atenolol 50 mg reduced overnight melatonin AUC by 64 percent compared to placebo [3]. That suppression directly prolongs sleep onset. Switching to a hydrophilic beta-blocker such as nadolol may preserve melatonin release and reduce sleep complaints.

Corticosteroids represent another frequent offender. Prednisone at doses above 20 mg daily increases wakefulness after sleep onset and delays initial sleep, partly through hypothalamic-pituitary-adrenal axis activation. Morning dosing helps, but it does not fully prevent the effect.

Stimulants and Wake-Promoting Agents

Amphetamine-based medications and methylphenidate are prescribed to more than 4.1 million U.S. adults for ADHD, and insomnia is the most commonly reported side effect. The effect is dose-dependent and predictable.

Methylphenidate extended-release has a functional duration of 8 to 12 hours. A patient who takes a dose at 8 a.m. may still have clinically relevant plasma levels at 8 p.m. The ADHD medication trial database maintained by the National Institute of Mental Health reports insomnia rates of 27 percent for mixed amphetamine salts XR vs. 13 percent for placebo [4]. Lisdexfetamide produces similar numbers. Timing the last dose before noon and avoiding booster IR doses after 2 p.m. are practical first steps.

Modafinil and armodafinil, prescribed for narcolepsy and shift-work disorder, have half-lives of 12 to 15 hours. Taking modafinil after 10 a.m. pushes wake-promoting dopaminergic activity well into the evening. The FDA label for modafinil lists insomnia in 5 percent of trial participants at 200 mg, rising to 11 percent at higher doses.

Caffeine, though not a prescription drug, deserves mention because it competes at adenosine A1 and A2A receptors with a half-life of 5 to 6 hours. A controlled study published in the Journal of Clinical Sleep Medicine found that 400 mg of caffeine taken 6 hours before bed still increased sleep latency by a mean of 20 minutes [5].

Other Drug Classes That Disrupt Sleep Onset

The list extends beyond stimulants and antidepressants. Thyroid hormone replacement at supraphysiologic doses produces a hypermetabolic state that fragments sleep architecture. Levothyroxine overreplacement, defined as a TSH below 0.4 mIU/L in a euthyroid target patient, increases self-reported insomnia by 42 percent according to a retrospective cohort study in Thyroid (N=830) [6].

Bronchodilators such as albuterol and theophylline activate beta-2 adrenergic receptors centrally. Theophylline has the added problem of being a methylxanthine (structurally related to caffeine), and at serum levels above 10 mcg/mL it produces measurable increases in sleep latency [7].

Decongestants containing pseudoephedrine act as indirect sympathomimetics. The American Academy of Sleep Medicine (AASM) clinical practice guideline on pharmacologic treatment of chronic insomnia notes: "Clinicians should conduct a thorough medication review, including OTC agents, before initiating hypnotic therapy" [8]. Diuretics taken in the evening cause nocturia, which fragments sleep. Nicotine replacement therapy patches that deliver nicotine overnight are another underrecognized contributor.

Some antihypertensives such as clonidine can paradoxically cause rebound insomnia when doses are missed, because alpha-2 agonist withdrawal triggers noradrenergic surges. Statins have been suspected of causing insomnia, but a meta-analysis of 8 randomized trials (N=2,420) found no statistically significant increase in sleep complaints compared to placebo [9].

First-Line Treatment: CBT-I Before Pills

The American College of Physicians (ACP) 2016 guideline recommends cognitive behavioral therapy for insomnia (CBT-I) as the initial treatment for chronic insomnia disorder in all adults [10]. Dr. Alon Avidan, director of the UCLA Sleep Disorders Center, has stated: "CBT-I is the most durable treatment we have for insomnia. Its effects persist for years after the therapy ends, unlike any medication on the market."

CBT-I typically includes stimulus control, sleep restriction, relaxation training, and cognitive restructuring. A meta-analysis of 20 RCTs (N=1,162) published in Annals of Internal Medicine showed that CBT-I reduced sleep-onset latency by a weighted mean of 19 minutes (95% CI: 14 to 24) and increased total sleep time by 26 minutes [10]. The benefit persisted at 12-month follow-up.

Digital CBT-I programs (dCBT-I), including FDA-cleared platforms like Pear Therapeutics' Somryst, have shown non-inferior outcomes in shorter time frames of 6 to 9 weeks. Not every patient can access or tolerate CBT-I, however, and pharmacotherapy remains necessary for many.

Prescription Drugs That Treat Sleep-Onset Insomnia

When behavioral therapy alone is insufficient, several drug classes carry FDA approval or strong evidence for reducing the time to fall asleep. The choice depends on the patient's age, comorbidities, substance use history, and whether sleep-maintenance difficulty coexists with sleep-onset trouble.

Dual Orexin Receptor Antagonists (DORAs)

Suvorexant (Belsomra) and lemborexant (Dayvigo) block orexin-A and orexin-B receptors to reduce wake drive. The SUNRISE-1 trial (N=1,006) demonstrated that lemborexant 5 mg reduced objective sleep-onset latency by 10.5 minutes vs. placebo and 12.5 minutes for the 10 mg dose over 30 nights of polysomnography [11]. According to the AASM 2017 clinical practice guideline, suvorexant is conditionally recommended for both sleep-onset and sleep-maintenance insomnia [8].

DORAs carry a lower risk of dependence compared to benzodiazepine receptor agonists. Their most reported side effect is next-morning somnolence (7 to 10 percent at approved doses). They should not be combined with strong CYP3A4 inhibitors such as ketoconazole, which can triple suvorexant plasma concentrations.

Benzodiazepine Receptor Agonists (Z-drugs)

Zolpidem (Ambien) remains the most widely prescribed sleep-onset medication in the United States, with over 25 million prescriptions dispensed in 2023. The immediate-release formulation at 5 mg (women) or 5 to 10 mg (men) reduces sleep latency by a mean of 23 minutes compared to placebo in pooled trial data [12]. The lower dose recommendation for women reflects sex-based pharmacokinetic differences: women clear zolpidem roughly 45 percent more slowly, increasing next-morning impairment risk, as the FDA noted in its 2013 label revision.

Zaleplon (Sonata) has an ultra-short half-life of one hour, making it appropriate for patients whose sole complaint is difficulty initiating sleep without any maintenance difficulty. It can even be taken during a middle-of-the-night awakening, provided at least four hours of bed time remain. Eszopiclone (Lunesta) covers both onset and maintenance with a longer half-life, but its metallic taste side effect occurs in up to 34 percent of patients at 3 mg [13].

All Z-drugs carry FDA boxed warnings for complex sleep behaviors including sleepwalking and sleep-driving. They are Schedule IV controlled substances and are generally recommended for short-term use of 2 to 4 weeks, though some patients use them intermittently for longer periods under clinician supervision.

Melatonin Receptor Agonist

Ramelteon (Rozerem) selectively activates MT1 and MT2 melatonin receptors in the suprachiasmatic nucleus. It is the only FDA-approved prescription insomnia medication without abuse potential or DEA scheduling. In a key trial (N=829), ramelteon 8 mg reduced subjective sleep-onset latency by 14 minutes vs. placebo [14]. The effect is modest compared to Z-drugs, but the safety profile makes ramelteon preferred for older adults and patients with a substance use history. It does not cause rebound insomnia on discontinuation.

Dr. Michael Sateia, lead author of the AASM pharmacologic insomnia guidelines, noted in the guideline commentary: "Ramelteon provides a lower-risk option for patients in whom the clinician prefers to avoid controlled substances, though its efficacy is more modest" [8].

Low-Dose Doxepin

Doxepin at 3 to 6 mg (brand name Silenor) is FDA-approved specifically for sleep-maintenance insomnia, not sleep onset. At these sub-antidepressant doses, it acts primarily as a histamine H1 antagonist. Pooled data from two phase III trials (N=831) showed it reduced wake after sleep onset by 22 to 29 minutes but had a smaller impact on initial sleep latency [15]. It may still benefit patients who report both onset and maintenance difficulty. Its advantage is the absence of controlled substance classification and minimal next-day impairment at the 3 mg dose.

Trazodone (Off-Label)

Trazodone at 25 to 100 mg is the most commonly prescribed off-label sedative for insomnia in the United States, with roughly 20 million prescriptions annually. A meta-analysis of 7 trials (N=429) found that trazodone reduced subjective sleep latency by about 10 minutes, a statistically significant but clinically modest effect [16]. The sedation comes from 5-HT2A antagonism and H1 blockade. Orthostatic hypotension, especially in older adults, and rare priapism in males are key safety concerns. The AASM guideline recommends against trazodone due to limited high-quality evidence, despite its widespread use [8].

How to Identify a Drug-Induced Sleep Problem

A structured medication timeline is the single most useful diagnostic step. Map when each medication was started or dose-changed against when the sleep complaint began. Look for onset within one to four weeks of a new drug or dose escalation.

Ask the patient about OTC products. Approximately 20 percent of adults who report insomnia are using at least one sympathomimetic decongestant or stimulant-containing supplement without recognizing its impact on sleep [17]. Energy drinks, pre-workout formulas, and weight-loss supplements often contain 200 to 400 mg of caffeine per serving.

Polysomnography is not routinely needed for drug-induced insomnia, but actigraphy worn for 7 to 14 days can objectively measure sleep latency patterns and confirm the clinical suspicion. The American Academy of Sleep Medicine practice parameters for actigraphy support its use in evaluating insomnia [18].

Tapering and Switching Strategies

When a causative medication is identified, the response depends on whether the drug is treating a condition that requires ongoing pharmacotherapy. For SSRIs causing insomnia in a patient with well-controlled depression, switching to mirtazapine (which promotes sleep through H1 and 5-HT2A antagonism) is a reasonable option. Mirtazapine 15 mg reduced sleep latency by 21 minutes in a four-week open-label study of depressed insomniacs (N=42) [19].

For beta-blockers causing insomnia, switching from propranolol to a hydrophilic agent (atenolol or nadolol) or adding exogenous melatonin 2 to 3 mg at bedtime can offset the melatonin suppression effect. A crossover trial (N=16) showed that melatonin 2.5 mg fully reversed propranolol-induced increases in sleep latency, restoring sleep-onset time to baseline values [20].

For patients who need to continue an activating medication and cannot tolerate a switch, adding a sleep-onset pharmacotherapy is the pragmatic approach. Prescribers should match the half-life of the sleep aid to the sleep complaint. Sleep-onset difficulty alone calls for zaleplon (1-hour half-life) or ramelteon (1 to 2.6-hour half-life). Combined onset and maintenance problems warrant suvorexant (12-hour half-life) or lemborexant (17 to 19-hour half-life at 5 mg).

Special Populations

Older adults metabolize most hypnotics more slowly, raising the risk of falls, next-day sedation, and cognitive impairment. The American Geriatrics Society Beers Criteria list benzodiazepines and Z-drugs as potentially inappropriate for adults 65 and older [21]. Ramelteon and low-dose doxepin carry the best safety data in this age group.

Pregnant patients face limited pharmacologic options. The FDA has not approved any hypnotic as safe in pregnancy. CBT-I is the recommended first-line intervention. Diphenhydramine is commonly used but has minimal evidence for efficacy beyond mild sedation, and its anticholinergic burden makes it inappropriate for regular use.

Patients with obstructive sleep apnea can safely use suvorexant and lemborexant, which do not worsen apnea-hypopnea index (AHI) in clinical trial subgroup analyses. Benzodiazepines and high-dose trazodone, by contrast, can increase airway collapsibility and should be avoided in untreated moderate-to-severe OSA [22].

Melatonin 0.5 to 3 mg taken 1 to 2 hours before the target bedtime may reduce sleep latency by 7 to 12 minutes based on a Cochrane review of 12 trials (N=597), with minimal adverse effects [23]. It is available without a prescription and is most effective for circadian-phase disorders such as delayed sleep-wake phase disorder.