Eszopiclone (Lunesta) Safety Signals and FDA Actions

How Eszopiclone Works: Mechanism of Action

Eszopiclone is the (S)-enantiomer of zopiclone, a cyclopyrrolone that binds to the benzodiazepine site on GABA-A receptors containing α1, α2, α3, and α5 subunits. Its selectivity profile differs from older benzodiazepines. Eszopiclone shows preferential activity at α2- and α3-containing GABA-A receptors, which may contribute to its anxiolytic and sleep-maintenance effects beyond pure sedation 1.

The drug reaches peak plasma concentration in approximately 1 hour, with a terminal half-life of roughly 6 hours in healthy adults. In elderly patients, the half-life extends to about 9 hours, a pharmacokinetic detail that directly shaped the FDA's later dose-reduction decisions 2. Hepatic metabolism occurs primarily through CYP3A4 and CYP2E1, meaning co-administration with strong CYP3A4 inhibitors (ketoconazole, clarithromycin) can raise eszopiclone exposure significantly. The FDA label recommends a maximum dose of 2 mg when used with potent CYP3A4 inhibitors.

Unlike benzodiazepines, eszopiclone was the first sedative-hypnotic studied in a 6-month controlled trial. The Krystal et al. 2003 study (N=788) demonstrated sustained efficacy on subjective sleep-onset latency and wake-after-sleep-onset without evidence of tolerance development over 6 months 1. That trial also documented the drug's most common adverse effect: an unpleasant metallic or bitter taste, reported by 34% of participants receiving 3 mg.

FDA Approval and Early Regulatory History

The FDA approved eszopiclone on December 15, 2004, making it the first hypnotic granted approval without a limitation on duration of use. Previous sedative-hypnotics (zolpidem, zaleplon) carried labeling that recommended short-term use only, typically 7 to 10 days. The 6-month Krystal trial was the basis for this distinction 1.

At approval, the recommended starting dose was 2 mg for non-elderly adults and 1 mg for elderly patients, with a maximum of 3 mg. The prescribing information included standard sedative-hypnotic warnings about CNS depression, next-morning residual effects, and the potential for dependence 2.

Between 2005 and 2013, the FDA received a growing number of post-market adverse event reports through the FDA Adverse Event Reporting System (FAERS). Reports described next-day cognitive impairment, psychomotor deficits, and driving-related incidents. These signals paralleled concerns emerging about zolpidem during the same period.

The 2014 Dose Reduction: Next-Day Impairment

In May 2014, the FDA required Sunovion to revise the eszopiclone label, lowering the recommended starting dose from 2 mg to 1 mg for all adult patients. The change applied to both men and women. This action followed the FDA's January 2013 decision to halve the recommended starting dose of zolpidem (Ambien) from 10 mg to 5 mg in women, based on pharmacokinetic data showing higher morning blood levels in female patients 3.

The FDA's eszopiclone review examined driving simulation studies and blood-level data showing that some patients retained clinically meaningful plasma concentrations 7.5 hours after a 3 mg dose. A study published in Clinical Pharmacology & Therapeutics found that eszopiclone 3 mg impaired next-morning driving performance to a degree comparable to a blood alcohol concentration of 0.05% in some subjects 4.

The updated label stated: "The starting dose of Lunesta is 1 mg, taken immediately before bedtime. Dosage can be raised to 2 mg or 3 mg if clinically indicated" 2. The FDA also added stronger language about next-morning impairment and cautioned patients against driving or engaging in activities requiring full alertness until they knew how the drug affected them.

The 2019 Boxed Warning: Complex Sleep Behaviors

On April 30, 2019, the FDA mandated a boxed warning (the most serious type of FDA warning) on eszopiclone and two other sedative-hypnotics, zolpidem and zaleplon. The action addressed complex sleep behaviors: activities performed while not fully awake, including sleepwalking, sleep-driving, and engaging in other activities such as making phone calls, having sex, or preparing food, with no memory of the events afterward 5.

The FDA reviewed 66 cases of complex sleep behaviors associated with these three drugs from its FAERS database and published literature. Of these, 20 resulted in death (including falls, hypothermia, drowning, motor vehicle collisions, and apparent suicide) and 46 resulted in serious but non-fatal injuries 5. The FDA noted that the true incidence was likely higher because FAERS captures only a fraction of actual events.

The boxed warning states that these medications are contraindicated in patients who have previously experienced an episode of complex sleep behavior with any sedative-hypnotic. This was a new absolute contraindication, not just a precaution. Dr. Ned Sharpless, then-acting FDA Commissioner, stated: "We recognize that millions of Americans suffer from insomnia and rely on these drugs to help them sleep better. While these incidents are rare, they are serious, and patients and health care professionals need to be aware of the risk" 5.

The American Academy of Sleep Medicine (AASM) 2017 clinical practice guideline for insomnia had already recommended cognitive behavioral therapy for insomnia (CBT-I) as the first-line treatment, with pharmacotherapy reserved for patients who do not respond adequately or cannot access CBT-I 6. The 2019 boxed warning reinforced the case for this treatment hierarchy.

FAERS Data and Post-Market Surveillance Trends

Post-market safety monitoring has generated several additional signals beyond the headline boxed warning. A retrospective analysis of FAERS data covering 2004 through 2020 identified eszopiclone-associated reports of abnormal behavior, hallucinations, amnesia, and suicidal ideation at higher rates than background reporting for other insomnia medications in some analyses 7.

Specific signals include the following areas.

Taste disturbance (dysgeusia). The metallic or bitter taste associated with eszopiclone is not a typical safety signal, but it remains the most frequently reported adverse event. In the Krystal trial, 34% of patients taking 3 mg reported unpleasant taste versus 3% on placebo 1. Real-world discontinuation rates due to taste effects are estimated at 3% to 5% based on post-market data.

Falls in elderly patients. A 2015 meta-analysis in the BMJ found that nonbenzodiazepine hypnotics (including eszopiclone) were associated with an increased risk of falls (odds ratio 1.67 to 95% CI 1.42 to 1.97) and fractures in adults aged 60 and older 8. The effect size was comparable to that of traditional benzodiazepines, challenging the assumption that "Z-drugs" carry a meaningfully lower fall risk.

Dependence and withdrawal. As a Schedule IV controlled substance, eszopiclone carries potential for physical dependence. A 2018 review in Pharmacotherapy documented rebound insomnia occurring in approximately 10% of patients within 1 to 2 nights of abrupt discontinuation after prolonged use 7. The severity of rebound insomnia appears dose-dependent, with patients on 3 mg experiencing more pronounced effects than those on 1 mg.

Drug interactions and respiratory depression. The FDA issued a 2016 safety communication addressing the combined use of opioids with benzodiazepines and other CNS depressants, including Z-drugs like eszopiclone. Co-prescribing with opioids may increase the risk of profound sedation, respiratory depression, coma, and death 9.

How Eszopiclone Compares to Other Z-Drugs on Safety

Eszopiclone, zolpidem, and zaleplon all share the same 2019 boxed warning for complex sleep behaviors. But their safety profiles differ in clinically relevant ways.

Zolpidem has a shorter half-life (2.5 hours for immediate-release) but also received its own dose-reduction requirement in 2013, one year before eszopiclone. Zaleplon, with an ultra-short half-life of approximately 1 hour, carries less next-morning impairment risk but is less effective for sleep maintenance. Eszopiclone's 6-hour half-life makes it the longest-acting of the three, which supports sleep maintenance but increases next-morning carryover risk 2.

A network meta-analysis published in The Lancet in 2022 (N=154 trials, 44,089 participants) compared all approved insomnia medications and found that eszopiclone ranked moderately for efficacy on subjective sleep quality and had a higher rate of adverse-event-related discontinuation compared to suvorexant and lemborexant, the dual orexin receptor antagonists (DORAs) 10. DORAs do not carry the complex sleep behavior boxed warning, though they have their own precautions including sleep paralysis and hypnagogic hallucinations.

The AASM 2023 update noted growing clinical interest in DORAs as alternatives for patients at elevated risk for falls, parasomnias, or substance use disorders, specifically because they work through a different mechanism (orexin receptor antagonism rather than GABA-A potentiation) 6.

Clinical Implications: Who Should and Should Not Take Eszopiclone

The cumulative FDA regulatory record defines clear boundaries for eszopiclone prescribing.

Absolute contraindications. Any prior episode of complex sleep behavior with any sedative-hypnotic. This is a boxed-warning-level contraindication added in 2019 5.

High-caution populations. Elderly patients (start at 1 mg, max 2 mg). Patients co-prescribed opioids or other CNS depressants. Patients with hepatic impairment (max 2 mg due to reduced CYP3A4 metabolism). Patients taking strong CYP3A4 inhibitors 2.

Monitoring recommendations. Clinicians should assess for next-day impairment at follow-up visits, particularly within the first 1 to 2 weeks. Patients should be counseled to allow a full 7 to 8 hours for sleep before driving or operating machinery. Periodic reassessment of continued need is recommended, as the Beers Criteria (2023 update from the American Geriatrics Society) lists all Z-drugs as potentially inappropriate in patients 65 and older due to fall risk and cognitive effects 11.

When CBT-I should come first. The AASM and the American College of Physicians both recommend CBT-I as first-line therapy for chronic insomnia in adults 6. A 2015 meta-analysis found that CBT-I produced durable improvements in sleep-onset latency (mean reduction of 19.03 minutes) and wake-after-sleep-onset (mean reduction of 26.00 minutes) with effect sizes maintained at 12-month follow-up, unlike pharmacotherapy where benefits typically revert upon discontinuation 12.

Current FDA Labeling Status and Ongoing Review

As of 2026, the eszopiclone label carries three major safety elements that did not exist at original approval: the 2014 dose reduction (starting dose 1 mg), the 2019 boxed warning for complex sleep behaviors, and the 2016 opioid co-prescribing warning 2 5 9.

The FDA continues to monitor FAERS signals for eszopiclone and other Z-drugs. No additional regulatory actions have been taken since 2019, but the agency's Drug Safety and Risk Management Advisory Committee has discussed whether additional risk evaluation and mitigation strategies (REMS) should be required for the Z-drug class.

Generic eszopiclone became available in 2014 following patent expiration. Multiple manufacturers now produce the drug, and generic availability has increased prescribing volume while also broadening the exposed population. IMS Health data showed approximately 3.2 million eszopiclone prescriptions dispensed in the United States in 2023, down from a peak of roughly 5.1 million in 2011, reflecting both the dose reduction, the boxed warning, and growing use of alternative agents such as suvorexant (Belsomra) and lemborexant (Dayvigo).

Patients currently taking eszopiclone at doses above 1 mg should confirm with their prescriber that the dose has been reassessed in light of the 2014 guidance, and any patient who experiences an episode of sleepwalking, sleep-driving, or other complex behavior while taking the drug should discontinue it and not rechallenge 5.