To Name :
To Email :
From Name :
From Email :
Comments :

Evidence-Based Reviews

Hypnotics and driving: FDA action, clinical trials show need for precautions

Drugs’ effects on performance and memory differ, depending on time since administration.

Vol. 6, No. 4 / April 2007

“Sleep driving” blamed on the hypnotic zolpidem was used as a defense last year in Virginia in a criminal case involving impaired driving. The defendant’s attorney argued that the defendant should not be held criminally liable because he was “essentially unconscious” and the accident therefore was involuntary.

The “sleep driving” defense failed when testimony revealed the defendant had taken 5 times the recommended zolpidem dose before the accident. The judge found him guilty of a felony charge of driving under the influence of a sleep medication.1

Sedative-hypnotics are increasingly being used to treat insomnia2-4 and as a result some patients try to drive while under the drugs’ sedating effects. Also, new FDA-ordered labeling for all 13 available prescription sleep aids warns of potential risks of “complex sleep-related behaviors,” including driving, phoning, and eating while asleep (Box 1).

Hypnotics can improve quality of life and well-being by addressing insomnia’s complications—hypertension, diabetes, coronary artery disease, depression, and anxiety5-7—but they also have been associated with impaired motor coordination and somnambulism. To help you and your patients weigh sleep medications’ relative risks and benefits, we report on clinical studies and court cases in the literature. Most of the data focus on zolpidem, by far the most prescribed hypnotic (Box 2).8,9

Box 1

New labeling for hypnotics: Sleep-related behaviors

Labeling of all sedative-hypnotic drugs now carries FDA-ordered precautions about “sleep-driving and other complex behaviors” that may occur without the patient being fully awake. FDA cited reports of patients preparing and eating food, making phone calls, and having sex after taking a sedative-hypnotic, usually without memory of the event. A warning also was added about rare, potentially fatal anaphylactic reactions in patients taking first or later doses of sleep medications.

Steven Galson, MD, MPH, director of FDA’s Center for Drug Evaluation and Research, said the labeling changes were needed to inform patients and prescribers about the risks of sleep aids that “are well-tolerated and effective for many people.”

Source: Walsh S, Rawlings K. FDA requests label change for all sleep disorder drug products. Available at

Zolpidem incidents and cases

In 2005, Americans filled 43 million prescriptions for sedative-hypnotics—26.5 million for zolpidem alone—compared with 29 million prescriptions in 2001.4 In addition to the new the FDA-requested warnings about sleep-related behaviors, zolpidem’s labeling cautions patients about operating heavy machinery, driving, or engaging in hazardous occupations after taking the drug. The manufacturer tells patients:

  • to ingest zolpidem only before going to bed
  • that they may experience residual sedation the following day.

Not all patients heed the precautions or follow dosing recommendations, however.

Impaired driving. Besides the “sleep driving” case in Virginia, a highly publicized zolpidem-related driving incident occurred May 4, 2006, when U.S. Representative Patrick Kennedy was involved in an accident after having taken zolpidem in combination with an antinausea medication. Another driving-related case has used zolpidem as a defense for impairment, but the court decided that the medication was not at fault because the defendant also had ingested alcohol.10

Other litigation. Although zolpidem-related impairment apparently has not been used successfully as a defense for a driving incident, class action suits alleging failure to disclose potentially harmful side effects have been filed against the manufacturer.

In Janet Makinen and others v. sanofi-aventis,11 at least 500 plaintiffs claim zolpidem is related to sleep-driving, sleep-eating, and other somnambulistic behaviors. Plaintiffs allege negligence, breach of implied warranties, fraud, unfair trade practices, express warranty violations, strict liability, and consumer fraud violations. Other suits claim dangerous sleep-related side effects with zolpidem use.12

What clinical evidence shows

Driving impairment. Clinical studies have shown conflicting results about driving impairment associated with zolpidem. The literature falls into 2 categories, based on treatment duration:

  • Zolpidem affects performance and memory within the first 4 to 5 hours of administration (Table 1).
  • Beyond 5 hours, no residual effects on performance have been identified (Table 2), and repeat nightly dosing has not caused impairment or tolerance.

Verster et al13 examined residual effects of benzodiazepines and the nonbenzodiazepines zolpidem, zopiclone (available in the United States as eszopiclone), and zaleplon on driving ability, as reported in studies of on-the-road driving, driving simulators, epidemiologic data, and closed-road driving. This review found that:

  • All sedative hypnotic benzodiazepines had statistically significant residual effects 10 to 11 hours after ingestion, with longer periods of impairment corresponding to medications with longer half-lives.
  • Zopiclone was associated with significant residual impairment for up to 10 hours after ingestion.
  • Zolpidem and zaleplon showed no significant impairment in driving 10 to 11 hours after ingestion. Impairment was found, however, when zolpidem was taken within 5 hours of driving.14-18

Table 1

Studies of zolpidem-associated driving skills impairment
(<5 hours after dosing)


Doses and timing

Driving skills assessments


Wilkinson, 199514 Blinded; 29 subjects

Zolpidem, 10 mg, 15 mg, and placebo in combination with an alcoholic drink (to reach a BAC of 0.08%) or placebo drink; testing 45 min, 130 min, and 230 min after administration

Visual backward masking test (approximates driving performance) and attention tests

Zolpidem produced significant impairment in combination with alcohol and when administered alone during peak effect assessment; alcohol did not potentiate zolpidem’s effects; additive effects of alcohol seen with 10-mg dose but not 15-mg dose of zolpidem

Rush et al, 199815 Double-blind, crossover; 9 subjects

Zolpidem, 7.5, 15, and 22.5 mg; quazepam, 15, 30, and 45 mg; triazolam, 0.1875, 0.375, and 0.5625 mg; testing ½, 1, 1½, 2, 2½, 3, 4, 5, and 6 hours after administration

Subject- and observer-rated questionnaires; tests of recall and delayed recognition

Performance-impairing effects of zolpidem were virtually indistinguishable from those of classic benzodiazepines, such as triazolam

Mattila et al, 199816 Randomized, placebo-controlled, double-blind, crossover; 12 subjects

Zolpidem, 15 mg; diazepam, 15 mg; oxazepam, 30 mg; zopiclone, 7.5 mg; alcohol testing before and 1, 3½, and 5 hours after administration

Simulated driving and other measures

Zolpidem impaired coordination, reaction, and cognition at 1 and 3½ hours; tracking remained impaired at 5 hours; all agents (especially zolpidem) impaired learning and memory

Mintzer et al, 199917 Double-blind, placebo-controlled; 16 subjects

Zolpidem, 15 mg/70 kg (dosed by subject weight); testing ½, 1, 2, and 3 hours after administration

Memory tasks (recall, fragment completion, recognition)

Zolpidem interfered with explicit but not implicit memory after administration; zolpidem produced a specific deficit in acquisition of contextual information

Verster et al, 200418 2-step randomized, placebo-controlled, double-blind, crossover; 30 subjects

Zolpidem, 10 mg and 20 mg; zaleplon, 10 mg and 20 mg; middle-of-the-night dosing; testing 4 hours after dosing

On-the-road driving and other tests of attention, learning, and thinking

Zolpidem, 10 mg and 20 mg, significantly impaired driving function; zolpidem, 20 mg, produced significant impairment on all psychomotor and memory tests; zaleplon, 10 mg and 20 mg, did not differ significantly from placebo

BAC: Blood alcohol concentration

Table 2

Studies of zolpidem-associated driving skills impairment
(>5 hours after dosing)


Doses and timing

Driving skills assessments


Fairweather et al, 199223 Randomized, placebo-controlled; 24 older volunteers taking no other medications

Zolpidem, 5 mg or 10 mg, or placebo taken before bedtime; testing 8.5 hours after administration

Numerous, including reactive time, memory, word recognition

Zolpidem consistently helped with sleep latency, with no residual performance deficits; no tolerance seen with repeated dosing

Bocca et al, 199924 Double-blind, crossover; 16 volunteers

Zolpidem, 10 mg; zopiclone, 7.5 mg; flunitrazepam,* 1 mg; and placebo given at 11 PM, with testing at 9 AM

Driving simulation and real time test drive; eye movements measured after driving tests

No residual effects with zolpidem; zopiclone impaired driving ability and increased saccadic latency; flunitrazepam impaired early morning driving and saccadic eye movements longer than zopiclone

Partinen et al, 200320 Randomized, placebo-controlled, double-blind, 3-period crossover; 18 women with insomnia

Zolpidem, 10 mg; temazepam, 20 mg; dosing at 2 AM, testing 5.5 hours after dosing

Driving simulation; delayed word recall and memory testing (FePsy test)

No statistically significant effects on driving ability with either drug; no significant differences in FePsy results compared with baseline or placebo

Staner et al, 200525 Randomized, placebo-controlled, double-blind, four-way crossover; 23 subjects with DSM-IV-TR diagnosis of insomnia

Zolpidem, 10 mg; zopiclone, 7.5 mg; lormetazepam,* 1 mg; 7 days of dosing; tests given 9 to 11 hours post-dosing

Driving simulation; EEG at rest and while driving

Zolpidem showed no impairment of driving ability and no EEG changes compared with placebo; driving impairment and EEG alterations were found with zopiclone and lormetazepam

* Hypnotics not approved in the United States but available elsewhere.

Acute effects (<5 hours)

Combined with alcohol. Wilkinson14 conducted a randomized, 6-way crossover study in which subjects received 10- or 15-mg doses of zolpidem or placebo plus an alcoholic beverage (enough to obtain a blood alcohol concentration [BAC] of ~0.08%) or placebo beverage. Tests given shortly after patients took the study medications showed that zolpidem caused statistically significant impairment both in combination with alcohol and alone during peak drug effect—identified as 45 minutes after ingestion. Alcohol did not potentiate the impairment associated with zolpidem.

Using a similar design, Mattila et al16 compared acute performance impairment associated with zolpidem, diazepam, oxazepam, and zopiclone. In this randomized, double-blinded, crossover study, all comparison medications impaired antecedent learning and memory, but zolpidem given at 15 mg had the greatest effect. Zolpidem impaired coordination, reactive functioning, and cognitive skills at 1 and 3.5 hours after administration, and simulated driving test performance remained impaired at 5 hours (approximately two half-lives of the medication). Of note is that the 15-mg zolpidem dose used in this study was shown by Wilkinson et al14 to be more impairing than the recommended maximum 10-mg dose.

A study from the University of Toronto19 that did not include zolpidem examined potential psychomotor performance deficits and sleepiness in a comparison of time-released melatonin, 6 mg; zaleplon, 10 mg; zopiclone, 7.5 mg; temazepam, 15 mg, and placebo. Tests were given to 9 men and 14 women, ages 21 to 53, just before drug administration and 7 hours later.

Zaleplon had the greatest effect on psychomotor performance, followed by temazepam and zopiclone. Aside from prolonged perceived sleepiness, melatonin and placebo did not interfere with performance testing.

Box 2

Zolpidem: Approved for ‘short-term’ insomnia

Zolpidem, a benzodiazepine receptor agonist, was the 7th most prescribed drug in the United States in 2005 (2006 data not available).8 It is FDA-approved for short-term treatment of insomnia, although “short-term” is not defined. Package labeling states:

This nonbenzodiazepine hypnotic has been shown to decrease sleep latency and increase sleep duration for up to 35 days in controlled clinical trials. Patients should be evaluated for a primary psychiatric or medical illness if insomnia does not remit after 7 to 10 days of treatment.

An imidazopyridine that acts as an agonist of GABA A1, zolpidem produces sedation while avoiding anticonvulsant, anxiolytic, and muscle relaxation effects. Available in 5- and 10-mg tablets, the drug is rapidly absorbed in the GI tract and excreted primarily through the kidneys. Its half-life is approximately 2.5 hours (approximately 3 hours in elderly patients). The most common side effects are daytime drowsiness, dizziness, and diarrhea; others include asthenia, hiccup, and diplopia.9

Middle-of-the-night dosing. Effects of zolpidem and zaleplon on driving ability, memory, and psychomotor performance were compared by Verster et al18 in a randomized, controlled trial. The double-blind, 5-period crossover design measured the effects of middle-of-the-night use of zaleplon, 10 or 20 mg; zolpidem, 10 or 20 mg; or placebo on:

  • driving ability 4 hours after administration
  • memory and psychomotor performance 6 hours after administration.

As expected, subjects taking zolpidem showed impairment on all measures. The 10- and 20-mg doses significantly impaired driving 4 hours after ingestion, with the 20-mg dose—twice the recommended maximum dose—producing greater impairment. The 20-mg dose—but not the 10-mg dose—also significantly impaired memory and psychomotor function. Zaleplon did not impair driving ability, memory, or psychomotor testing.

Did you miss this content?
Psychosis and catatonia after dancing with a dangerous partner