EVIDENCE-BASED ANSWER
Short-acting hypnotics such as zolpidem (Ambien) or zaleplon (Sonata) are the preferred hypnotics in the elderly because of an improved side-effect profile compared with traditional hypnotics such as benzodiazepines (strength of recommendation: B, based on extrapolations of randomized controlled trials). Zolpidem and zaleplon have a quick onset and short duration of action, making them less likely to cause residual sedation, cognitive changes, and falls than benzodiazepines. More comparative clinical trials in the elderly are needed to determine if zolpidem and zaleplon are truly safer than benzodiazepines in this population. Hypnotics should be prescribed on a short-term, intermittent basis as part of a comprehensive treatment plan that addresses any underlying causes of poor sleep.
Evidence summary
Zolpidem and zaleplon
Zolpidem and zaleplon differ structurally from benzodiazepines but act at the benzodiazepine receptor.1 Due to their rapid absorption and short half-lives, they are particularly helpful for patients who have trouble falling asleep.2 They have been shown to decrease sleep latency, increase total sleep time, and increase sleep efficiency without disturbing sleep architecture or adversely affecting memory.1
Comparative studies in the elderly have demonstrated that zolpidem is as effective as triazolam,3 and that zaleplon is more effective than placebo at decreasing sleep latency and improving sleep quality.4 Tolerance, withdrawal symptoms, or rebound insomnia occur less frequently than with benzodi-azepines,1 but zolpidem increased risk of hip fracture in a case control study (adjusted odds ratio=1.95, 95% confidence interval, 1.09–3.51).5
Side effects of zolpidem and zaleplon are considered dose-related, and a lower dose of 5 mg is recommended for older patients.2 Efficacy of intermittent use of zolpidem has been demonstrated in clinical studies,1 a practice that could potentially decrease risk of side effects. Overall, if a hypnotic is desired for an older adult, zolpidem and zaleplon are preferred because of their improved side-effect profiles compared with older hypnotics such as benzodiazepines, chloral hydrate, over-the-counter sleep aids, and antidepressants (see Table ).
TABLE 1
Adverse effects of hypnotics in the elderly
| Hypnotic | Adverse effect |
|---|---|
| Benzodiazepines | Somnolence, anterograde amnesia, falls, hip fracture, rebound insomnia, tolerance, dependence, impaired sleep architecture2,3,5 |
| Antihistamines | Somnolence, dry mouth, constipation, urinary retention, blurred vision, cognitive changes3 |
| Valerian | Headache, excitability, uneasiness, cardiac disturbances, insomnia, drowsiness, withdrawal symptoms10 |
| Melatonin | Headache, depressive symptoms, daytime fatigue and drowsiness, dizziness, abdominal cramps, reduced alertness10 |
| Chloral hydrate | Nausea, vomiting, diarrhea, may increase effects of warfarin, overdose potential3,8 |
| Tricyclic antidepressants | Dry mouth, constipation, urinary retention, blurred vision, cognitive changes, orthostatic hypotension, somnolence, worsening of chronic heart failure, overdose potential, cardiac conduction abnormalities2,3 |
| Trazodone | Somnolence, orthostatic hypotension, dry mouth, priapism3 |
| Zolpidem | Drowsiness, headache, dizziness, somnolence, fatigue, agitation, nightmares, diarrhea, myalgia, arthralgia, anterograde amnesia1,10 |
| Zaleplon | Headache, dizziness, somnolence, short-term amnesic effects, next-day memory impairment, mild rebound insomnia1,10 |
Benzodiazepines
Benzodiazepines have been used since the 1960s for their hypnotic, anxiolytic, anticonvulsant, muscle-relaxing, and amnesic properties. A recent meta-analysis showed that benzodiazepines improve sleep latency by only 4.2 minutes compared with placebo.6 Although benzodiazepines increase sleep time and efficiency, patients quickly develop tolerance to the hypnotic effects.7 Additional problems associated with benzodiazepines include dependence, rebound insomnia, residual sedation, falls, hip fractures, and detrimental effects on sleep architecture.7