Insomnia and other sleep problems significantly affect the quality of life of many older patients. Up to one third of older persons have been prescribed a medication for sleep problems, but the common agents are associated with falls, ataxia, memory problems, and other adverse effects. Glass and colleagues used a meta-analysis to study the balance of benefits and risks of sedative hypnotic use in older patients.

The authors searched multiple electronic databases for relevant studies, reviewed the bibliographies of articles and reviews, and requested data from unpublished studies to identify all pertinent information. For inclusion in the meta-analysis, studies had to involve patients with an average age of at least 60 years and use standard definitions for insomnia. Studies of barbiturates and chloral hydrate were excluded because these drugs are not recommended in older patients. The outcome variables considered were reported sleep quality, latency (ease of getting to sleep), number of awakenings, and total sleep time.

Only 20 of the 120 studies initially identified met criteria for inclusion in the meta-analysis. Benzodiazepines were the most common drugs studied (830 patients), followed by zaleplon (Sonata; 609 patients), zolpidem (Ambien; 384 patients), zopiclone (Imovane, not available in the United States; 106 patients), and diphenhydramine (Benadryl; 14 patients). Placebos were taken by 468 patients.

Overall, sedatives were associated with a statistically significant improvement in sleep quality. Using a seven-point scale, sedative and placebo groups had a difference of only 0.11 for reported sleep quality, but this difference was equivalent to a number needed to treat of 13 for an improvement in sleep quality. The mean increase in sleep time with sedatives was 25 minutes (12.8 compared with 37.8 minutes). Average number of awakenings decreased with sedatives. The authors conducted a funnel plot analysis showing a possible publication bias in favor of the positive outcomes of sleep quality and total sleep time.

Based on pooled data from 16 studies with 2,220 patients, the number needed to harm was six. The most common adverse reports were fatigue, headache, nightmares, and nausea. Cognitive effects, daytime fatigue, and impairment of performance tasks were significantly more common with sedative use. Psychomotor problems were more common with sedative use, including seven serious falls and one motor vehicle collision, but the difference did not reach statistical significance.

The authors conclude that sedative hypnotics are associated with a small improvement in some aspects of sleep in older patients, but this therapy also carries significant risk of adverse events. They added that adverse events can have serious consequences for older patients that may outweigh any benefit from improved sleep.