Sleeping Pill Action Owes Much to Placebo Effect

Author: Daniel Weiss, Senior Editor

An analysis of published and unpublished trials of Z drugs finds that almost half their ability to help users fall asleep is due to the placebo effect.

Almost half of the assistance that users derive from the popular sleeping pills known as Z drugs is due to the placebo effect, according to the results of an analysis of previous trials published in the December 17, 2012, edition of BMJ. The drugs, technically known as non-benzodiazepine hypnotics, are the most commonly prescribed hypnotic agents worldwide.
 
The researchers drew on randomized double-blind placebo-controlled trials submitted to the FDA as part of the new drug approval process for eszopiclone, zaleplon, and zolpidem. The FDA dataset offered the advantage of including unpublished trials as well as published ones, which have been found to significantly overstate drug effectiveness. However, all trials included were industry sponsored, which has been shown to enhance the outcome of clinical trials.
 
Data for the analysis was drawn from 13 studies that included 65 separate drug-placebo comparisons and 4378 participants. The mean duration of the studies was 33.9 days, the mean age of participants was 49.6 years, and 61% of study participants were female. Zolpidem was assessed in 8 studies, eszopiclone in 3 studies, and zaleplon in 3 studies. (One study assessed both zolpidem and zaleplon.)
 
The primary outcomes for the analysis were polysomnographic sleep latency, or how long it takes to fall asleep as measured by laboratory equipment, and subjective sleep latency, or how long it takes to fall asleep according to participants’ perception. The results found that Z drugs produce significant, although relatively small, reductions in polysomnographic sleep latency (weighted standardized mean difference of -0.36) and subjective sleep latency (-0.33) compared with placebo. In all, the combined effect of drug and placebo effects were found to reduce the time it took to fall asleep by an average of 22 minutes as measured by laboratory equipment and an average of 7 minutes as measured subjectively.
 
The results were stronger with higher doses of drugs, longer duration of treatment, and trials including a higher portion of younger or female patients. In all, though, the non-specific placebo response (which includes the placebo effect, regression toward the mean, and improvement due to the natural course of the condition) accounted for almost half the effect of the response to the drugs. As a result, the researchers note, the remaining drug effect should be balanced against harms associated with the drugs. They also suggest that psychological approaches to treating insomnia should be considered, given that the placebo effect is psychological in nature.