For millions of people who take benzodiazepine receptor agonist (BZRA) sleep medications, the prospect of stopping them is daunting. These drugs, which include common benzodiazepines and Z-drugs, can lead to dependency, cognitive impairment, and falls, especially with long-term use. Newer classes of sleep medications, such as melatonin receptor agonists and orexin receptor antagonists, are safer alternatives. But the central clinical question has remained: once a patient is already on a BZRA, does adding or switching to a novel hypnotic actually help them reduce or stop the older drug?
A new study published in Sleep Medicine provides a clear, if sobering, answer. It works, but only if you act early. The longer a patient has been on BZRAs, the harder it becomes to reduce them.
The study, led by Dr. Nobutaka Ayani of Kyoto Prefectural University of Medicine and colleagues at two hospitals in Japan, tracked 190 patients who were already taking BZRA hypnotics and then began a novel hypnotic. The researchers followed them for one year to see whether they were able to reduce their BZRA use by at least one agent.
What They Found
Overall, 46 percent of patients successfully reduced their BZRA use within one year after starting a novel hypnotic. But that overall number hides a dramatic split based on how long patients had been taking BZRAs before the switch.
For patients who had used BZRAs for less than one year, the success rate was 69 percent. Nearly seven out of ten patients in this group were able to reduce their older medication. For patients with one year or more of prior BZRA use, the success rate fell to 36 percent. The difference was highly statistically significant, with a p-value below 0.0001.
The researchers then adjusted for other factors that could influence the outcome, such as age, sex, and the specific type of hypnotic prescribed. The adjusted odds ratios paint an even starker picture. Compared with patients who had no regular prior BZRA use, those with 1 to 5 years of prior use had an odds ratio of 0.21 (95 percent confidence interval: 0.08 to 0.56) for successful reduction. Those with 5 or more years of prior use had an odds ratio of 0.16 (95 percent CI: 0.06 to 0.43). In plain language, a patient who has been on BZRAs for 1 to 5 years is roughly five times less likely to successfully reduce their use than someone who has not been a regular user. At 5 years or more, the odds drop to about six times less likely.
The study drew from a larger pool of 3,802 patients who were prescribed a total of 5,114 hypnotics across the two hospitals between May 2021 and April 2022. Of those prescriptions, 38 percent were for novel hypnotics, meaning the newer drug classes are already a significant share of clinical practice. The 190 patients who met the inclusion criteria for the analysis had all been on BZRAs and then initiated a novel hypnotic during the study window.
The novel hypnotics in question included melatonin receptor agonists, which help regulate the sleep-wake cycle through the body’s natural circadian pathways, and orexin receptor antagonists, which promote sleep by blocking the brain’s wake-promoting signals. Both classes are considered safer than BZRAs, with lower risks of dependence, tolerance, and next-day impairment.
Why It Matters
Insomnia is one of the most common complaints in primary care and sleep medicine, and BZRAs have been a mainstay of treatment for decades. But evidence for their long-term use has been steadily eroding. Chronic BZRA use is associated with memory problems, daytime sedation, increased risk of falls and fractures in older adults, and physical dependence with withdrawal symptoms upon discontinuation.
Clinical guidelines from organizations such as the American Academy of Sleep Medicine and the British Association for Psychopharmacology have increasingly recommended limiting BZRA use to short courses, often no more than four weeks. They advise switching to safer alternatives or using non-pharmacological treatments like cognitive behavioral therapy for insomnia.
The problem is that in practice, many patients end up staying on BZRAs for months or years. Once established, the habit can be extremely difficult to break. The new study directly addresses this real-world challenge by showing that introducing a novel hypnotic can serve as a bridge to BZRA reduction, but only within a relatively narrow window of opportunity.
For clinicians, the message is straightforward: if a patient is on a BZRA, consider introducing a novel hypnotic early in the course of treatment. Waiting until the patient has been on BZRAs for a year or more means the odds of successfully reducing the older drug drop dramatically. For patients who have been on BZRAs for five years or longer, the likelihood of success with a simple medication switch is low, and more intensive strategies such as gradual tapering combined with behavioral therapy may be needed.
These findings also have implications for the broader conversation about deprescribing in sleep medicine. The concept of deprescribing, systematically reducing or stopping medications that are no longer appropriate has gained traction in geriatrics and psychiatry. This study adds specific, actionable data to that discussion for one of the most widely prescribed classes of medications in the world.
Limits
The study has several important limitations. It was retrospective in design, meaning the researchers analyzed existing medical records rather than randomly assigning patients to different treatments. This introduces the possibility of selection bias, patients who were offered novel hypnotics may have differed from those who were not in ways that affected their outcomes. The sample size was relatively small, with only 190 patients meeting the inclusion criteria, which limits the precision of the subgroup analyses. The study was conducted at just two hospitals in Japan, so the findings may not generalize to other countries or healthcare systems where prescribing practices, patient demographics, and insurance coverage differ. Additionally, the primary outcome was reduction by at least one BZRA agent, not complete discontinuation, so the clinical significance of a partial reduction may vary from patient to patient.
Bottom Line
Introducing a novel hypnotic such as a melatonin receptor agonist or orexin receptor antagonist can help patients reduce their use of older BZRA sleep medications, but the window for successful switching is narrow. Patients who have been on BZRAs for less than one year have a roughly 69 percent chance of reducing their use, while those on BZRAs for a year or more have only a 36 percent chance. The findings reinforce the importance of avoiding long-term BZRA therapy whenever possible and of acting early to introduce safer alternatives when a switch is clinically indicated.
Source
Ayani N, Matsumoto Y, Kurokawa T, et al. Long-term effects of introducing novel hypnotics to users of benzodiazepine receptor agonist hypnotics. Sleep Med. 2026;147:109148. doi:10.1016/j.sleep.2026.109148. PMID: 42470902.