By Sumathi Reddy

Debra Davis has had trouble sleeping for more than five years. Once she started taking a prescription sleep medication to help with her insomnia, it was hard to stop.

“I was taking it daily for about a year,” says Ms. Davis, a 58-year-old nurse practitioner in Denver.

Sleep doctors say this is a common problem: Medications intended as short-term fixes become a regular habit. And when patients stop taking the medications, anxiety takes over.

Studies show a majority of patients take sleep medications for at least a year, and 30% for at least five years.

“A good percentage of people have made at least one attempt to get off of it and haven’t succeeded,” says Jack Edinger, a psychology professor who specializes in insomnia at National Jewish Health in Denver.

With an eye on reducing that problem, Dr. Edinger is studying whether patients trying to get off sleep medications do better if they don’t know what they’re taking. In a pilot study, he found that patients who taper their medications in a blinded way—taking their normal dose of medication some nights, and less or no medication on others—do better at staying off their medications in the long term. In July he got federal funding to do a five-year, $2.5 million study in a larger group of 260 patients.

For the pilot study, 78 participants got six weeks of cognitive behavioral therapy for insomnia (CBTI) while on their regular dose of medication.

The subjects were then split into three groups. One group of subjects had their medications reduced 25% every two weeks, and a second had theirs reduced by 10% every two weeks. They couldn’t see the dosage of their medication as it tapered over 20 weeks. Members of a control group and took their medication at the normal dosage.

At the end of the 20 weeks, the control group subjects were offered the chance to taper the dosage of their sleep medication at a rate that they could see.

All subjects saw similar results after their tapering: The majority successfully got off their medication. But three months later, 73% of the participants in the blinded groups remained off their drugs, while only 35% of those in the control group did.

The preliminary results were presented at the annual SLEEP conference in San Antonio in June. The results didn’t include 17 patients who were in various stages of the study, but they will be included in the final results, which Dr. Edinger hopes to have published next year. He says the results were statistically significant even though the sample size was small.

“When patients knowingly reduce or eliminate their drugs on a particular night, their anxiety level goes up and they start wondering how they’re going to do,” Dr. Edinger says. “That anxiety is really what’s getting in the way. It’s not the fact that they don’t have a drug in their system.”

There hasn’t been much other research on reducing sleep medication without patients’ knowledge. Researchers in a 2012 study examined patients taking Ambien for a year who switched to placebo for a week at three separate points during the year without knowing it. They experienced no rebound insomnia.

Daniel Buysse, a psychiatry professor at the University of Pittsburgh School of Medicine, says Dr. Edinger’s recent experiment got around the anxiety patients feel when trying to wean themselves from sleep medications. Still, he says the small number of patients in the study so far makes it difficult to draw conclusions. And he wonders how such a practice could be implemented clinically given ethical questions about giving patients placebo pills.

“It’s an interesting notion that you may be able to eliminate patients’ negative expectations and fears,” says Dr. Buysse, who wasn’t involved in the study but has worked with Dr. Edinger on other studies. “The problem though is that typically, we are reluctant in clinical medicine to prescribe placebos or to involve any type of what might be perceived as deception.”

Along with the ethical issues of giving patients placebos, there are logistical challenges of giving patients varying doses of pills in a week, or a placebo.

Dr. Edinger’s logistical solution: Patients can buy a seven-day pill box and empty capsules. They can then have a friend or family member distribute different doses of their medication in the capsules so they don’t know what the dose is on different days and keep a sleep diary to see if there’s any difference.

David Neubauer, an associate psychiatry professor at Johns Hopkins University School of Medicine, called the experiment clever.

He also noted the positive effect the CBTI sessions may have had, as well as meeting with a doctor regularly. “This was a pretty labor-intensive process,” Dr. Neubauer says. “It’s much more than simply, ‘Here’s a blinded taper.’ I think this whole approach really highlights the fact that there are psychological and behavioral dimensions for sleep for every single one of us.”

Ms. Davis, who was in the blinded tapering group in Dr. Edinger’s study, says the sleep behavior and techniques she learned as part of the CBTI helped her immensely.

“It helped my anxiety around sleep,” she says. But some of her bad habits, like watching TV in bed, have started creeping back. She says she sometimes takes sleep medication once or twice a month.