Masked tapering plus augmented CBTI increased long‑term benzodiazepine receptor agonist discontinuation versus standard CBTI with unmasked taper, without worsening insomnia.

Background

Benzodiazepine receptor agonist (BZRA) hypnotics offer modest sleep benefits and increase fall risk in older adults. Guidelines recommend deprescribing BZRAs and using cognitive behavioral therapy for insomnia (CBTI) as first-line treatment. Placebo-related mechanisms (eg, expectancy) contribute meaningfully to perceived hypnotic effects. Whether masking the taper dose and adding exercises targeting placebo/nocebo mechanisms improves BZRA discontinuation was unknown.

Patients

• Adults aged ≥55 years using lorazepam, alprazolam, clonazepam, temazepam, and/or zolpidem ≥2 nights/week for ≥3 months for current or prior insomnia.
• N=188 randomized; mean age 69.8 years (SD 8.3); 65.4% male; 91.5% used BZRA ≥1 year.
• Baseline BZRA use: mean 5.9 nights/week; mean dose 3.9 mg diazepam-equivalent.
• Insomnia severity: 48.4% had moderate-to-severe insomnia.
• Key exclusions: high risk for withdrawal complications, poor CBTI candidacy, discontinuation inappropriate, unstable housing.

Intervention

Masked Taper plus CBTI-Augmented Program (MTcap)

• Eight 60-minute individual CBTI sessions over 9 weeks (+ up to two 30-minute withdrawal check-ins).
• Pharmacist-compounded, encapsulated nightly doses with progressive inert filler; 7-day blister packs masked actual dose and included placebo-only capsules as taper progressed.
• Augmented behavioral content targeting expectancy/placebo and nocebo mechanisms: education on placebo effects, adverse effect risk visualization, experiential exercises, pros/cons of hypnotic use, relaxation and cognitive techniques (eg, diaphragmatic breathing, thought records, worry journal), and relapse prevention.
• Participants guessed nightly dose and tracked daytime symptoms; taper unmasked with feedback in session 8.
• Taper pace: approximately 25% dose reduction per week across 9 weeks.

Control

Standard CBTI plus Supervised (Unmasked) Gradual Taper (SGT)

• Same CBTI structure/content and check-ins.
• Paper taper schedule (same 25%/week over 9 weeks) with pill cutter; no masking or augmented expectancy-targeting exercises.

Outcome

• Primary: Percentage achieving BZRA discontinuation 6 months after treatment (7-day self-reported medication logs; subset urine confirmation).
• Secondary:
  • Insomnia Severity Index (ISI) at 1 week post-treatment and 6 months.
  • Percentage discontinued at 1 week post-treatment.
  • BZRA frequency (nights/week) and dose (diazepam-equivalent mg) at 1 week and 6 months.
  • Dysfunctional Beliefs About Sleep–Medication subscale (DBAS-16 MED) at 1 week and 6 months.

Study Design

• Phase 2, randomized, parallel-group clinical trial at an academic medical center and a Department of Veterans Affairs medical center (California).
• Enrollment: December 2018–November 2023; analyses: November 2023–July 2024.
• Randomization stratified by site, gender, and drug half-life; assessors blinded; intention-to-treat analyses.
• Trial registration: ClinicalTrials.gov NCT03687086.

Level of Evidence

Level I (individual randomized controlled trial).

Follow up period

• Primary endpoint at 6 months after treatment ended.
• Additional assessment at 1 week post-treatment.

Results

Primary outcome: BZRA discontinuation at 6 months

• MTcap: 73.4% (64/87 assessed) vs SGT: 58.6% (52/89 assessed); absolute difference 14.8%.
• Effect size: OR 1.95 (95% CI 1.03–3.70).
• NNT = 7 to achieve one additional long-term discontinuation.
• Urine testing in a subset (n=25) was 100% concordant with logs.

Key secondary outcomes

• Discontinuation at 1 week post-treatment: MTcap 88.4% vs SGT 67.4%; absolute difference 21%.
  • Effect size: OR 3.68 (95% CI 1.67–8.12).
  • NNT = 5.
• BZRA frequency (nights/week): Greater reduction with MTcap at 1 week (difference −1.31; 95% CI −2.05 to −0.57); no significant difference at 6 months.
• BZRA dose (diazepam-equivalent mg): No significant between-group differences at 1 week or 6 months.
• Insomnia Severity Index (ISI): Clinically meaningful improvements in both groups at 1 week (~6–8 point reductions) and maintained at 6 months; no between-group differences.
• DBAS-16 MED: Improved in both groups; no between-group differences.

Safety

• Adverse events leading to discontinuation or acute care were infrequent and similar between groups (eg, falls: 2 MTcap, 1 SGT; chest tightness/viral illness: 1 MTcap; dyspnea: 1 SGT).

Limitations

• Self-reported medication logs (urine testing only in a subset) may misclassify use.
• Excluded individuals with significant cognitive impairment and those outside California, limiting generalizability.
• COVID-19–related procedural adjustments; potential changes in non-BZRA sleep aids were not restricted.
• Observed between-group difference (14.8%) was smaller than the 20% assumed for power.
• Expectancy measure (DBAS-16 MED) may not fully capture targeted mechanisms.

Funding

Supported by the U.S. Department of Veterans Affairs Health Services Research & Development (Merit Award IIR17-234) and the National Institutes of Health (NIA and NHLBI awards), with additional institutional support. Funders had no role in study design, conduct, analysis, or publication decisions.

Citation

Fung CH, Alessi C, Martin JL, et al. Masked Taper With Behavioral Intervention for Discontinuation of Benzodiazepine Receptor Agonists: A Randomized Clinical Trial. JAMA Internal Medicine. 2024;184(12):1448-1456. doi:10.1001/jamainternmed.2024.5020. ClinicalTrials.gov: NCT03687086.