Diazepam
| Evidence Level: L3 | Predicted Indications: 10 |
Table of Contents
Diazepam: From an Established Benzodiazepine to Insomnia
One-Sentence Summary
Diazepam is a benzodiazepine that is currently marketed in the US, but the supplied record does not list its labeled indications. The TxGNN model predicts it may be effective for insomnia, with 24 registered clinical trials and 18 publications retrieved for this concept. Most of that evidence concerns hypnotic tapering, discontinuation and dependence. The only direct clinical comparison is a 1981 outpatient study, so the actual support for diazepam efficacy in insomnia is thin.
Quick Overview
| Item | Content |
|---|---|
| Predicted New Indication | Insomnia |
| TxGNN Prediction Score | 99.9997% |
| Evidence Level | L3 |
| US Market Status | ✓ Marketed |
| Number of NDAs | 20 |
| Recommended Decision | Hold |
Why is This Prediction Reasonable?
Detailed mechanism-of-action data is not available in the record. Diazepam is a positive allosteric modulator of the GABA-A receptor. Enhancing inhibitory GABA signaling produces sedative-hypnotic effects, so the mechanistic rationale for insomnia is plausible and well established.
Benzodiazepines as a class have long been used for anxiety and insomnia. A review in the literature set (PMID 39581171) describes diazepam as a prototypical GABA-A modulator that gives therapeutic effects but also sedation-related side effects.
Direct proof is limited, however. The registered trials mostly study hypnotic discontinuation, tapering and dependence, not diazepam efficacy. The only direct clinical evidence is a 1981 comparison in which diazepam was the comparator. The model prediction is therefore best treated as a research question rather than an established new use.
Clinical Trial Evidence
Of the 24 trials retrieved, the 10 most relevant are listed below. None tests diazepam as an insomnia treatment.
| Trial Number | Phase | Status | Enrollment | Key Findings |
|---|---|---|---|---|
| NCT04050176 | Phase 3 | Active, not recruiting | 260 | Blinded vs open-label hypnotic tapering combined with CBT for insomnia. Tests discontinuation, not diazepam efficacy. |
| NCT05935553 | Phase 2/3 | Recruiting | 93 | Baclofen to improve benzodiazepine titration in dependence. Addresses the downside of long-term use. |
| NCT02831894 | Phase 2 | Completed | 74 | Effect of tapering pace on hypnotic discontinuation in insomnia. Provides withdrawal context. |
| NCT03461042 | Phase 4 | Completed | 17 | Ramelteon added during dose reduction or interruption of benzodiazepine and non-benzodiazepine hypnotics in chronic insomnia. |
| NCT04751851 | N/A | Completed | 128 | Acceptance and Commitment Therapy added to a benzodiazepine withdrawal program in hypnotic-dependent insomnia. |
| NCT01893632 | Phase 2 | Terminated | 2 | Gabapentin for benzodiazepine dependence. Terminated with only 2 participants. |
| NCT03687086 | N/A | Completed | 188 | Non-drug mechanism to help older adults stop sleeping pills. Indirect. |
| NCT02648776 | N/A | Unknown | 1400 | Prospective cohort on risk and benefit of hypnotics in elderly patients at a Taiwanese medical center. |
| NCT04364321 | N/A | Unknown | 74 | Single-dose clonazepam vs intermittent diazepam for recurrent febrile seizures in children. Not an insomnia trial. |
| NCT05646693 | Phase 2 | Unknown | 58 | Adepsique (amitriptyline, perphenazine, diazepam) in chronic tinnitus. Diazepam is only one component of the combination. |
Literature Evidence
Of the 18 publications retrieved, the 9 most relevant are listed below.
| PMID | Year | Type | Journal | Key Findings |
|---|---|---|---|---|
| 6113175 | 1981 | Comparative clinical trial (double-blind) | J Int Med Res | 100 outpatients with insomnia, 7 days. Lormetazepam 1 mg was significantly better than diazepam 5 mg at shortening time to fall asleep and prolonging uninterrupted sleep. |
| 39581171 | 2024 | Review | Bioorg Chem | Reviews GABA-A receptor modulators. Diazepam is a positive allosteric modulator with therapeutic effects but also sedation-related side effects. |
| 36692463 | 2023 | Meta-analysis | Acta Pharm | Assesses tranquilizers in elderly patients by dose, outcomes and adverse effects. The supplied abstract gives no results. |
| 35196378 | 2022 | Nonrandomized intervention | Fam Pract | Discontinuation of chronic benzodiazepine use in primary care. The supplied abstract gives no results. |
| 6114852 | 1981 | Review | Drugs | Triazolam review. Its short half-life is described as more suitable for insomnia than longer-acting benzodiazepines. Does not evaluate diazepam directly. |
| 35228700 | 2022 | Preclinical (mouse) | Nat Neurosci | Long-term diazepam impaired dendritic spine plasticity and caused cognitive impairment via TSPO. A safety signal. |
| 40583063 | 2025 | Clinical and mechanistic study | Cell Mol Biol Lett | Prolonged use of benzodiazepines and Z-drugs was linked to higher breast cancer risk. A long-term safety signal. |
| 40350874 | 2025 | Preclinical (mouse) | Zhongguo Zhong Yao Za Zhi | Diazepam served only as the positive control in a herbal-extract study. |
| 37776625 | 2023 | Preclinical (rat) | J Pharm Biomed Anal | Diazepam served only as the positive control in a metabolomics study of a Chinese medicine for insomnia. |
US Market Information
There are 20 authorizations in total. Five are listed below. The record gives no approved-indication text for any of them.
| Authorization Number | Product Name | Dosage Form | Approved Indication |
|---|---|---|---|
| ANDA217843 | Diazepam (ASCLEMED USA INC.) | Tablet | Not provided in the record |
| NDA020648 | Diazepam (Oceanside Pharmaceuticals) | Gel | Not provided in the record |
| ANDA070325 | Diazepam (Cardinal Health 107, LLC) | Tablet | Not provided in the record |
| ANDA070325 | Diazepam (Mylan Pharmaceuticals Inc.) | Tablet | Not provided in the record |
| ANDA206477 | Diazepam (Chartwell RX, LLC) | Solution | Not provided in the record |
Other forms in the record include injection solution and spray.
Safety Considerations
Please refer to the package insert for safety information. No warning, contraindication or drug-interaction data was available in the record.
The supplied literature points to concerns that should be weighed in any insomnia use:
- Dependence, tolerance and withdrawal: much of the trial evidence exists because long-term hypnotic use causes dependence that is hard to stop.
- Cognitive impairment: long-term diazepam impaired cognition in a mouse study (PMID 35228700).
- Older adults: fall risk and cognitive effects are especially relevant in this group.
- Possible cancer signal: prolonged benzodiazepine and Z-drug use was linked to higher breast cancer risk (PMID 40583063).
Conclusion and Next Steps
Decision: Hold
Rationale: The GABA-A mechanism makes the prediction plausible. However, no modern trial tests diazepam for insomnia, and the only direct clinical study is from 1981. The main long-term safety concerns are dependence, cognitive impairment and fall risk. The package-insert safety data is also missing, which blocks the safety screening step.
The other nine predicted indications are weaker still: they are graded L4 to L5 and are also on Hold.
To proceed, the following is needed:
- Package insert warnings, contraindications and labeled indications for the US products
- Detailed mechanism of action data from DrugBank
- A systematic review comparing diazepam with current first-line insomnia treatments, including CBT for insomnia and newer hypnotics
- A safety plan for short-term use, dependence risk and older-adult populations, if a trial is considered
Disclaimer
This content is for research purposes only and does not constitute medical advice. Clinical validation is required before any clinical application.