Diazepam

證據等級: L5

預測適應症: 10 個

Diazepam: From Anxiety & Seizure Management to Insomnia

One-Sentence Summary

Diazepam is a classic benzodiazepine historically established for anxiety disorders, alcohol withdrawal, muscle spasms, and seizure control. The TxGNN model predicts it may be effective for Insomnia, with 24 clinical trials and 18 publications currently supporting this direction. The mechanistic basis is well-founded, though meaningful long-term safety concerns — including dependence, disrupted sleep architecture, and cognitive impairment — substantially constrain its therapeutic application.

Quick Overview

Item	Content
Original Indication	Regulatory data not retrieved (known pharmacological uses: anxiety disorders, seizures, muscle spasms, alcohol withdrawal)
Predicted New Indication	Insomnia (disease)
TxGNN Prediction Score	99.9997%
Evidence Level	L2
US Market Status	Data retrieval gap (0 NDA records found; Valium/Diazepam historically FDA-approved since 1963)
Number of NDAs	0 (data gap)
Recommended Decision	Proceed with Guardrails

Why is This Prediction Reasonable?

Although the formal DrugBank mechanism of action record was not retrieved for this dataset, Diazepam’s pharmacology is thoroughly characterized in published literature. Diazepam acts as a positive allosteric modulator (PAM) of the GABA-A receptor, binding to the benzodiazepine site between the α and γ subunits. This enhances the frequency of chloride channel opening in response to GABA, causing neuronal hyperpolarization and generalized CNS depression. The net effects include reduced sleep latency, increased total sleep time, and prolonged Stage 2 NREM sleep — making the hypnotic application pharmacologically straightforward and mechanistically inevitable.

The connection between Diazepam’s established anxiolytic properties and insomnia is clinically meaningful because insomnia frequently co-occurs with anxiety. The hyperarousal state that delays sleep initiation is directly addressed by GABAergic inhibition. This is why benzodiazepines were historically the standard of care for insomnia before z-drugs (zolpidem, zaleplon) and dual orexin receptor antagonists became available. PMID 6113175, the most direct RCT in this dataset, confirms Diazepam’s measurable hypnotic efficacy in patients with sleep-onset disorders — establishing it as the active benchmark comparator in subsequent insomnia research.

However, the prediction’s near-perfect score (99.9997%) reflects mechanistic inevitability rather than unexplored potential. The principal concern is not short-term efficacy but long-term risk. Chronic use suppresses REM sleep and slow-wave sleep, disrupts NREM slow oscillations and spindle coupling critical for memory consolidation in older adults (PMID 40570297, 2025, Sleep), and impairs dendritic spine structural plasticity via mitochondrial TSPO activation, causing persistent cognitive decline in animal models (PMID 35228700, 2022, Nature Neuroscience). Current clinical guidelines position benzodiazepines as short-term bridging agents only, with CBTi (Cognitive Behavioral Therapy for Insomnia) as the preferred first-line treatment.

Clinical Trial Evidence

Trial Number	Phase	Status	Enrollment	Key Findings
NCT05935553	Phase 2/3	Recruiting	93	Baclofen (GABA-B agonist) as adjunct to optimize Diazepam titration during BZD dependence treatment; Diazepam used as the reference standard, providing direct dosing and safety data relevant to insomnia management
NCT04050176	Phase 3	Active, Not Recruiting	260	Blinded vs. open-label BZD tapering combined with CBTi for chronic insomnia; evaluates discontinuation rates and the placebo contribution to hypnotic-dependent patients — defines boundary conditions for BZD use
NCT03461042	Phase 4	Completed	17	Ramelteon adjunct during (non-)BZD dose reduction algorithm for chronic insomnia; addresses transitional strategy from BZD to safer hypnotic alternatives
NCT04751851	N/A	Completed	128	Acceptance and Commitment Therapy (ACT) vs. standard support added to BZD withdrawal program in adults with hypnotic-dependent insomnia; optimal withdrawal modalities assessed
NCT07417813	N/A	Recruiting	121	Lemborexant (dual orexin antagonist) for insomnia comorbid with psychiatric disorders; benchmarks next-generation agents against BZD/Z-drug background
NCT02648776	N/A	Unknown	1,400	Prospective Taiwanese cohort examining risk-benefit of BZD hypnotics in elderly patients; pharmacokinetics, clinical outcomes, and long-term safety — directly relevant to chronic insomnia management with BZDs
NCT03687086	N/A	Completed	188	Novel cognitive mechanisms for helping older adults discontinue hypnotics; demonstrates that tapering alone is insufficient — motivational and cognitive targets must also be addressed
NCT02281175	N/A	Completed	114	PASSE-65+ psychosocial program for gradual BZD weaning in elderly; 12-week randomized trial demonstrating feasibility of structured discontinuation with psychological support
NCT02831894	Phase 2	Completed	74	Comparison of fast vs. slow BZD tapering in treatment-seeking insomnia patients; individual traits (anxiety sensitivity, personality) predicted successful discontinuation
NCT05646693	Phase 2	Unknown	58	Antioxidant therapy combined with Adepsique® (amitriptyline + perphenazine + diazepam) for chronic tinnitus; direct diazepam-containing intervention arm with inflammatory cytokine and oxidative stress outcomes

Literature Evidence

PMID	Year	Type	Journal	Key Findings
6113175	1981	RCT	J Int Med Res	Double-blind 7-day trial (n=100): Lormetazepam 1 mg vs. Diazepam 5 mg for sleep disorders; Diazepam meaningfully reduced sleep latency and prolonged uninterrupted sleep, confirming direct hypnotic efficacy
39581171	2024	Review	Bioorganic Chem	Comprehensive review of GABA-A receptor modulators; Diazepam identified as the prototypical PAM with established activity for anxiety, epilepsy, and insomnia; sedation, tolerance, and dependence highlighted as key clinical limitations
40570297	2025	Cohort Study	Sleep	Chronic BZD/BZRA use in older adults with insomnia significantly disrupts NREM slow oscillations and spindle coupling — functions critical for memory consolidation; supports strict short-term use only
35228700	2022	Preclinical	Nature Neuroscience	Long-term diazepam impairs dendritic spine structural plasticity via mitochondrial TSPO, increasing microglial engulfment and causing persistent cognitive impairment in mice; raises dementia risk concern for chronic use
7595266	1995	Review	J Fam Pract	Systematic review of BZD therapy for insomnia in community-dwelling elderly; short-term sleep lab efficacy confirmed across 10 trials; long-term use associated with increased fall and injury risk
6135990	1983	Review	N Engl J Med	Landmark NEJM review establishing clinical pharmacokinetics and rational prescribing of BZDs; forms the foundational evidence base for Diazepam as a hypnotic agent
7525193	1994	Review/Guidelines	Drugs	Evidence-based guidelines for rational BZD use; recommends limiting hypnotic prescriptions to transient or short-term insomnia; longer half-life agents (including Diazepam) associated with greater daytime residual sedation and hangover
40583063	2025	Study	Cell Mol Biol Lett	Retrospective clinical analysis: prolonged BZDR use (Diazepam and Zolpidem) significantly exacerbates breast cancer risk; mechanistic evidence via GABA-A receptor signaling — safety signal for long-term use in women
29479317	2018	Review/Meta-analysis	Front Pharmacol	Meta-analysis of Suanzaoren formulae for insomnia RCTs; Diazepam consistently used as active comparator, affirming its benchmark hypnotic status in controlled insomnia research
37776625	2023	Animal Study	J Pharm Biomed Anal	Naoling Pian vs. Diazepam (positive control) in PCPA-induced insomnia rats; Diazepam produced significant sleep prolongation and behavioral normalization, validating its preclinical sedative-hypnotic efficacy profile

Safety Considerations

Please refer to the package insert for safety information.

Conclusion and Next Steps

Decision: Proceed with Guardrails

Rationale: Diazepam has a well-established mechanistic basis for treating insomnia (GABA-A positive allosteric modulation → CNS inhibition → reduced sleep latency and increased sleep duration), confirmed by a direct RCT (PMID 6113175) and endorsed in multiple guideline-level reviews spanning four decades. The L2 evidence level is grounded in real clinical data. However, the core clinical evidence is aging (1980s–1990s), and more recent findings document significant long-term risks: disrupted sleep architecture, cognitive impairment, dependency, and potential oncological associations. Current clinical practice has largely transitioned to CBTi and newer pharmacological agents for chronic insomnia.

To proceed, the following is needed:

Retrieve the US FDA package insert and complete NDA records via the FDA Orange Book (NDA 013263 for Valium and associated generics) to fill the US regulatory data gap
Obtain formal DrugBank MOA documentation (DG002) for standardized mechanistic reporting
Define a strict duration limit for any proposed use (2–4 weeks per current international guidelines)
Establish clear patient exclusion criteria: adults aged ≥65 years, patients with prior or current substance use disorder, obstructive sleep apnea, and pregnant or nursing individuals
Develop a pre-planned dose-tapering exit protocol to minimize discontinuation syndrome and rebound insomnia
Evaluate whether this repurposing pathway provides incremental clinical value over already-approved alternatives (z-drugs, ramelteon, lemborexant/suvorexant) before committing to a formal development program
Disclaimer

This content is for research purposes only and does not constitute medical advice. Clinical validation is required before any clinical application.