Anesthesia-Informed Approaches to Insomnia: Shared Neural Circuits, Biomimetic Sleep, and Emerging Interventional Strategies

Abstract

Background

Insomnia affects approximately 10–30% of adults, burdening quality of life, cognition, and health. Although cognitive behavioral therapy for insomnia (CBT-I) and approved pharmacotherapies remain standard care, limited accessibility, durability, and efficacy in refractory populations have motivated interest in mechanism-based alternatives, including approaches derived from anesthetic pharmacology.

Objective

This narrative review examines the convergent neural circuits underlying anesthesia-induced and physiological sleep, critically appraises evidence for anesthetic agents—dexmedetomidine, ketamine/esketamine, and propofol—delivered through non-traditional modalities, and evaluates perioperative sleep optimization, with explicit attention to evidence level and limitations.

Methods

We searched PubMed/MEDLINE, Embase, Web of Science, and the Cochrane Library for English-language publications through June 2026. Evidence from randomized controlled trials (RCTs), systematic reviews, observational studies, case reports, and mechanistic studies was synthesized narratively and graded using the Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence.

Results

Dexmedetomidine induces “biomimetic sleep” via alpha-2 adrenergic agonism, resembling natural non-rapid eye movement stage 2 sleep more closely than propofol. However, most supporting evidence is mechanistic, perioperative, or from small pilot studies rather than insomnia-specific RCTs. Patient-controlled sleep systems, intranasal delivery, and bionic sleep therapy are promising but preliminary, with incomplete long-term safety data, particularly regarding cardiovascular effects and withdrawal. Ketamine and esketamine may enhance slow-wave sleep, but most data derive from comorbid depression, confounding attribution. Stellate ganglion block, acupuncture, and closed-loop stimulation remain investigational and susceptible to placebo effects. Perioperative strategies improve postoperative sleep in RCTs, but extrapolation to chronic insomnia is limited by fundamental etiological differences.

Conclusion

Anesthesia-informed approaches provide mechanistic insights and investigational directions rather than established alternatives to CBT-I or approved hypnotics. Priorities include adequately powered insomnia-specific RCTs, comparison with standard treatments, long-term safety evaluation, and phenotype-guided selection.

Introduction

Sleep constitutes a fundamental physiological process essential for immune regulation, cognitive restoration, metabolic homeostasis, and emotional stability.1 Insomnia is the most commonly reported sleep disorder in clinical practice, with prevalence estimates ranging from 10–30% in the general adult population and increasing markedly to 40–60% in perioperative and critically ill cohorts.2,3 According to the International Classification of Sleep Disorders, Third Edition (ICSD-3), insomnia is characterized by difficulty initiating or maintaining sleep, early morning awakening with inability to return to sleep, or non-restorative sleep despite adequate opportunity and circumstances, resulting in significant daytime functional impairment.4 Chronic insomnia not only diminishes health-related quality of life but also confers increased risk for cardiovascular disease, metabolic syndrome, type 2 diabetes, neurodegenerative disorders, and psychiatric comorbidities, particularly major depressive disorder and generalized anxiety disorder.5

Current evidence-based management of insomnia rests on two pillars. Cognitive behavioral therapy for insomnia (CBT-I) is recommended as first-line treatment by international guidelines, including the update of the European Insomnia Guideline, on the basis of durable efficacy and an favorable safety profile.6 Pharmacological options include benzodiazepine receptor agonists (traditional benzodiazepines and non-benzodiazepine “Z-drugs”), melatonin receptor agonists, histamine H1 antagonists, certain sedating antidepressants, and—more recently—dual orexin receptor antagonists (DORAs; suvorexant, lemborexant, and daridorexant), which selectively target wake-promoting pathways rather than globally suppressing central nervous system activity.7–9 Nevertheless, important limitations persist: access to trained CBT-I providers remains restricted in many health systems; a substantial proportion of patients respond incompletely or relapse; and available hypnotics carry risks of tolerance, dependence, cognitive and psychomotor impairment, respiratory depression, and—in the case of DORAs—cost and variable availability.6,7,9 These gaps are most apparent in patients with chronic refractory insomnia, in whom multiple conventional treatments have failed. It is within this specific context—not as replacements for established therapy—that anesthesia-informed approaches warrant critical examination.

Any discussion of anesthetic agents for sleep must begin with a clear conceptual distinction among three states that are frequently conflated. Physiological sleep is a spontaneously reversible, homeostatically regulated state that preserves arousability to salient stimuli, exhibits ultradian cycling between NREM and rapid eye movement (REM) stages, and supports memory consolidation and synaptic homeostasis. Sedation is a drug-induced continuum of depressed consciousness in which some of these features—spontaneous arousability, stage cycling, and restorative processes—are partially preserved, whereas general anesthesia is a controlled, drug-induced state comprising unconsciousness, amnesia, analgesia, and immobility in which natural sleep architecture is typically absent.10 The therapeutic rationale examined in this review therefore rests not on inducing unconsciousness per se, but on whether specific anesthetic agents can engage endogenous sleep circuits sufficiently to reproduce restorative features of natural sleep—an empirical question that, as discussed below, is only partly answered.

In parallel with these conceptual developments, anesthesiology has witnessed remarkable advances in understanding how anesthetic agents interact with endogenous sleep–wake regulatory circuits. The framework of “shared circuits” between anesthesia and natural sleep posits that general anesthesia engages overlapping neural substrates with physiological sleep–wake regulation, encompassing key subcortical nuclei including the ventrolateral preoptic nucleus (VLPO), locus coeruleus (LC), tuberomammillary nucleus (TMN), lateral hypothalamus, and dorsal raphe nucleus, as well as thalamocortical networks that govern consciousness and arousal states.10 This convergence has catalyzed development of non-traditional therapeutic modalities at the interface of anesthesia and sleep medicine: patient-controlled sleep (PCSL) systems using dexmedetomidine;11 bionic sleep therapy (BST) combining propofol with dexmedetomidine;12 intranasal dexmedetomidine delivery;13,14 stellate ganglion block (SGB) combined with dexmedetomidine;15 and acupuncture-point stimulation for perioperative sleep disturbance.16 Intracranial electroencephalography (EEG) studies have demonstrated that dexmedetomidine produces more sleep-like brain activity than propofol in human participants,17 and a recent crossover trial reported that pharmacological enhancement of glymphatic transport during a single sleep opportunity increased beta-amyloid and tau clearance by approximately 9–10%.18

Three considerations that directly determine the risk–benefit balance of these approaches have, however, received insufficient attention in earlier discussions and are addressed throughout this review. First, obstructive sleep apnea (OSA) is highly prevalent among patients presenting with refractory insomnia—comorbid insomnia and sleep apnea (COMISA) affects an estimated 30–50% of each patient population—and sedative agents that reduce upper airway tone or ventilatory drive may be hazardous in undiagnosed OSA, mandating systematic screening before any anesthesia-informed intervention.19 Second, repeated or prolonged dexmedetomidine exposure is associated with withdrawal phenomena, including rebound sympathetic overactivity, tachycardia, and hypertension,20,21 and ketamine carries recognized risks of misuse and lower urinary tract toxicity,22 so that safety frameworks developed for single perioperative exposures cannot simply be transferred to chronic insomnia treatment. Third, drug responses differ by sex and age—for example, propofol requirements vary systematically with both factors23—implying that dosing strategies extrapolated from undifferentiated trial populations may be inappropriate for individual patients.

This review critically analyzes: (1) the shared neurobiological mechanisms between anesthesia and sleep–wake regulation; (2) the effects of specific anesthetic agents on sleep architecture and their therapeutic applications through non-traditional modalities; (3) emerging interventional and non-pharmacological treatments for chronic and refractory insomnia; (4) perioperative sleep optimization strategies and the limits of their extrapolation to primary insomnia; and (5) future directions encompassing pharmacogenomics, artificial intelligence, and novel drug delivery systems. Throughout, we distinguish established evidence from preliminary findings, grade the strength of evidence explicitly, and position anesthesia-informed approaches relative to current first-line therapies.

Review Methodology

A narrative format was chosen because the field spans heterogeneous evidence types—from molecular and animal mechanistic studies to perioperative RCTs and single-patient case reports—for which a single systematic question with predefined eligibility criteria would be premature, and because several of the interventions discussed (eg, patient-controlled sleep systems, bionic sleep therapy) are supported by only one or two primary studies, precluding meaningful quantitative synthesis. We acknowledge the inherent risk of selection bias in this format and have sought to mitigate it through the transparent procedures described below and through deliberate inclusion of negative, inconclusive, and conflicting findings.

We searched PubMed/MEDLINE, Embase, Web of Science, and the Cochrane Library from database inception through June 2026. Core search strings combined agent- or technique-specific terms (eg, “dexmedetomidine”, “ketamine”, “esketamine”, “propofol”, “stellate ganglion block”, “acupuncture”, “transcutaneous electrical acupoint stimulation”, “closed-loop acoustic stimulation”, “transcranial magnetic stimulation”) with sleep-related outcome terms (eg, “insomnia”, “sleep quality”, “sleep architecture”, “polysomnography”, “postoperative sleep disturbance”, “slow-wave sleep”, “glymphatic”). Reference lists of included articles and relevant reviews were screened manually to identify additional studies. Conference abstracts, non-English publications, and studies without extractable sleep-related outcomes were excluded; preprints were included.

Because of the heterogeneity of the evidence base, we did not apply a single risk-of-bias instrument. Instead, each source cited for a therapeutic claim was classified according to the Oxford Centre for Evidence-Based Medicine (OCEBM) 2011 Levels of Evidence,24 and the strength of each conclusion in this review is matched to that classification (Table 1). Where evidence derives exclusively from perioperative populations, animal models, or uncontrolled observations, this is stated explicitly in the text. Study selection and evidence classification were performed by the first author and verified independently by two co-authors, disagreements were resolved by consensus.

Table 1.

Evidence Summary for Anesthesia-Informed Insomnia Approaches, Graded by the Oxford Centre for Evidence-Based Medicine (OCEBM) 2011 Levels of Evidence

Intervention	Proposed Mechanism	Representative Evidence	Population	OCEBM Level*	Main Limitations
Dexmedetomidine biomimetic sedation	Alpha-2A agonism, LC suppression, VLPO disinhibition	Intracranial EEG;17 spindle/N3 studies;25,26 crossover PSG trial27	Healthy volunteers; surgical patients	2 (for acute physiological outcomes)	Surrogate EEG outcomes; possible homeostatic displacement;28 restorative equivalence to natural sleep unproven
Patient-controlled sleep (PCSL)	Repeated biomimetic sleep restoring homeostasis	Pilot study;11 retrospective TRD cohort29	Refractory insomnia; treatment-resistant depression	4	Single research group; uncontrolled/retrospective; IV access, monitoring, cost, regulatory barriers
Intranasal dexmedetomidine	Same as above; slower absorption	RCTs;13,14,30 dose-finding trial31	Perioperative and procedural patients (one trial enrolled chronic insomnia patients undergoing surgery30)	2 (perioperative outcomes)	No RCT in non-surgical chronic insomnia; long-term nasal safety unknown
Ketamine/esketamine	NMDA antagonism; BDNF-mediated plasticity	PSG study;32 RCT in comorbid insomnia–depression;33 pilot;34 long-term depression safety data35	Depression/treatment-resistant depression with insomnia	2–4	Sleep benefit confounded by antidepressant effect; misuse and urinary toxicity;22 no primary-insomnia trials
Propofol/bionic sleep therapy (BST)	GABA-A potentiation + alpha-2 agonism	Single case report12	Single refractory-insomnia patient	4–5	Hypothesis-generating only; airway and cardiorespiratory risk; misuse potential
Stellate ganglion block (SGB)	Cervical sympatholysis reducing hyperarousal	Perioperative combination RCT;15 insomnia pilot;36 dosing study;37 systematic review38	Perioperative patients; small insomnia cohorts	2–4	No sham-controlled trial; transient benefit; placebo/regression-to-mean unexcluded36
Acupuncture/TEAS	Multisite neuromodulation; endogenous opioids; melatonin	Perioperative meta-analysis of 9 RCTs;16 sham-controlled RCT (negative on primary outcome)39	Perioperative patients; cancer-related insomnia	1–2 (inconsistent)	Placebo-dominated responses;39 subjective outcomes; few insomnia-specific trials
Closed-loop acoustic stimulation (CLAS)	Phase-locked slow-wave enhancement	Mechanistic crossover study28	Healthy young adults	2 (mechanistic)	No insomnia-population trial; homeostatic-displacement caveat
rTMS + dexmedetomidine	Cortical inhibition + alpha-2 agonism; anti-inflammatory	Single small RCT40	Chronic insomnia disorder	2 (single small RCT)	Independent replication absent
Perioperative sleep preservation (dexmedetomidine, opioid-sparing analgesia, circadian measures)	Multimodal sleep protection	Meta-analysis of RCTs;41 RCTs;42,43 critical care guideline44	Surgical and ICU patients	1–2	Short-term outcomes only; not evidence for chronic insomnia treatment

Notes: *OCEBM 2011 Levels of Evidence for treatment benefits: 1 = systematic review of randomized trials; 2 = randomized trial; 3 = non-randomized controlled cohort; 4 = case series or case-control; 5 = mechanism-based reasoning.24 Levels here reflect the best available evidence for the stated indication, not for chronic insomnia treatment overall.

Shared Neurobiological Mechanisms Between Anesthesia and Sleep–Wake Regulation

Both anesthesia and natural sleep are characterized by reversible unconsciousness, reduced responsiveness to external stimuli, and specific patterns of cortical EEG activity.10 However, fundamental differences persist: natural sleep retains the capacity for rapid arousal in response to salient stimuli, supports memory consolidation and synaptic homeostasis, and exhibits characteristic ultradian cycling between NREM and REM sleep stages—features that are disrupted or absent during pharmacologically induced unconsciousness.10 These distinctions matter clinically: producing an EEG pattern that resembles sleep is not equivalent to restoring the full restorative function of sleep, and claims of therapeutic benefit must therefore rest on clinical outcomes rather than on electrophysiological resemblance alone.

Convergent Neural Circuits and Receptor Mechanisms

The ventrolateral preoptic nucleus (VLPO) serves as the principal sleep-promoting nucleus, containing GABAergic and galaninergic neurons that are activated during sleep and send inhibitory projections to multiple arousal-promoting nuclei throughout the brainstem and hypothalamus.10 Seminal work established the “flip-flop switch” model of sleep–wake regulation, in which mutual inhibition between the VLPO and arousal centers generates stable, discrete state transitions.45 Evidence demonstrates that GABAergic anesthetics including propofol, isoflurane, and sevoflurane activate VLPO neurons, thereby promoting unconsciousness through engagement of endogenous sleep pathways.45 However, optogenetic and chemogenetic studies have revealed important dissociations: while some anesthetics converge on VLPO-dependent mechanisms, isoflurane-induced unconsciousness does not require an intact VLPO, indicating that anesthesia and sleep, despite their similarities, are not identical states at the circuit level.46

Dexmedetomidine produces a state neurophysiologically resembling natural NREM sleep through a mechanism fundamentally distinct from that of GABAergic agents. By acting on presynaptic alpha-2A receptors in the locus coeruleus, dexmedetomidine suppresses noradrenergic neurotransmission to the VLPO and other arousal centers, thereby disinhibiting VLPO neurons and activating endogenous sleep-promoting pathways.47 This mechanism—engaging rather than bypassing physiological sleep circuits—produces what has been termed “biomimetic sleep,” characterized by EEG features resembling NREM stage 2 sleep, including prominent sleep spindles and slow-wave activity.25,26 A human intracranial EEG study provided direct evidence: dexmedetomidine produces more sleep-like brain activity than propofol, with changes in spectral power, network entropy, and functional connectivity that align more closely with natural N2 sleep than with GABAergic anesthesia.17 Functional connectivity analyses have further demonstrated that dexmedetomidine reduces connectivity within the default mode network, frontoparietal network, and limbic system—changes associated with EEG slow-delta and spindle oscillations.48 Importantly, however, biomimetic EEG features do not guarantee equivalent restorative or cognitive benefit compared with natural sleep: whether dexmedetomidine-induced states reproduce the synaptic homeostasis, memory consolidation, and subjective restoration of physiological sleep remains incompletely established, and at least one study suggests partial interference with subsequent sleep homeostasis (Dexmedetomidine: Biomimetic Sleep and Novel Delivery Modalities).28

Ketamine, a non-competitive NMDA receptor antagonist, modulates the activity of sleep–wake regulatory circuits and may promote slow-wave sleep through downstream effects on brain-derived neurotrophic factor (BDNF) signaling and synaptic homeostasis.32 The orexin (hypocretin) system plays a critical role in stabilizing wakefulness. DORAs selectively inhibit the orexinergic arousal system, facilitating sleep without disrupting natural sleep architecture.8 Pharmacogenomic studies have demonstrated that HCRTR1/2 gene variants significantly influence sleep–wake transitions and hemodynamic stability during propofol, dexmedetomidine, and remifentanil anesthesia, providing evidence that individual genetic variation in the orexin system may predict differential responses to anesthetic agents.49 It should be noted that while orexin receptor agonists were previously investigated for acceleration of anesthesia emergence, clinical development of these compounds for this indication has been discontinued.

Neuroinflammatory, Glymphatic, and Neuroplastic Mechanisms

Chronic insomnia is increasingly recognized as a state of low-grade inflammation, with elevated circulating levels of pro-inflammatory cytokines including interleukin-1β (IL-1β), interleukin-6 (IL-6), and tumor necrosis factor-alpha (TNF-α) observed in patients with persistent sleep disturbances.50 In preclinical models, dexmedetomidine attenuates neuroinflammation through inhibition of microglial activation and reduction of pro-inflammatory cytokine release via MAPK/ERK or NLRP3 pathways;51,52 whether these anti-inflammatory effects contribute meaningfully to clinical sleep outcomes in humans is unknown.

The glymphatic system—a cerebrospinal fluid–interstitial fluid exchange system dependent on aquaporin-4 (AQP4) channels in astrocytic end-feet—facilitates clearance of metabolic waste products including beta-amyloid and tau during sleep in animal models.53 Dexmedetomidine has been reported to enhance glymphatic clearance in preclinical studies through suppression of central noradrenergic tone and induction of slow-wave-rich states.54 Recent human work has begun to test these mechanisms directly: a companion study validated a plasma-based assay demonstrating glymphatic clearance of amyloid beta and tau from brain to plasma in humans,55 and a crossover trial in healthy older adults (currently available as a preprint) reported that a pharmacological combination of intravenous dexmedetomidine with the alpha-1 agonist midodrine increased beta-amyloid and tau clearance by approximately 9–10% during a single sleep opportunity, with Bayesian mediation analysis identifying increased EEG slow-wave activity as a mediator.18 Critically, dexmedetomidine alone in that trial did not increase clearance—plausibly because drug-induced hypotension triggered cerebrovascular counter-regulation that offset the slow-wave benefit—and only the combination with midodrine achieved enhanced clearance.18 These findings are best regarded as proof of mechanism rather than evidence of clinical benefit: the long-term clinical implications of pharmacologically enhanced glymphatic clearance for neurodegenerative risk remain entirely under investigation, and any suggestion that anesthetic-induced sleep prevents neurodegeneration would be premature.

BDNF and neuroplasticity represent another proposed link between anesthetic pharmacology and sleep regulation. BDNF is essential for activity-dependent synaptic plasticity, neuronal survival, and cognitive function.32 In patients with major depression, ketamine-induced enhancement of BDNF signaling has been associated with improvements in slow-wave sleep activity specifically in treatment responders (r = 0.595, P = 0.032 for slow-wave activity; r = 0.626, P = 0.022 for high-amplitude slow waves), suggesting that ketamine-induced synaptic plasticity may underlie sleep enhancement in patients who achieve antidepressant response.32 This observation aligns with the synaptic homeostasis hypothesis of sleep, which posits that slow-wave sleep serves to downscale and rebalance synaptic connections strengthened during wakefulness,56 but it also illustrates a recurring interpretive caveat in this literature: because the data derive from depressed populations, it is difficult to separate direct hypnotic effects from sleep improvement secondary to mood improvement (Ketamine and Esketamine: NMDA Receptor Antagonism and Sleep Modulation).

The Gut–Brain Axis: An Emerging Frontier

The gut–brain axis has been increasingly implicated in sleep regulation. The gut microbiome produces neuroactive metabolites including short-chain fatty acids, gamma-aminobutyric acid, serotonin, and melatonin.57 Anesthetic agents, particularly propofol and volatile anesthetics, can alter gut microbiome composition in preclinical and early clinical studies, leading to dysbiosis that may persist into the postoperative period.57 Such dysbiosis could plausibly increase intestinal permeability, allowing bacterial endotoxins and inflammatory mediators to enter systemic circulation and cross the blood–brain barrier, thereby activating neuroinflammatory pathways that disrupt sleep architecture. This pathway remains largely hypothetical in humans: no adequately powered trial has yet tested whether perioperative modulation of the microbiome (eg, with probiotics) improves sleep outcomes, and we include this topic to highlight a mechanistic research direction rather than an actionable therapeutic strategy.

Anesthetic Agents and Non-Traditional Therapeutic Applications

Dexmedetomidine: Biomimetic Sleep and Novel Delivery Modalities

Dexmedetomidine, a highly selective alpha-2 adrenergic receptor agonist with approximately 8-fold greater selectivity for the alpha-2A subtype than clonidine, has emerged as one of the most studied agents for sleep-related applications due to its unique pharmacological profile.58 Unlike GABAergic anesthetics that produce a pharmacodynamically distinct state, dexmedetomidine induces biomimetic sleep through engagement of endogenous VLPO-mediated pathways.

Mechanism and EEG Evidence

Intracranial EEG recordings confirm that dexmedetomidine sedation produces patterns resembling NREM stage 2 sleep, with prominent “dex-spindles” and progressive shift through theta to slow-delta oscillations.25 Delta power increases with depth of dexmedetomidine anesthesia in a manner similar to progressive stages of NREM sleep, whereas propofol-induced changes are smaller and more localized.17 Network entropy decreases with depth of dexmedetomidine anesthesia in a pattern closely mirroring natural sleep stages,17 and an independent pilot study demonstrated that dexmedetomidine can promote slow-wave (N3-like) activity in humans.26 These electrophysiological observations must, however, be interpreted alongside evidence of possible homeostatic interference. Smith et al demonstrated that daytime dexmedetomidine sedation with closed-loop acoustic stimulation alters subsequent slow-wave sleep homeostasis: post-sedation sleep showed reduced N3 duration (mean difference −17.1 minutes, 95% CI −30.0 to −8.2, P = 0.015) and reduced dissipation of slow-wave pressure (P = 0.02), with higher daytime dexmedetomidine doses correlating with greater decreases in subsequent NREM sleep (r = −0.61, P = 0.03).28 These data can be read in two ways: they support the biomimetic hypothesis insofar as dexmedetomidine-induced states appear to discharge part of the homeostatic NREM “pressure,” but they simultaneously raise the possibility of a homeostatic debt or displacement—daytime pharmacological sleep may partially substitute for, and thereby reduce, subsequent nocturnal deep sleep. For patients whose therapeutic goal is consolidation of nighttime sleep, this is a potentially adverse consequence that has received little attention and that could limit the suitability of daytime or around-the-clock dexmedetomidine regimens. It argues for evening-timed dosing and objective polysomnographic monitoring in future insomnia trials.

Beyond perioperative and experimental settings, evidence from independent research groups is beginning to emerge. Schnider et al conducted a randomized crossover trial of subanesthetic oromucosal dexmedetomidine (40 µg) in healthy volunteers and found objective polysomnographic benefits—sleep onset latency reduced by approximately 11.5 minutes, NREM sleep duration increased by approximately 37 minutes, and slow-wave energy increased by 23% in the first part of the night—without suppression of cortisol or melatonin secretion and without impairment of next-day alertness.27 Although conducted in non-insomniac volunteers, this study is important because it provides objective, placebo-controlled, non-perioperative evidence from a group independent of the teams that developed patient-controlled sleep, and it identifies a low-dose, non-invasive route with a favorable acute safety profile.

Patient-controlled sleep (PCSL) systems enable patients to self-administer pre-programmed dexmedetomidine doses via intravenous infusion pumps with patient-controlled bolus capability and safety lockout intervals. In a pilot study by An et al, 15 of 20 patients with chronic intractable insomnia—defined as persistent sleep disturbance for more than one year despite treatment with at least two conventional hypnotic agents—completed the treatment protocol.11 The PCSL device delivered dexmedetomidine at a continuous background infusion of 0.2 µg/kg/h with patient-triggered boluses of 0.3 µg/kg and a 10-minute lockout interval. The primary outcome, the Pittsburgh Sleep Quality Index (PSQI), decreased from a baseline of 14.33 ± 3.66 to 8.40 ± 6.15 after treatment (P < 0.05), with 12 of 15 patients demonstrating improvement.11 The investigators hypothesized that sustained efficacy may involve restoration of sleep homeostasis through repeated induction of biomimetic sleep states, and noted that dexmedetomidine doses gradually decreased during the treatment course in responsive patients.11 A subsequent retrospective study from the same group reported that esketamine combined with dexmedetomidine PCSL in 233 patients with treatment-resistant depression was associated with response rates of approximately 62.00%, 59.7360%, and 58₀ 49% at 1, 3, and 6 months without serious adverse events.29 These findings are encouraging but must be weighed against substantial limitations: the PCSL evidence derives almost entirely from a single research group, the pivotal study was uncontrolled with a small sample, the retrospective design of the larger follow-up precludes causal attribution, and the requirement for intravenous access with physiological monitoring restricts the approach to specialized centers. Feasibility, monitoring burden, cost, and regulatory barriers—an intravenous anesthetic delivery system used chronically outside the operating room or intensive care unit—remain largely unaddressed and are discussed further in Safety Considerations for Repeated or Chronic Anesthetic-Based Interventions.

Intranasal Dexmedetomidine Delivery

Intranasal administration offers high bioavailability (65–82%), rapid absorption through the nasal mucosa, and avoidance of first-pass hepatic metabolism.13,14 Several RCTs have examined this route, although almost exclusively in perioperative or procedure-related contexts. Yang et al, in a randomized three-blinded trial comparing intranasal dexmedetomidine (1.0 µg/kg), oral lorazepam (0.05 mg/kg), and placebo for preoperative anxiety-related insomnia, found intranasal dexmedetomidine superior to both comparators for sleep initiation, sleep quality, and next-day functioning without increasing adverse effects.13 He et al reported that preoperative intranasal dexmedetomidine (1.0 µg/kg) improved perioperative sleep quality and reduced postoperative neurocognitive deficits in elderly patients undergoing general anesthesia.14 Wu et al conducted a randomized double-blind trial in 110 older patients with chronic insomnia undergoing total hip or knee arthroplasty: intranasal dexmedetomidine (2.0 µg/kg) improved postoperative sleep quality versus placebo on the Leeds Sleep Evaluation Questionnaire (P < 0.0001), with improvements in total sleep time, sleep onset latency, and sleep efficiency, and reduced PSQI and Insomnia Severity Index scores, without significant adverse events.30 It should be emphasized that even this largest trial was performed in a surgical population, and its primary outcomes were postoperative; evidence that intranasal dexmedetomidine benefits non-surgical chronic insomnia is currently lacking. The slower onset with intranasal administration may be advantageous, allowing gradual sleep onset that more closely mimics natural sleep initiation, and slower absorption may attenuate peak hemodynamic effects (Safety Considerations for Repeated or Chronic Anesthetic-Based Interventions).

Dosage Optimization

Li et al, using a biased-coin up-and-down sequential allocation design, determined that 1.75 µg/kg represents the 95% effective dose (ED95) for adequate sedation in patients with chronic insomnia disorder (n = 60).31 Dose-dependent hypotension and bradycardia remain the principal acute safety concerns, most prominent during intravenous bolus administration; the intranasal and oromucosal routes appear to mitigate these effects through slower absorption and lower peak plasma concentrations.13,27 Long-term safety data for repeated or chronic dexmedetomidine administration in insomnia—including cumulative cardiovascular effects, withdrawal phenomena, and nasal mucosal tolerability—remain limited and are addressed in Safety Considerations for Repeated or Chronic Anesthetic-Based Interventions.

Ketamine and Esketamine: NMDA Receptor Antagonism and Sleep Modulation

Ketamine produces complex, dose-dependent effects on sleep architecture. Duncan et al, using nocturnal polysomnography in patients with treatment-resistant depression, demonstrated that a single intravenous ketamine infusion (0.5 mg/kg) increased total sleep time, slow-wave sleep, and REM sleep compared with baseline, with total slow-wave activity increasing particularly during the first NREM sleep cycle.32 Importantly, the correlation between slow-wave activity improvements and BDNF changes was observed exclusively in ketamine responders.32 This pattern—sleep improvement tracking antidepressant response—illustrates the central interpretive difficulty of this literature: most ketamine and esketamine sleep data originate from patients with comorbid depression, frequently receiving concomitant psychiatric medications, so observed sleep benefits may be mediated by mood improvement (or by study drug interactions) rather than reflecting a direct hypnotic action. Evidence in primary insomnia without depression is essentially absent, and polysomnographic findings in depressed populations should not be extrapolated to insomnia in general.

Esketamine, the S-enantiomer of ketamine with approximately four-fold greater affinity for NMDA receptors, has been studied primarily in treatment-resistant depression.34 A randomized trial by Ding et al recruited 84 patients with comorbid insomnia and depression (78 analyzed) and reported that dexmedetomidine plus esketamine produced superior sleep outcomes compared with dexmedetomidine alone: total sleep time increased by 163.32 ± 45.67 versus 111.53 ± 38.29 minutes; N3 duration was greater (62.4 ± 18.3 versus 41.7 ± 15.6 minutes); and sleep continuity index was higher (0.78 ± 0.12 versus 0.64 ± 0.14, all P < 0.001), with greater improvement in Montgomery–Åsberg Depression Rating Scale scores and BDNF increases proposed as mediators.33 Again, because both mood and sleep improved in parallel, attribution of the sleep benefit specifically to a hypnotic mechanism is not possible from these data. Long-term safety data exist for esketamine in depression: the SUSTAIN-3 open-label extension study (n = 1148, up to approximately 6.5 years of intermittent intranasal esketamine) identified no new safety signals, with 6.4% of patients discontinuing because of adverse events.35 Nevertheless, recognized class risks—including dissociation, blood pressure elevation, lower urinary tract symptoms and ulcerative cystitis with repeated exposure, and misuse or diversion potential—must inform any consideration of ketamine-based agents for chronic insomnia, a use for which neither regulatory approval nor dedicated long-term safety data exist.22

Perioperative Applications

Low-dose esketamine infusion during gynecological laparoscopic surgery has been reported to reduce postoperative sleep disturbance incidence.59 A meta-analysis confirmed that perioperative ketamine or esketamine improves the subjective quality of recovery after surgery across multiple surgical populations.60 Esketamine for patient-controlled intravenous analgesia significantly reduced postoperative sleep disturbance in elderly patients after total hip or knee arthroplasty,61 and intravenous esketamine improved postoperative sleep disturbance, anxiety, and depression in elderly patients undergoing laparoscopic abdominal surgery.62 These are perioperative analgesia-adjunct studies and should not be read as evidence for insomnia treatment outside the surgical setting.

Propofol: Context-Dependent Effects and Bionic Sleep Therapy

Propofol acts primarily as a positive allosteric modulator at GABA-A receptors, producing rapid-onset, dose-dependent unconsciousness. Evidence regarding propofol’s effects on postoperative sleep quality is mixed: propofol-based anesthesia was associated with better postoperative sleep quality than comparator regimens in some studies,63 while other studies indicate transient disruption resolving within 24 hours, consistent with its pharmacokinetic profile.64 Patients with preoperative sleep disturbances require higher propofol target plasma concentrations to achieve equivalent anesthetic depth compared with normal sleepers, potentially attributable to alterations in GABA–glutamate balance associated with chronic sleep deprivation.65 Propofol requirements and recovery profiles also differ systematically by sex and age, underscoring the need for individualized dosing.23 In addition, propofol has well-documented misuse and dependence potential among healthcare professionals with access to the drug, and prolonged infusion carries the rare but serious risk of propofol infusion syndrome; both considerations argue strongly against unsupervised or chronic use.

Bionic Sleep Therapy (BST)

BST involves the controlled intravenous administration of propofol combined with dexmedetomidine under monitored conditions to deliberately induce a sustained pharmacological sleep-like state. The entire published evidence base for BST in insomnia currently consists of a single case report: a patient with 21-year refractory insomnia underwent three sessions using target-controlled propofol infusion (1.0–1.5 µg/mL) combined with dexmedetomidine, with reported improvement in objective sleep efficiency from 0% to 51%.12 The proposed multi-target mechanism—simultaneous GABA-A potentiation and alpha-2 agonism to reduce cortical and subcortical hyperarousal, “resetting” of dysregulated sleep–wake rhythms, and modulation of neuroinflammation and oxidative stress—remains speculative.12 We emphasize that BST is a preliminary, hypothesis-generating observation rather than a therapeutic option: it requires anesthetic drug administration with the attendant risks of airway compromise and cardiorespiratory depression, and it cannot be recommended outside a research protocol with specialized monitoring and trained anesthesia personnel.

Interventional and Non-Pharmacological Modalities at the Anesthesia–Sleep Interface

Stellate Ganglion Block for Chronic Insomnia

Stellate ganglion block (SGB)—a minimally invasive procedure involving local anesthetic injection adjacent to the stellate ganglion in the cervical sympathetic chain—has been proposed as an intervention for chronic insomnia, on the rationale that sympatholysis attenuates the hyperarousal state that characterizes the disorder. The stellate ganglion provides sympathetic innervation to the head, neck, and upper extremities; its blockade reduces sympathetic tone both peripherally and, potentially, within central arousal circuits.

The evidence base remains early and heterogeneous. A systematic review of SGB for postoperative sleep disturbance concluded that available studies are small and of variable quality.38 Insomnia-specific data are limited to a handful of clinical studies: a comparative study of SGB with different ropivacaine volumes reported improved sleep quality in patients with insomnia,37 and a recent single-arm prospective pilot study of high-frequency SGB in 72 patients with insomnia disorder found PSQI improvement from 15.35 to 10.59 that exceeded the minimal clinically important difference—yet the authors themselves cautioned that the uncontrolled design cannot exclude placebo effects or regression to the mean and called for sham-controlled trials.36 A systematic review and meta-analysis specifically examining SGB for chronic insomnia in the English-language peer-reviewed literature remains lacking as of 2026, and broader reviews of SGB for psychiatric indications describe a clinical research landscape dominated by small, uncontrolled reports.66 Reported adverse events are generally mild and transient, including Horner’s syndrome and hoarseness, but the procedure carries rare serious risks (vascular injection, seizures, pneumothorax) that must be weighed against the still-unproven benefit.37,66

Two studies have examined SGB combined with dexmedetomidine in surgical populations. Guo et al, in a prospective randomized trial of elderly patients undergoing hip arthroplasty, found that the combination achieved significantly higher Richards-Campbell Sleep Questionnaire scores during the first three postoperative nights than either dexmedetomidine alone or control, alongside superior analgesia; however, the benefit was transient, with no significant PSQI difference at one month, and the single-center design limits generalizability.15 Zhao et al, in a retrospective evaluation of patients undergoing cardiac valve replacement with extracorporeal circulation, reported that the combination attenuated perioperative oxidative stress and reduced postoperative cognitive dysfunction compared with dexmedetomidine alone, findings the authors attributed to hemodynamic stability and suppression of sympathetic excitation.67 Both studies are perioperative, short-term, and—in the latter case—retrospective; they do not establish efficacy for chronic insomnia.

Overall, SGB for insomnia should be regarded as investigational. Given the known potency of expectation and placebo effects in invasive procedures—and the absence to date of any sham-controlled trial—claims of efficacy await adequately powered, sham-controlled RCTs with objective sleep outcomes and extended follow-up.

Acupuncture and Acupoint Stimulation for Perioperative Sleep

Acupuncture and related techniques, including transcutaneous electrical acupoint stimulation (TEAS) and auricular acupressure, have been studied as adjunctive therapies for perioperative sleep disturbance. A systematic review and meta-analysis by Liu et al, encompassing 9 RCTs with 719 patients, reported that acupuncture-point stimulation improved perioperative subjective sleep quality (SMD: −1.36; 95% CI: −1.71 to −1.01; P < 0.00001), increased total sleep time, and improved sleep efficiency.16 Proposed mechanisms include stimulation of A-delta and C fibers, increased endogenous opioid release in the hypothalamus, enhanced GABAergic neurotransmission in the VLPO, modulation of serotonergic and noradrenergic transmission in the raphe nuclei and locus coeruleus, and increased melatonin secretion.16 Functional MRI studies have reported that acupuncture at sleep-related points deactivates the default mode network and limbic system—patterns resembling those observed during natural sleep and dexmedetomidine sedation.

However, the acupuncture literature is particularly vulnerable to placebo and expectancy effects, and sham-controlled evidence in insomnia is sobering: in a randomized sham-controlled trial of acupuncture for chemotherapy-associated insomnia in breast cancer, acupuncture was not superior to sham acupuncture on the primary Insomnia Severity Index outcome (between-group difference −0.4, P = 0.609), with both arms improving substantially.39 This result illustrates that much of the observed benefit in uncontrolled or non-sham-controlled acupuncture studies may reflect non-specific contextual effects rather than point-specific physiological action. It does not necessarily negate clinical utility—a reproducible placebo-dominated response may still benefit some patients—but it changes the evidentiary status of the intervention and argues for rigorous blinding in future trials. Most perioperative meta-analytic data also derive from subjective outcomes in surgical populations rather than from objective polysomnography in insomnia patients.

Transcutaneous electrical acupoint stimulation (TEAS) offers a non-invasive alternative that can be administered by nursing staff without specialized acupuncture training. An RCT reported that TEAS at Shenmen (HT7) and Sanyinjiao (SP6) for 30 minutes before bedtime on three consecutive postoperative nights improved sleep quality, reduced pain scores, and decreased analgesic consumption in patients undergoing total knee arthroplasty.16 Pending sham-controlled replication, TEAS is best viewed as a low-risk component of multimodal perioperative sleep management rather than a validated insomnia treatment.

Closed-Loop Acoustic Stimulation and Neuromodulation

Closed-loop acoustic stimulation (CLAS) enhances slow-wave activity by delivering auditory stimuli phase-locked to EEG slow oscillations. When combined with dexmedetomidine sedation in healthy young adults, CLAS altered subsequent slow-wave sleep homeostasis—reducing N3 duration and slow-wave pressure dissipation on the post-sedation night—suggesting that the paired intervention discharges part of the homeostatic NREM need, but with the same caveat regarding potential displacement of nocturnal deep sleep discussed in Dexmedetomidine: Biomimetic Sleep and Novel Delivery Modalities.28 This technology bridges anesthetic sedation with sleep engineering and represents a non-pharmacological augmentation of biomimetic sleep, but it remains experimental, and no insomnia-population trials have been reported.

Repetitive transcranial magnetic stimulation (rTMS) combined with dexmedetomidine has been examined in chronic insomnia disorder. Zheng et al conducted an RCT in 42 participants randomized to dexmedetomidine alone (n = 23) or rTMS (1 Hz over the left dorsolateral prefrontal cortex, 100% resting motor threshold, 4 weeks) plus dexmedetomidine (n = 19).40 The combination group showed significantly greater decreases in PSQI, Hamilton Anxiety Rating Scale, and Hamilton Depression Rating Scale scores at treatment termination and at 1-, 3-, and 6-month follow-up (all P < 0.01), with reduced light sleep and awakening and increased deep and REM sleep; IL-6 and TNF-α concentrations were lower, and glutathione peroxidase and superoxide dismutase levels higher, than with dexmedetomidine alone (P < 0.05), suggesting anti-inflammatory and antioxidant mechanisms.40 The authors also reported that rTMS permitted gradual reduction of dexmedetomidine dosage by the end of treatment.40 These results are promising but derive from a single small trial; independent replication with larger samples and objective sleep measurement is needed before the combination can be considered beyond a research setting.

Perioperative Sleep Optimization Strategies

Impact of Surgery and Anesthesia on Postoperative Sleep

Postoperative sleep disturbance (PSD) is highly prevalent, affecting up to 69.4% of surgical patients according to a recent systematic review and meta-analysis,3 and has implications for recovery, pain perception, cognitive function, and patient satisfaction. PSD manifests across all stages of surgical care: preoperative anxiety-related insomnia, intraoperative disruption of sleep architecture by anesthetic agents, and postoperative sleep fragmentation due to pain, environmental factors, medications, and altered circadian rhythms.68

Anesthetic agents exert specific effects on sleep architecture that persist into the postoperative period.68 Most anesthetics reduce REM and slow-wave sleep in the immediate recovery period, with effects lasting from hours to days depending on agent, dose, and duration of administration. Opioids, frequently used for postoperative analgesia, further compound this disruption by suppressing REM sleep and increasing arousal frequency.69,70 These alterations can impair cognitive and physical recovery, increase pain sensitivity through sleep-loss-induced hyperalgesia, and contribute to postoperative delirium—particularly in elderly patients.71

Dexmedetomidine for Perioperative Sleep Enhancement

Dexmedetomidine has been studied extensively for perioperative sleep optimization. Multiple RCTs and a meta-analysis of randomized trials indicate that perioperative dexmedetomidine administration improves postoperative sleep quality across surgical populations.41 In cardiac surgical patients, low-dose dexmedetomidine infusion (0.2–0.4 µg/kg/h) improved sleep quality after ultrafast-track extubation.42 In cesarean section patients, intraoperative dexmedetomidine during combined spinal-epidural anesthesia was associated with improved postoperative sleep quality (PSQI: 4.88 ± 1.15 versus 6.42 ± 1.21 on postoperative day 1, P < 0.001) and reduced anxiety across three postoperative days.43

Proposed mechanisms include: (1) production of biomimetic sleep with preserved N2 architecture and enhanced slow-wave activity; (2) reduction of sympathetic tone and stress response; (3) opioid-sparing analgesia that minimizes opioid-related sleep disruption; (4) anxiolytic effects reducing preoperative anxiety-related insomnia; and (5) potential enhancement of glymphatic clearance, although the clinical relevance of the last remains speculative (Neuroinflammatory, Glymphatic, and Neuroplastic Mechanisms).

Multimodal Anesthesia and Opioid-Sparing Strategies

Multimodal anesthesia incorporating opioid-sparing strategies represents a fundamental approach to perioperative sleep preservation. By minimizing opioid exposure and employing diverse analgesic modalities, these strategies protect natural sleep architecture while maintaining effective pain control.69,70 Recommended components include acetaminophen, non-steroidal anti-inflammatory drugs (NSAIDs), gabapentinoids, regional anesthesia techniques (peripheral nerve blocks, neuraxial anesthesia, fascial plane blocks), ketamine infusions, and non-pharmacological interventions. Regional anesthesia provides excellent analgesia while avoiding systemic opioid effects on sleep.

The Society of Critical Care Medicine (SCCM) 2025 focused update to the PADIS guidelines, led by Lewis et al, emphasizes multimodal analgesia with opioid-sparing strategies for intensive care unit (ICU) patients and conditionally recommends dexmedetomidine over propofol for sedation in mechanically ventilated adults when light sedation and/or delirium reduction are the highest priorities.44 Benzodiazepines are associated with disrupted sleep architecture, delirium, and prolonged mechanical ventilation, and current guidelines recommend against routine benzodiazepine premedication.44

Melatonin, Circadian Interventions, and ICU Sleep Management

Melatonin, the primary hormone regulating circadian rhythm, plays a critical role in sleep–wake regulation. Anesthesia and surgery disrupt endogenous melatonin secretion through surgical stress-induced cortisol release, light exposure in operating rooms and ICUs, and direct effects of anesthetic agents on pineal function.72–74 Exogenous melatonin has been shown to improve postoperative sleep quality; 6 mg taken before bedtime increased total sleep time and reduced sleep latency after laparoscopic cholecystectomy and breast cancer surgery.75,76

Non-pharmacological circadian interventions are also effective: noise reduction, light control (eye masks, dim lighting at night), earplugs, and daytime natural light exposure improve postoperative sleep quality.77 In the ICU, the ABCDEF bundle (Assess, prevent, and manage pain; Both spontaneous awakening and breathing trials; Choice of analgesia and sedation; Delirium assessment, prevention, and management; Early mobility and exercise; Family engagement) serves as an integrated management framework.44 Dexmedetomidine-based sedation is associated with more favorable sleep architecture than propofol and better outcomes than benzodiazepines in critically ill patients.44 Sleep disruption in the ICU affects up to 80% of patients and contributes to delirium, prolonged mechanical ventilation, and impaired functional recovery.78,79

From Perioperative Evidence to Chronic Insomnia: Limits of Extrapolation

Much of the RCT-grade evidence cited in this review—including most dexmedetomidine, esketamine, acupuncture, and SGB trials—derives from perioperative populations. This evidence is relevant but not directly transferable to primary chronic insomnia, for at least three reasons. First, the etiologies differ fundamentally: perioperative sleep disturbance is driven by surgical trauma, anesthetic residue, pain, environmental noise, and acute circadian disruption, and typically resolves over days to weeks, whereas primary chronic insomnia is maintained by hyperarousal, conditioned wakefulness, and circadian dysregulation over months to years. An intervention that restores sleep acutely after surgery need not modify the chronic maintaining mechanisms of insomnia disorder. Second, perioperative trials usually measure short-term outcomes (nights 1–3, occasionally one month), providing no information on the durability required of an insomnia treatment. Third, perioperative studies enroll patients selected for surgery rather than for insomnia phenotype, and concomitant analgesic, anxiolytic, and anesthetic exposures confound attribution. We therefore treat perioperative evidence as proof of principle and hypothesis-generating for chronic insomnia applications, and we have flagged perioperative provenance wherever it supports a therapeutic claim (Table 1).

Matching Interventions to Insomnia Phenotypes

Insomnia is not a unitary condition. Clinically, patients present with sleep-onset, sleep-maintenance, or early-morning-awakening phenotypes (frequently mixed), and mechanistically oriented research distinguishes endotypes such as hyperarousal-predominant, emotion-linked (comorbid depression/anxiety), and somatic-comorbidity-associated insomnia. The interventions reviewed here differ in mechanism and pharmacokinetics in ways that plausibly map onto these phenotypes—although we emphasize that no anesthesia-informed intervention has yet been prospectively tested in a phenotype-stratified design, and the matching proposed below is hypothesis-generating rather than evidence-based (Table 2).

Table 2.

Hypothesis-Generating Matching of Investigational Approaches to Insomnia Phenotypes and Endotypes (No Approach Has yet Been Prospectively Validated in a Phenotype-Stratified Trial)

Phenotype/Endotype	Typical Features	Mechanistically Aligned Investigational Approach	Rationale	Evidence Status and Caveats
Sleep-onset predominant	Prolonged sleep-onset latency; preserved maintenance	Intranasal or oromucosal low-dose dexmedetomidine13,27	Gradual sleep initiation without sustained next-morning sedation	Acute PSG benefit shown in non-insomniac volunteers only; insomnia RCTs lacking
Sleep-maintenance, hyperarousal endotype	High wake after sleep onset; short objective sleep; sympathetic predominance	Low-dose dexmedetomidine regimens; CLAS;28 SGB36	Slow-wave enhancement and sympatholysis target hyperarousal	Preliminary; sham and expectancy effects untested; displacement of nocturnal deep sleep possible28
Early-morning awakening	Terminal insomnia; often older age	No matched approach; caution advised	—	Age-related pharmacokinetic sensitivity raises adverse-event risk23
Emotion-linked (comorbid depression/anxiety)	Insomnia with mood disorder	Ketamine/esketamine adjunct; esketamine + PCSL29,33	Mood–sleep coupling; BDNF-related plasticity	Sleep benefit not separable from antidepressant effect; misuse and urinary risks22
Somatic-comorbidity-associated (incl. COMISA)	OSA features, pain, medical illness	Treat underlying condition first (eg, OSA therapy); opioid-sparing multimodal analgesia perioperatively69,70	Sedatives may worsen undiagnosed OSA19	OSA screening is a prerequisite before any anesthetic-based trial
Perioperative sleep disturbance	Acute, surgery-related, self-limiting	Dexmedetomidine; TEAS; melatonin; multimodal strategies16,41,75,76	Established in the perioperative context	Applicable now for surgical patients; not transferable to chronic insomnia (From Perioperative Evidence to Chronic Insomnia: Limits of Extrapolation)

For sleep-onset-predominant insomnia, agents with rapid but gentle onset and short duration—intranasal or oromucosal dexmedetomidine—are mechanistically attractive because they facilitate sleep initiation without sustained next-morning sedation.13,27 For sleep-maintenance-predominant insomnia with objective short sleep duration and a hyperarousal profile, interventions that enhance slow-wave sleep or blunt sympathetic tone (low-dose dexmedetomidine regimens, CLAS, SGB) align with the presumed endotype, albeit with only preliminary support.28,36 For emotion-linked insomnia with comorbid depression, ketamine/esketamine-based approaches have the most supporting data—but, as discussed in Ketamine and Esketamine: NMDA Receptor Antagonism and Sleep Modulation, their sleep benefits cannot currently be separated from antidepressant effects, and polypharmacy in these populations further obscures attribution.32,33 For somatic-comorbidity-associated insomnia, and in particular for the large subgroup with comorbid OSA (COMISA), sedative anesthetic agents may be harmful rather than helpful: undiagnosed OSA is common in refractory insomnia,19 and airway and ventilatory effects of alpha-2 agonists, GABAergic agents, and NMDA antagonists must be considered before any trial of these agents. Finally, early-morning-awakening phenotypes, and older patients in general, warrant particular caution because age-related pharmacokinetic changes prolong drug effects and increase sensitivity to hemodynamic adverse events.23

Prospective validation of phenotype-guided selection—using clinical subtyping combined with objective measures such as polysomnography-derived sleep depth, quantitative EEG arousal indices, and eventually pharmacogenomic markers49,80—is a priority for the field and a prerequisite for any credible claim of “precision” insomnia therapy.

Safety Considerations for Repeated or Chronic Anesthetic-Based Interventions

Because the therapeutic concepts reviewed here imply repeated or chronic exposure to agents that were developed and approved for acute perioperative or intensive care use, safety deserves treatment as a first-order issue rather than an afterthought.

Cardiovascular Effects and Withdrawal

Dexmedetomidine predictably causes dose-dependent bradycardia and hypotension, and hypertension can occur with loading doses; these effects are most pronounced with intravenous bolus administration and are attenuated—but not eliminated—by intranasal or oromucosal routes.13,27 Of particular concern for any chronic regimen, withdrawal phenomena after prolonged dexmedetomidine exposure are well documented in critical care: in adult ICU cohorts, clinically significant withdrawal after infusions of 72 hours or longer occurred in roughly one-third of patients, with rebound hypertension predominating,20 and a pediatric systematic review and meta-analysis similarly found that more than one-third of children developed hypertension or tachycardia after cessation.21 Although insomnia regimens use lower doses, sympathetic rebound after repeated nightly dosing has not been systematically studied and cannot be assumed to be absent. Chronic suppression of noradrenergic tone also carries theoretical risks of tolerance and altered stress responsiveness. Any clinical development program for dexmedetomidine-based insomnia therapy should therefore incorporate predefined hemodynamic monitoring, tapering protocols, and withdrawal surveillance.

Route- and Device-Specific Risks

Repeated intranasal administration raises the possibility of local nasal mucosal toxicity, which has not been systematically evaluated for dexmedetomidine in chronic use; pediatric meta-analytic data for intranasal dexmedetomidine indicate low but non-trivial rates of hypotension and bradycardia even with single perioperative doses. Patient-controlled intravenous systems add device-related risks—infection, thrombosis, line displacement, pump malfunction, and misuse or diversion—that are unfamiliar in insomnia care and would require governance frameworks comparable to those for ambulatory infusion therapies.11

Agent-Specific Chronic Risks

Propofol carries misuse and dependence potential, and chronic or high-dose exposure risks propofol infusion syndrome; its use outside monitored settings is unacceptable Ketamine and esketamine are associated with lower urinary tract toxicity (including ulcerative cystitis) and misuse liability with repeated exposure; a recent systematic appraisal concluded that the addiction risk in therapeutic settings is real, albeit incompletely quantified, and requires active management.22 Long-term esketamine safety data exist only for intermittent use in depression under strict risk-management programs.35

Respiratory Risk and OSA Screening

All sedative anesthetic agents can reduce upper airway tone or blunt ventilatory drive. Given the high prevalence of undiagnosed OSA among patients with refractory insomnia,19 we consider OSA screening (eg, validated questionnaires followed by sleep testing when indicated) an ethical prerequisite for enrollment in any trial or clinical program of anesthetic-based insomnia therapy. This single step would both protect patients and prevent confounding of trial results by unrecognized COMISA.

Population-Specific Considerations

Sex- and age-related differences in anesthetic pharmacology are clinically material: propofol requirements are higher in younger women and decline with age, with corresponding differences in recovery time,23 and older patients are more susceptible to hypotension, bradycardia, and delirium. Pediatric and pregnant populations lack any supporting evidence for these indications. Dosing paradigms developed in undifferentiated trial populations should therefore not be generalized without stratified study.

Monitoring and Governance

Finally, implementation barriers are non-trivial: continuous pulse oximetry and blood pressure monitoring, trained personnel, emergency airway capability, regulatory status (all such uses are off-label), and cost all constrain translation. These requirements currently confine legitimate use of intravenous anesthetic-based sleep interventions to research protocols in specialized centers.

Positioning Within the Current Insomnia Treatment Landscape

For anesthesia-informed approaches to be clinically useful, their intended place in the treatment pathway must be stated plainly. CBT-I remains the universally recommended first-line treatment for chronic insomnia, with durable benefits and minimal risk.6 Approved pharmacotherapies—including DORAs, whose efficacy has been quantified in network meta-analysis9—constitute the evidence-based second pillar, with guideline support for defined durations of use.6

Against this benchmark, none of the interventions reviewed here can currently be recommended as first-line, and none should be represented as an alternative to CBT-I or approved hypnotics. A defensible positioning, consistent with the evidence graded in Table 1, is as follows. Dexmedetomidine-based approaches (intranasal/oromucosal routes and PCSL) are investigational options for rigorously defined refractory insomnia—failure of CBT-I and at least two pharmacological classes—within research protocols or specialized multidisciplinary programs, after OSA has been excluded.11,27 Ketamine/esketamine-based approaches are relevant primarily to insomnia comorbid with treatment-resistant depression, where they should be regarded as adjuncts to antidepressant care rather than hypnotics per se.32,33 BST is hypothesis-generating only.12 SGB, acupuncture/TEAS, CLAS, and rTMS combinations are investigational adjuncts, with perioperative TEAS/acupuncture having the strongest (though still subjective-outcome-based) support for procedure-related sleep disturbance.16 Perioperative sleep-preserving strategies (Perioperative Sleep Optimization Strategies), by contrast, are applicable now within enhanced-recovery and ICU practice, supported by RCTs and guidelines,41,44 and should not be conflated with chronic insomnia treatment.

A pragmatic research agenda follows directly: head-to-head trials of the most mature anesthesia-informed candidates (eg, intranasal/oromucosal dexmedetomidine) against CBT-I, DORAs, or both; phenotyped enrollment (Matching Interventions to Insomnia Phenotypes); objective polysomnographic and patient-reported outcomes; and protocolized long-term safety follow-up. Until such trials exist, these approaches belong in the research setting.

Future Directions

Precision Approaches and Pharmacogenomics

The field is moving toward approaches that tailor therapy to individual patient characteristics. Pharmacogenomic factors influencing response to dexmedetomidine (CYP2A6 polymorphisms affecting metabolism), ketamine (BDNF Val66Met genotype affecting neuroplasticity response), and propofol (CYP2B6 and UGT1A9 variants) may eventually enable individualized dosing and agent selection,80 although none of these markers is yet clinically actionable for insomnia. Combined effects of HCRTR1/2 gene variants and non-genetic factors significantly influence sleep–wake transitions and hemodynamic stability during propofol, dexmedetomidine, and remifentanil anesthesia, providing proof of concept that genetic variation in the orexin system modulates anesthetic response.49 A case report identified a rare SCN9A variant (rs16851751) associated with profound post-procedural sedation and apnea following dexmedetomidine–propofol co-administration, highlighting the potential clinical relevance of ion channel pharmacogenetics.81 Biomarkers including BDNF levels, inflammatory cytokine profiles, and quantitative EEG signatures may predict treatment response, and combinatorial approaches pairing pharmacological agents with CBT-I, rTMS, or sleep-hygiene optimization may exceed monotherapy outcomes—hypotheses that all require prospective validation.

Artificial Intelligence and Closed-Loop Systems

Artificial intelligence (AI) and machine learning are increasingly applied at the intersection of anesthesia and sleep medicine. Machine learning algorithms trained on intraoperative physiological data—including EEG-derived indices, heart rate variability, and blood pressure waveforms—have been reported to predict postoperative sleep disturbance with accuracies exceeding 80% in derivation cohorts, and deep learning models applied to preoperative polysomnography have identified patients at risk for adverse anesthetic outcomes; external validation in prospective cohorts remains limited. Closed-loop anesthesia delivery systems use real-time EEG feedback to adjust drug infusion rates, and closed-loop acoustic stimulation can enhance slow-wave activity during dexmedetomidine sedation,28 offering a potential amplifier of biomimetic sleep—subject to the homeostatic-displacement caveat discussed above. Integration of AI with wearable sleep monitoring may eventually enable continuous ambulatory assessment of perioperative sleep architecture.

Novel Drug Delivery and Digital Health

Potential future developments include transdermal dexmedetomidine patches for sustained overnight delivery, implantable or programmable delivery devices with physiological feedback control, and closed-loop systems that adjust dosing from real-time EEG. Such technologies could, if proven safe, move anesthetic-based insomnia therapy from hospital to outpatient settings; none currently exists beyond conceptual or early developmental stages. Wearable sleep trackers, smartphone-based sleep diaries, and digital therapeutic platforms (including prescription digital therapeutics delivering CBT-I, such as Somryst) can already facilitate monitoring of treatment response and delivery of behavioral interventions,82 and represent the most immediately deployable bridge between anesthetic pharmacology and scalable insomnia care.

Addressing Critical Knowledge Gaps

Several critical knowledge gaps require well-designed clinical trials: (1) long-term safety data for repeated dexmedetomidine administration, including cardiovascular effects, withdrawal phenomena, nasal tolerability, and tolerance; (2) adequately powered RCTs comparing anesthetic-based insomnia therapies against CBT-I and DORAs, with objective sleep outcomes; (3) prospectively validated patient-selection criteria based on insomnia phenotype, comorbidity (particularly OSA), and biomarkers; (4) durability of treatment effects and optimal maintenance and tapering strategies; (5) clarification of whether sleep improvements with ketamine/esketamine reflect direct hypnotic effects or antidepressant mediation, using insomnia-specific populations; (6) sham-controlled RCTs of SGB and acupuncture in insomnia-specific populations with polysomnographic endpoints, given the demonstrated potency of placebo effects in this space;36,39 (7) prospective validation of pharmacogenomic-guided personalization; and (8) cost-effectiveness analyses comparing these approaches with conventional therapy to inform policy and reimbursement.

Limitations

This review has limitations that qualify its conclusions. First, as a narrative rather than systematic review, it is susceptible to selection and citation bias despite the transparent search procedures described in Review Methodology; no study flow diagram, formal risk-of-bias scoring, or quantitative synthesis was undertaken, and publication bias favoring positive results cannot be excluded—particularly for SGB, acupuncture, and device-based interventions where small uncontrolled studies predominate. Second, much of the evidence derives from perioperative, depressed, or critically ill populations rather than from patients with primary chronic insomnia, limiting generalizability (From Perioperative Evidence to Chronic Insomnia: Limits of Extrapolation). Third, heterogeneity of interventions, dosing regimens, comparators, and outcome measures precludes pooled effect estimates, and no head-to-head trials against CBT-I or approved hypnotics exist. Fourth, long-term safety data for repeated anesthetic-agent exposure in insomnia are almost entirely absent; safety inferences drawn from critical care and depression populations may not transfer. Fifth, several mechanistic discussions (glymphatic clearance, gut–brain axis, pharmacogenomics, AI-guided systems) rely on indirect, preclinical, or preliminary evidence and should be read as hypotheses. Finally, some cited sources were not obtainable in full text or remain available only as preprints or in-press records at the time of writing; these are identified in the reference list and should be interpreted with corresponding caution.

Conclusion

What is currently known. Anesthesia and sleep share defined neural circuits, and among anesthetic agents dexmedetomidine most closely reproduces the electrophysiology of natural NREM sleep in humans. Perioperative administration of dexmedetomidine, opioid-sparing multimodal analgesia, and circadian interventions improves postoperative sleep in randomized trials and is supported by critical care guidelines. These findings establish proof of mechanism and a legitimate research direction.

What remains unproven. No anesthesia-informed intervention has demonstrated efficacy for primary chronic insomnia in an adequately powered, insomnia-specific RCT. Evidence for PCSL, intranasal dexmedetomidine in non-surgical patients, ketamine/esketamine as hypnotics, BST, SGB, and CLAS is preliminary, derives from small, single-group, perioperative, or comorbid-depression populations, and—in the case of invasive and device-based procedures—is vulnerable to placebo effects. Long-term safety of repeated anesthetic exposure, including withdrawal, cardiovascular, urinary, and device-related risks, is undemonstrated, and benefits cannot yet be weighed against those of CBT-I and approved pharmacotherapies, which remain the standard of care.

Accordingly, anesthesia-informed approaches should be regarded as investigational adjuncts whose rational development will require phenotype-stratified, sham- or active-controlled trials with objective outcomes and protocolized safety surveillance. Whether the mechanistic elegance of biomimetic sleep can be converted into durable clinical benefit for patients with insomnia is the central open question for the field.

Funding Statement

No external funding was received for the preparation of this article.

Data Sharing Statement

This article is a review and does not report or analyze original data.

Author Contributions

Yu Wu: Conceptualization, Investigation, Writing – original draft, Writing – review & editing.

Xiongwei Yu: Investigation, Methodology, Funding acquisition, Project administration, Writing – review & editing.

Houxin Sun: Investigation, Visualization, Writing – review & editing.

Fang Huang: Validation, Writing – review & editing.

Jinbao Wang: Conceptualization, Supervision, Funding acquisition, Project administration, Writing – review & editing.

All authors took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this article.