The preceding four modules of this chapter examined the major sedative-hypnotic drug classes in sequence: benzodiazepines, non-benzodiazepine hypnotics, barbiturates and anesthetic sedatives, and the toxicology and clinical management of the class as a whole. This integrative module serves a different purpose: to provide the neurobiological and comparative framework that unifies the pharmacology of all these agents. Understanding why different drugs produce different clinical effects requires a firm grasp of the neuroscience of sleep itself — the circuits that generate it, the homeostatic and circadian forces that regulate it, and the specific mechanisms by which pharmacological agents interact with or disrupt normal sleep architecture.
This module covers the neurobiology of sleep including sleep stage architecture, the two-process model of sleep regulation, circadian timing mechanisms, and the orexin system as a master switch of arousal; the class-by-class effects of sedative-hypnotic drugs on sleep stage composition; a systematic comparative pharmacology table spanning all major drug classes; a diagnostic and treatment algorithm for insomnia disorder grounded in current guideline recommendations; the clinical positioning of benzodiazepines within the broader landscape of anxiety disorder pharmacotherapy; and an overview of emerging and investigational agents that represent the next generation of sleep and sedation pharmacology. Together these sections provide the integrative foundation needed to generate and answer the higher-order clinical questions that define EXPANDED tier question bank material.
Normal human sleep is not a uniform state but a dynamic cycle of physiologically distinct stages that repeat in an approximately 90-minute ultradian rhythm throughout the night. Sleep is broadly divided into non-rapid eye movement (NREM) sleep and rapid eye movement (REM) sleep. NREM sleep is itself subdivided into three stages based on electroencephalographic (EEG) characteristics: N1 (light sleep, theta waves 4–8 Hz, constituting approximately 5% of total sleep time), N2 (intermediate sleep, characterized by sleep spindles at 12–15 Hz and K-complexes, constituting approximately 45–55% of total sleep time), and N3 (slow-wave sleep or deep sleep, characterized by high-amplitude delta waves at 0.5–4 Hz, constituting approximately 15–20% of total sleep time in young adults and declining progressively with age).1 REM sleep is characterized by desynchronized low-amplitude mixed-frequency EEG activity resembling wakefulness, rapid conjugate eye movements, skeletal muscle atonia (mediated by active inhibition of spinal motor neurons), and vivid dreaming; it constitutes approximately 20–25% of total sleep time and occurs predominantly in the latter half of the night as sleep cycles progress.
The functional significance of sleep stages is not merely descriptive. N3 slow-wave sleep is the stage during which growth hormone is predominantly secreted, immune consolidation occurs, and declarative memory (facts and events) is consolidated through hippocampal-neocortical dialogue involving sharp-wave ripples and slow oscillations. REM sleep serves distinct functions including procedural and emotional memory consolidation, threat simulation and emotional processing, and synaptic homeostasis. The clinical implication is that pharmacological disruption of specific sleep stages has functional consequences beyond simply altering subjective sleep quality — N3 suppression impairs physical recovery and memory consolidation, and REM suppression disrupts emotional regulation and contributes to the REM rebound phenomenon (intense, often disturbing dreams) seen after discontinuation of REM-suppressing agents.1
The two-process model of sleep regulation, proposed by Borbely and colleagues, provides the most clinically useful framework for understanding both normal sleep and the mechanism of action of hypnotic drugs. The model posits that sleep timing and depth are governed by the interaction of two independent processes: Process S (homeostatic sleep pressure) and Process C (circadian alerting signal).2
Process S represents the accumulation of sleep pressure during wakefulness, driven primarily by the progressive accumulation of adenosine in the basal forebrain and other sleep-promoting regions. Adenosine is a byproduct of neuronal energy metabolism, and its extracellular concentration rises monotonically during waking in proportion to prior wake duration. Adenosine acts on A1 and A2A receptors in sleep-promoting neurons of the ventrolateral preoptic area (VLPO) to increase NREM sleep propensity, and on arousal-promoting neurons (orexinergic, noradrenergic, histaminergic) to inhibit wakefulness. Caffeine's mechanism of action is competitive antagonism of adenosine receptors, directly opposing this homeostatic sleep drive. During sleep, adenosine is cleared and Process S decays exponentially, reaching its nadir near the end of the sleep period. The depth of N3 slow-wave sleep at sleep onset is directly proportional to the magnitude of Process S — the longer and more intensely one has been awake, the more deep sleep is generated in the first sleep cycle.2
Process C represents the circadian alerting signal generated by the suprachiasmatic nucleus (SCN) of the hypothalamus, the brain's master circadian clock. The SCN generates a near-24-hour oscillation in arousal promotion that rises throughout the day, peaks in the early evening (opposing the rising sleep pressure of Process S to maintain wakefulness), and then falls sharply in the evening — the circadian gate opening for sleep. The timing of Process C is entrained to the external light-dark cycle via retinal photoreceptors (particularly melanopsin-containing intrinsically photosensitive retinal ganglion cells) projecting to the SCN via the retinohypothalamic tract. Light exposure in the evening delays the circadian phase (later sleep timing) while morning light advances it (earlier sleep timing) — the basis of light therapy for circadian sleep disorders. The circadian signal is communicated to the periphery primarily through the SCN's regulation of melatonin secretion from the pineal gland, which serves as a hormonal signal of darkness and contributes to the circadian gate for sleep onset.3
The orexin (hypocretin) system, described in detail in CNS-02, provides a third regulatory layer that stabilizes the boundary between wakefulness and sleep. Orexinergic neurons of the lateral hypothalamus are active during wakefulness and provide excitatory drive to all major monoaminergic and cholinergic arousal nuclei — the locus coeruleus (norepinephrine), dorsal raphe (serotonin), tuberomammillary nucleus (histamine), and basal forebrain (acetylcholine). This tonic excitatory drive prevents the intrusion of sleep into wakefulness and maintains consolidated arousal. Orexin neuronal activity is in turn inhibited by the VLPO sleep-promoting neurons during sleep, creating a mutually inhibitory flip-flop switch that generates rapid, stable transitions between sleep and wakefulness rather than gradual drifts.4 The loss of orexinergic neurons in narcolepsy type 1 destabilizes this switch, producing pathological intrusions of sleep and REM physiology (cataplexy, sleep paralysis, hypnagogic hallucinations) into wakefulness. Dual orexin receptor antagonists (DORAs) therapeutically reduce orexin-mediated wake drive, facilitating the flip-flop switch to transition to sleep without the broad GABAergic suppression of traditional hypnotics.
The VLPO is the primary sleep-promoting nucleus, containing GABAergic and galaninergic neurons that inhibit arousal nuclei during sleep. VLPO activity is itself regulated by adenosine (activating), circadian signals, and temperature (mild cooling promotes VLPO activity and sleep onset, which explains the role of thermoregulation in sleep hygiene). Damage to the VLPO in animal models produces profound insomnia; in humans, inflammatory or structural damage to sleep-promoting circuits (as in fatal familial insomnia, a prion disease affecting the thalamus and hypothalamus) can produce complete, ultimately fatal loss of sleep. The understanding that specific circuits generate sleep has enabled the rational development of pharmacological agents targeted to the specific nodes of the sleep-wake switch, including the orexin receptor antagonists that act at the flip-flop switch's wake-promoting arm rather than globally enhancing GABAergic inhibition.1
Every sedative-hypnotic drug class alters sleep architecture in ways that follow directly from its mechanism of action. Understanding these class-specific effects is clinically important for two reasons: it explains why patients on different agents report different sleep quality despite similar total sleep times, and it provides the mechanistic rationale for agent selection when sleep quality — not just sleep quantity — is a therapeutic priority.
Benzodiazepines produce the most pronounced pharmacological disruption of normal sleep architecture among commonly used hypnotics. Through non-selective GABA-A receptor (GABA-A) potentiation at α1, α2, α3, and α5 subunit-containing receptors, they reliably suppress N3 slow-wave sleep — the most physically restorative stage — and suppress rapid eye movement (REM) sleep, reducing both REM duration and dream intensity. The net effect is that despite increasing total sleep time and reducing sleep onset latency, benzodiazepine-induced sleep is pharmacologically shallow, N3-depleted, and REM-reduced. Sleep spindle activity in N2 is typically increased, which may account for the characteristic "benzodiazepine sleep" on polysomnography — abundant spindle-rich N2 that registers as sleep but lacks the restorative properties of N3. Patients on chronic benzodiazepines frequently report feeling unrefreshed despite prolonged sleep time, a direct clinical consequence of N3 suppression. Upon discontinuation, rebound increases in both N3 and REM sleep are characteristic, with REM rebound producing intensely vivid and often disturbing dreams that contribute to rebound insomnia and reinforce continued use.5
Z-drugs produce sleep architecture effects that are qualitatively similar to but quantitatively less pronounced than benzodiazepines, particularly at lower doses where α1 selectivity is most expressed. At standard therapeutic doses, zolpidem produces minimal suppression of N3 slow-wave sleep, a meaningful distinction from benzodiazepines driven by its relative preference for α1-containing GABA-A receptors (which mediate sedation) over α2/α3 receptors (which mediate anxiolysis and muscle relaxation and are expressed in brain regions that regulate deeper sleep stages). REM sleep is generally preserved or only modestly reduced. This architecture advantage diminishes at higher doses, with extended-release formulations, and in elderly patients who are more sensitive to α1-mediated suppression of slow-wave sleep at any dose. Eszopiclone, with less α1 selectivity than zolpidem, produces more N3 suppression than zolpidem at equivalent sedating doses. The practical implication is that low-dose immediate-release zolpidem represents the least architecturally disruptive GABA-active hypnotic option, though the advantage is dose- and formulation-dependent rather than absolute.6
Ramelteon and tasimelteon produce no significant disruption of sleep architecture at therapeutic doses. Because their mechanism involves circadian phase-setting and facilitation of sleep onset through melatonin receptor type 1 (MT1)/melatonin receptor type 2 (MT2) receptor agonism at the suprachiasmatic nucleus (SCN) rather than direct neuronal inhibition, they do not suppress N3 or REM sleep. Polysomnographic studies confirm preservation of normal sleep stage distribution with ramelteon, with the primary effect being modest reduction in sleep onset latency (10–20 minutes) without alteration in subsequent sleep stages.3 The pharmacological trade-off is that this architecture-preserving profile comes with substantially weaker hypnotic efficacy than GABA-active agents — appropriate for mild to moderate sleep-onset insomnia and circadian rhythm disorders, but insufficient for patients with severe insomnia or significant sleep maintenance difficulties.
DORAs represent the most favorable sleep architecture profile of any pharmacologically active hypnotic class. By selectively reducing orexin-mediated wake drive at the flip-flop switch, suvorexant and lemborexant facilitate the transition from wakefulness to sleep without imposing pharmacological changes on the intrinsic sleep-generating machinery. Multiple polysomnographic studies confirm that DORAs preserve N3 slow-wave sleep and may modestly increase REM sleep — the latter consistent with a reduction in orexin-mediated REM suppression that normally reinforces stable non-rapid eye movement (NREM) sleep in the early part of the night.7 The clinical consequence is that dual orexin receptor antagonist (DORA)-induced sleep most closely resembles the sleep stage composition of natural, unmedicated sleep among all available pharmacological hypnotics. For patients in whom sleep quality and restorative function are primary treatment goals — athletes, cognitively demanding occupations, patients with conditions where N3 or REM sleep has particular therapeutic importance — DORAs represent the pharmacologically superior choice.
Dexmedetomidine's α2-adrenergic agonism at the locus coeruleus produces a neurobiological state that more closely resembles N2 natural sleep than any other IV sedative. EEG studies during dexmedetomidine sedation demonstrate spontaneous sleep spindles and slow oscillations consistent with NREM sleep physiology, in contrast to propofol and benzodiazepine-based sedation which produce EEG patterns that lack these natural markers.8 This unique neurobiological profile — sedation through inhibition of the arousal system rather than activation of the sleep system's inhibitory arm — preserves the capacity for arousal and natural sleep cycling to a greater degree than other IV sedatives. In the ICU, this may partially explain the lower delirium burden observed with dexmedetomidine compared to benzodiazepine infusions, as delirium is strongly associated with circadian disruption and loss of natural sleep architecture during critical illness.
Barbiturates produce the most profound and indiscriminate suppression of all sleep stages, including complete suppression of REM sleep at anesthetic doses. At sedating doses, barbiturates abolish normal sleep architecture and produce a pharmacological coma with EEG burst-suppression at high doses — the mechanistic basis of barbiturate coma for refractory intracranial hypertension. Propofol-induced sedation generates EEG patterns with some features resembling N2/N3 NREM sleep (slow oscillations, spindle-like activity at certain doses), but lacks the cyclic architecture of natural sleep and completely suppresses REM. Patients maintained on propofol infusions in the ICU accumulate profound REM sleep debt, which may contribute to post-ICU cognitive and emotional sequelae including post-traumatic stress disorder (PTSD), consistent with the known functions of REM sleep in emotional memory processing.1
The following systematic comparison covers all major sedative-hypnotic classes across the dimensions most relevant to clinical prescribing decisions and question bank examination. For each class, the key variables are mechanism and receptor target, primary clinical indication, onset and duration, dependence liability, overdose risk and reversibility, sleep architecture effect, controlled substance status, and defining clinical niche.
Mechanism: Positive allosteric modulation of GABA-A receptors (GABA-A) at the α-γ subunit interface; non-selective across α1, α2, α3, α5 subunits; increases frequency of chloride channel opening. Receptor dependence absolute — no direct channel activation without GABA. Indications: Anxiety disorders, insomnia (short-term), acute seizures and status epilepticus, alcohol withdrawal, procedural sedation, muscle spasm. Onset: Variable by agent and route (IV diazepam/midazolam: seconds to minutes; oral agents: 30–60 minutes). Duration: Highly variable (triazolam t½ 2–4 hours to diazepam active metabolite t½ up to 200 hours). Dependence liability: High, particularly with short-acting high-potency agents; physical dependence within weeks; withdrawal potentially life-threatening. Overdose risk: Moderate as monotherapy (ceiling effect from GABA dependence); dramatically amplified by opioid or alcohol co-ingestion. Reversibility: Full with flumazenil (short-acting antagonist — resedation risk). Sleep architecture: Suppresses N3 and rapid eye movement (REM); increases N2 spindle activity. Controlled status: Schedule IV. Clinical niche: Acute anxiety, seizures, alcohol withdrawal, procedural sedation; second-line for chronic anxiety and insomnia.5,9
Mechanism: Positive allosteric modulation of GABA-A receptors with relative α1 selectivity at standard doses; binds same site as benzodiazepines. Indications: Insomnia (sleep onset for zolpidem/zaleplon; sleep onset and maintenance for eszopiclone). Duration: Ultra-short (zaleplon t½ ~1 hour) to intermediate (eszopiclone t½ ~6 hours). Dependence liability: Lower than benzodiazepines but real; Schedule IV; documented withdrawal and rebound insomnia. Overdose risk: Similar to benzodiazepines — amplified by co-ingestion. Reversibility: Flumazenil reverses. Sleep architecture: Less N3 suppression than benzodiazepines at therapeutic doses; REM largely preserved. Controlled status: Schedule IV. Unique safety concern: Black box warning for complex sleep behaviors (sleep-driving, sleepwalking). Clinical niche: Short-term insomnia pharmacotherapy; zolpidem IR for sleep onset; eszopiclone for mixed onset/maintenance.6
Mechanism: Selective melatonin receptor type 1 (MT1)/melatonin receptor type 2 (MT2) receptor agonism in the suprachiasmatic nucleus (SCN); phase-sets the circadian clock; does not directly activate sleep-generating circuits. Agents: Ramelteon (FDA-approved for sleep-onset insomnia and circadian rhythm disorders in totally blind); tasimelteon (non-24-hour sleep-wake disorder; Smith-Magenis syndrome). Dependence liability: None established; not scheduled. Overdose risk: Minimal — no CNS depression at therapeutic doses. Sleep architecture: No disruption; normal stage distribution preserved. Clinical niche: Sleep-onset insomnia especially when avoidance of scheduled medications is a priority (substance use disorder history, elderly); circadian rhythm disorders. Key limitation: Modest hypnotic efficacy; ineffective for sleep maintenance insomnia.3
Mechanism: Dual orexin receptor type 1 (OX1R)/orexin receptor type 2 (OX2R) competitive antagonism; removes orexin wake-promoting drive without activating inhibitory sleep circuits. Agents: Suvorexant (10–20 mg); lemborexant (5–10 mg). Indications: Both sleep-onset and sleep-maintenance insomnia. Dependence liability: Schedule IV; lower than Z-drugs in preclinical and clinical studies; physical dependence less pronounced. Overdose risk: Low compared to GABA-active agents; no respiratory depression at therapeutic doses; sleep paralysis, cataplexy-like episodes at higher doses. Sleep architecture: Best preservation of normal N3 and REM among active hypnotics; may increase REM. Clinical niche: Sleep-onset and maintenance insomnia with favorable architecture profile; preferred in elderly, OSA (with caution), PTSD-associated insomnia. Unique adverse effects: Cataplexy-like episodes, hypnagogic hallucinations (mechanistically consistent with orexin blockade mimicking narcolepsy physiology).7
Mechanism: GABA-A receptor potentiation at β subunit pore sites, increasing duration of channel opening; direct channel activation at high concentrations (GABA-independent). Current clinical indications: Phenobarbital for neonatal seizures, refractory status epilepticus, alcohol/benzodiazepine withdrawal, resource-limited epilepsy management; pentobarbital for refractory intracranial pressure (ICP) and pediatric procedural sedation; thiopental largely unavailable in the US. Dependence liability: High; severe withdrawal syndrome; physical dependence similar to benzodiazepines. Overdose risk: High; no ceiling effect; no reversal agent; narrow therapeutic index. Sleep architecture: Profound suppression of REM; suppression of N3 at sedating doses; burst-suppression at anesthetic doses. Controlled status: Schedule II (amobarbital, pentobarbital, secobarbital) and Schedule IV (phenobarbital, butabarbital). Clinical niche: Specific anticonvulsant indications; withdrawal management; no role in general insomnia treatment.9
Propofol: GABA-A potentiation; no reversal agent; rapid context-insensitive offset after short infusions; propofol infusion syndrome (PRIS) risk with prolonged high-dose infusions; first-line for procedural sedation and ICU sedation in most patients. Dexmedetomidine: α2-adrenergic agonism; arousable sedation; minimal respiratory depression; bradycardia and hypotension; preferred in ICU patients requiring neurological assessment or at high delirium risk. Ketamine: NMDA antagonism; sympathomimetic; bronchodilator; analgesic at sub-dissociative doses; preferred for procedural sedation in hemodynamically compromised patients and bronchospasm; emergence reactions reduced by midazolam pretreatment; emerging role in treatment-resistant depression (TRD). Etomidate: GABA-A modulation; hemodynamic stability; adrenocortical suppression; preferred for RSI in hemodynamically unstable patients without sepsis contraindication. Remimazolam: Esterase-cleaved ultra-short benzodiazepine; context-insensitive offset; flumazenil reversible; no CYP interactions; procedural sedation.8,10
Low-dose doxepin (3–6 mg): H1 antagonism; FDA-approved for insomnia; not scheduled; no dependence; sleep maintenance indication. Trazodone (50–150 mg): H1 and 5-HT2A serotonin receptor (HT2A) antagonism; most widely prescribed off-label hypnotic; not scheduled; useful in comorbid depression. Mirtazapine (7.5–15 mg): H1 and 5-HT2A antagonism; not scheduled; weight gain limiting; useful when appetite stimulation desired. Quetiapine (25–100 mg): H1 and 5-HT2A antagonism; full antipsychotic adverse effect burden; reserve for patients with concurrent psychiatric indication. Sleep architecture effects: All preserve N3 and REM to a greater degree than GABA-active agents; low-dose doxepin has specific evidence for sleep maintenance without architecture disruption.
Insomnia disorder is defined by the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) and the International Classification of Sleep Disorders (ICSD-3) as dissatisfaction with sleep quality or quantity associated with one or more of the following: difficulty initiating sleep, difficulty maintaining sleep, or early morning awakening with inability to return to sleep. These sleep difficulties must occur despite adequate opportunity for sleep, cause clinically significant distress or impairment in daytime functioning, occur at least 3 nights per week, persist for at least 3 months (chronic insomnia disorder), and not be better explained by another sleep disorder or the physiological effects of a substance.9 Acute insomnia (duration less than 3 months) is typically triggered by identifiable stressors and often resolves spontaneously; chronic insomnia disorder involves perpetuating factors (conditioned arousal, sleep-incompatible behaviors, cognitive hyperarousal) that maintain the disorder independently of the original precipitant — a distinction with important therapeutic implications.
Comorbid insomnia — insomnia occurring in the context of a psychiatric or medical condition — is far more common than primary insomnia and requires treatment directed at both conditions. The historical practice of treating the underlying condition and expecting insomnia to resolve has been replaced by a model of treating insomnia as a co-occurring condition in its own right, as evidence shows that untreated insomnia worsens outcomes of comorbid depression, anxiety, chronic pain, and cardiovascular disease through bidirectional mechanisms.9
Cognitive behavioral therapy for insomnia (CBT-I) is the first-line treatment for chronic insomnia disorder across all major clinical practice guidelines, including those of the American Academy of Sleep Medicine (AASM), the American College of Physicians (ACP), and the European Sleep Research Society. CBT-I produces durable improvements in sleep onset latency, sleep efficiency, and wake after sleep onset that are maintained at long-term follow-up, in contrast to pharmacological agents whose effects diminish after discontinuation and which carry adverse effect and dependence risks. CBT-I components include sleep restriction therapy (limiting time in bed to consolidate sleep drive), stimulus control (restricting bed use to sleep and sex to break conditioned arousal), sleep hygiene education, cognitive restructuring of dysfunctional sleep beliefs, and relaxation training. Digital CBT-I platforms (Sleepio, Somryst — FDA cleared) substantially expand access beyond in-person therapy and are supported by randomized controlled trial evidence, making them a particularly relevant option in primary care and rural medicine settings where sleep specialist access is limited.9
When pharmacotherapy is indicated — either because CBT-I is unavailable, has failed, or because rapid symptom control is clinically urgent — agent selection should be guided by the primary sleep complaint, patient characteristics, comorbidities, and risk factors. A practical decision framework proceeds as follows.
Sleep-onset insomnia (difficulty falling asleep, with adequate sleep maintenance once asleep): Short-acting Z-drugs are effective — zolpidem IR 5–10 mg, zaleplon 5–10 mg (can also be used for middle-of-night awakening if 4 hours of sleep remain). Ramelteon 8 mg is a non-scheduled first choice when controlled substance prescribing is problematic. Low-dose suvorexant 5–10 mg or lemborexant 5 mg offer sleep-onset efficacy without architecture disruption. For patients with comorbid depression and sleep-onset insomnia, trazodone 50–100 mg is a reasonable first option.
Sleep-maintenance insomnia (difficulty staying asleep, with adequate sleep onset): Eszopiclone 1–3 mg, zolpidem ER 6.25–12.5 mg, or suvorexant 10–20 mg or lemborexant 10 mg (superior evidence for wake after sleep onset (WASO) reduction). Low-dose doxepin 3–6 mg is FDA-approved specifically for sleep maintenance and is particularly appropriate when a non-scheduled agent is preferred. For middle-of-night awakening with at least 4 hours remaining, low-dose sublingual zolpidem 1.75–3.5 mg (Intermezzo) is approved for this specific indication.
Mixed onset and maintenance insomnia: DORAs (suvorexant, lemborexant) are the most versatile agents, with evidence for both endpoints. Eszopiclone covers both complaints. In patients with comorbid depression or anxiety, antidepressant-based approaches (trazodone, mirtazapine, doxepin) address both conditions simultaneously.
Special population overlays: Elderly patients — start with ramelteon or low-dose doxepin; if scheduling is acceptable, dual orexin receptor antagonist (DORA) at lowest dose; avoid benzodiazepines and Z-drugs (Beers Criteria). Patients with substance use disorder history — ramelteon first choice; low-dose doxepin second; avoid all scheduled agents where possible. Patients with obstructive sleep apnea (OSA) on continuous positive airway pressure (CPAP) — DORAs preferred if hypnotic needed; confirm CPAP adherence. Patients with PTSD — DORAs preferred (rapid eye movement (REM) preservation); prazosin if nightmares are the dominant complaint. Patients with severe hepatic disease — ramelteon contraindicated (CYP1A2 (cytochrome P450 1A2) dependent); LOT benzodiazepines if indicated; dose reduction for all agents.9
The mismatch between the chronic nature of insomnia disorder (by definition persisting ≥3 months) and the short-term evidence base for most hypnotic agents (typically 7–35 days labeling) creates a persistent clinical tension. Eszopiclone and suvorexant have the most robust long-term (6–12 month) efficacy and safety data. When pharmacotherapy extends beyond the short-term labeling, clinicians should document explicit rationale, ensure CBT-I has been offered or attempted, and schedule regular reassessment of continued need. Discontinuation should be planned from the outset of prescribing: structured tapering (rather than abrupt cessation) combined with concurrent CBT-I produces the highest rates of successful long-term discontinuation, with randomized trial evidence supporting success rates of 60–80% with this approach.9
Anxiety disorders collectively represent the most prevalent category of psychiatric illness, affecting approximately 30% of adults at some point in their lifetime. The major anxiety disorders — generalized anxiety disorder (GAD), panic disorder, social anxiety disorder (SAD), specific phobias, and separation anxiety disorder — share a core pathophysiology of dysregulated threat appraisal involving hyperactive amygdala responses, impaired prefrontal cortical regulation of fear circuits, and dysregulation of serotonergic, noradrenergic, and GABAergic neurotransmission. The pharmacological approaches to anxiety treatment reflect these neurobiological targets: serotonergic and noradrenergic reuptake inhibitors address the monoaminergic dysregulation over a weeks-long latency period, while benzodiazepines provide immediate GABAergic inhibition of limbic hyperactivity at the cost of the adverse effects and dependence liability discussed throughout this chapter.11
Selective serotonin reuptake inhibitors (SSRIs) and serotonin-norepinephrine reuptake inhibitors (SNRIs) are the established first-line pharmacotherapy for all major anxiety disorders with a chronic or recurrent course. Their anxiolytic efficacy, established across randomized controlled trials in GAD, panic disorder, social anxiety disorder, and PTSD, is comparable to benzodiazepines for most patients but without dependence liability, tolerance, or the cognitive impairment associated with GABAergic agents.11 The critical clinical limitation is onset latency: 2–6 weeks are required before clinically meaningful anxiolytic effects are established, during which patients may experience transient worsening of anxiety (particularly jitteriness and insomnia from serotonergic activation in the first 1–2 weeks). This latency period is the primary clinical indication for short-term benzodiazepine bridging. FDA-approved agents for specific anxiety disorders include: GAD — escitalopram, paroxetine, duloxetine, venlafaxine; panic disorder — sertraline, paroxetine, fluoxetine, venlafaxine, clonazepam, alprazolam; social anxiety disorder — sertraline, paroxetine, venlafaxine; PTSD — sertraline, paroxetine.
The pharmacologically rational use of benzodiazepines in anxiety disorders is as a bridging agent during the SSRI/SNRI latency period, not as primary monotherapy. A typical bridging protocol initiates the SSRI or SNRI at the first visit with concurrent short-term benzodiazepine prescription (typically lorazepam 0.5–1 mg twice daily or clonazepam 0.25–0.5 mg twice daily) for a defined period of 2–4 weeks, with a clear plan to taper and discontinue the benzodiazepine once SSRI/SNRI efficacy is established. This strategy manages the acute symptom burden while avoiding chronic benzodiazepine exposure. The bridging period requires clear patient communication that the benzodiazepine is temporary, with explicit documentation of the discontinuation plan. In patients with a history of substance use disorder, benzodiazepine bridging carries substantially higher risk and alternatives — buspirone, hydroxyzine, or pregabalin in appropriate contexts — should be considered.11
Buspirone is a partial agonist at 5-5-HT1A serotonin receptor (HT1A) serotonin receptors and a weak dopamine D2 antagonist, approved for generalized anxiety disorder. It produces anxiolysis without sedation, cognitive impairment, muscle relaxation, or anticonvulsant effects, and has no dependence liability or cross-tolerance with benzodiazepines. Its onset latency (1–4 weeks) makes it unsuitable for acute anxiety management but appropriate as a long-term anxiolytic alternative for GAD. Clinical use is limited by this latency, by modest efficacy compared to SSRIs and benzodiazepines in head-to-head studies, and by the fact that patients previously treated with benzodiazepines frequently find buspirone subjectively unsatisfying due to the absence of the reinforcing immediate CNS effects of benzodiazepines — a clinical phenomenon reflecting both receptor-level adaptation and pharmacological expectation. Buspirone is most useful in benzodiazepine-naive patients with GAD who require long-term anxiolytic therapy without the risks of benzodiazepine use.11
Panic disorder: Alprazolam and clonazepam are FDA-approved and effective, providing rapid suppression of panic attacks. However, the long-term management of panic disorder now centers on CBT (particularly interoceptive exposure) and SSRI/SNRI therapy. Benzodiazepines in panic disorder carry the specific risk of interfering with extinction learning — the neurobiological process underlying CBT efficacy — by blunting the physiological arousal responses that are necessary for exposure therapy to work. This pharmacological antagonism of behavioral therapy represents a meaningful clinical trade-off that should inform treatment planning. Social anxiety disorder: Benzodiazepines provide situational relief (performance anxiety, acute social fears) but do not address the cognitive and behavioral maintaining factors of SAD that respond to CBT and SSRIs. PTSD: Benzodiazepines are generally not recommended in PTSD. Evidence from randomized trials and observational studies suggests they do not reduce PTSD symptom severity and may worsen outcomes by interfering with fear extinction, increasing rapid eye movement (REM) suppression (counterproductive in a condition where REM-dependent emotional processing is impaired), and increasing risk of substance use disorder co-morbidity. Current PTSD treatment guidelines endorse SSRIs/SNRIs, trauma-focused CBT, and EMDR as evidence-based interventions, with benzodiazepines relegated to a secondary role for specific comorbid symptoms only.11
Neurosteroids are endogenous steroids synthesized in the brain (and periphery) that modulate GABA-A receptor (GABA-A) function through binding sites distinct from both the classical benzodiazepine site and the barbiturate site. The most pharmacologically relevant neurosteroid is allopregnanolone (3α-hydroxy-5α-pregnan-20-one), a progesterone metabolite that is a potent positive allosteric modulator of GABA-A receptors, including receptor configurations that contain δ subunits and are located extrasynaptically — a population not targeted by benzodiazepines. These extrasynaptic δ-subunit-containing receptors mediate tonic GABAergic inhibition and are particularly highly expressed in the hippocampus, thalamus, and cerebellum, contributing to mood regulation and stress responsivity.12
Brexanolone (Zulresso) is an IV formulation of synthetic allopregnanolone, FDA-approved in 2019 for postpartum depression — the first drug approved specifically for this indication. It is administered as a 60-hour continuous IV infusion in a certified healthcare setting and produces rapid and sustained antidepressant effects in women with moderate-to-severe postpartum depression (PPD), with response rates substantially higher than standard antidepressants in this specific population. Its mechanism involves restoration of the neurosteroid milieu disrupted by the precipitous drop in progesterone and its metabolites at delivery, which is thought to unmask vulnerability to PPD in susceptible individuals. Brexanolone's CNS depression is a significant safety concern during the infusion, requiring continuous pulse oximetry monitoring and prohibiting driving for 12 hours after infusion completion.
Zuranolone (Zurzuvae, oral neurosteroid) received FDA approval in 2023 for major depressive disorder and postpartum depression — the first oral neurosteroid and the first oral drug with demonstrated rapid antidepressant effects (within 3 days in clinical trials). It is taken once daily at bedtime for a 14-day treatment course, producing antidepressant effects that persist beyond the treatment period. At 50 mg, zuranolone produces next-day sedation and driving impairment that requires counseling and precautions analogous to those for sedative-hypnotics. The mechanism of action as a GABA-A modulator at synaptic and extrasynaptic receptors overlaps with conventional sedative-hypnotics, but its primary indication and rapid antidepressant effect at sub-sedating doses distinguish it pharmacologically. Zuranolone does not require Risk Evaluation and Mitigation Strategy (REMS) enrollment (unlike brexanolone), though driving precautions apply on days of use and the following morning.12
Remimazolam (Byfavo), introduced in CNS-03, represents a new paradigm for benzodiazepine-based procedural sedation. Its esterase-metabolized context-insensitive pharmacokinetics, CYP-independence, flumazenil reversibility, and minimal cardiovascular effects position it as a potential replacement for midazolam in procedural sedation settings. Post-approval clinical experience is accumulating in endoscopy, bronchoscopy, and minor surgical procedures. Its clinical advantage over propofol in procedural sedation settings outside the operating room includes full reversibility, substantially lower cardiovascular depression risk, and the ability to be administered by non-anesthesiologist providers in appropriate settings.10
The expanding clinical application of ketamine and its S-enantiomer esketamine beyond anesthesia and pain management into psychiatry represents one of the most significant pharmacological developments of the past decade. IV ketamine infusion therapy for TRD and acute suicidality is now offered at specialized centers across the United States, and the infrastructure for this service is growing rapidly in response to clinical need. The pharmacology of ketamine's antidepressant mechanism — NMDA antagonism, AMPA potentiation, brain-derived neurotrophic factor (BDNF)-mTOR signaling, and possibly synaptogenesis in prefrontal cortical circuits — represents a mechanistic departure from all prior antidepressants and has generated enormous interest in NMDA receptor modulation as a target for rapidly acting antidepressants. Multiple next-generation NMDA-targeted agents are in clinical development, including SAGE-718 (a positive allosteric modulator of NMDA receptors rather than an antagonist) and rapastinel (a partial NMDA receptor agonist), which aim to capture antidepressant activity without the dissociative adverse effects of ketamine.13
Several investigational agents targeting novel sleep-wake mechanisms are in clinical development. KNX100 and related compounds are selective orexin receptor 2 (OX2R) antagonists (as opposed to the dual orexin receptor 1 (OX1R)/OX2R blockade of suvorexant and lemborexant), based on evidence that OX2R is the more important orexin receptor subtype for sleep promotion and that selective OX2R blockade may produce sleep with fewer of the cataplexy-like adverse effects associated with OX1R blockade. Selective histamine H1 antagonists with improved selectivity over existing agents (doxepin) are in development for sleep maintenance. GABA-A receptor subtype-selective agents targeting α2 and α3 subunits selectively (to provide anxiolysis without sedation or dependence) have been an active area of drug development for two decades, though translating receptor subtype selectivity from in vitro to in vivo conditions has proven challenging. The fundamental pharmacological insight driving this work — that the adverse effects of classical benzodiazepines reflect α1 receptor engagement while anxiolytic and muscle relaxant effects reflect α2/α3 engagement — remains a compelling basis for the development of next-generation anxiolytics without sedation and dependence liability.4
Hobson JA. Sleep is of the brain, by the brain and for the brain. Nature. 2005;437(7063):1254–1256.
doi:10.1038/nature04283Borbely AA, Daan S, Wirz-Justice A, Deboer T. The two-process model of sleep regulation: a reappraisal. J Sleep Res. 2016;25(2):131–143.
doi:10.1111/jsr.12371Kato K, Hirai K, Nishiyama K, et al. Neurochemical properties of ramelteon (TAK-375), a selective MT1/MT2 receptor agonist. Neuropharmacology. 2005;48(2):301–310.
doi:10.1016/j.neuropharm.2004.09.007Sakurai T. The neural circuit of orexin (hypocretin): maintaining sleep and wakefulness. Nat Rev Neurosci. 2007;8(3):171–181.
doi:10.1038/nrn2092Ashton H. The diagnosis and management of benzodiazepine dependence. Curr Opin Psychiatry. 2005;18(3):249–255.
doi:10.1097/01.yco.0000165594.60434.84Sanna E, Busonero F, Talani G, et al. Comparison of the effects of zaleplon, zolpidem, and triazolam at various GABA-A receptor subtypes. Eur J Pharmacol. 2002;451(2):103–110.
doi:10.1016/S0014-2999(02)02191-XHerring WJ, Connor KM, Ivgy-May N, et al. Suvorexant in patients with insomnia: results from two 3-month randomized controlled clinical trials. Biol Psychiatry. 2016;79(2):136–148.
doi:10.1016/j.biopsych.2014.10.003Wunsch H, Kahn JM, Kramer AA, Rubenfeld GD. Dexmedetomidine in the care of critically ill patients from 2001 to 2007: an observational cohort study. Anesthesiology. 2010;113(2):386–394.
doi:10.1097/ALN.0b013e3181e74116Sateia MJ, Buysse DJ, Krystal AD, Neubauer DN, Heald JL. Clinical practice guideline for the pharmacological treatment of chronic insomnia in adults: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2017;13(2):307–349.
doi:10.5664/jcsm.6470Kilpatrick GJ. Remimazolam: non-clinical and clinical profile of a new sedative/anesthetic agent. Front Pharmacol. 2021;12:690875.
doi:10.3389/fphar.2021.690875Baldwin DS, Anderson IM, Nutt DJ, et al. Evidence-based pharmacological treatment of anxiety disorders, post-traumatic stress disorder and obsessive-compulsive disorder: a revision of the 2005 guidelines from the British Association for Psychopharmacology. J Psychopharmacol. 2014;28(5):403–439.
doi:10.1177/0269881114525674Meltzer-Brody S, Colquhoun H, Riesenberg R, et al. Brexanolone injection in post-partum depression: two multicentre, double-blind, randomised, placebo-controlled, phase 3 trials. Lancet. 2018;392(10152):1058–1070.
doi:10.1016/S0140-6736(18)31551-4Murrough JW, Iosifescu DV, Chang LC, et al. Antidepressant efficacy of ketamine in treatment-resistant major depression: a two-site randomized controlled trial. Am J Psychiatry. 2013;170(10):1134–1142.
doi:10.1176/appi.ajp.2013.13030392