◈ Nootpedia
Compact view
Research pass: thorough Pharmaceutical · Oral SKIP-PERMANENT HIGH

GHB / GBL

Extended Research
◈ Extended Research ◈

Our depth — beyond the mirror

Deeper analysis, verdict reasoning, and per-archetype recommendations from our research team.

▸ Our verdict SKIP-PERMANENT HIGH

GHB has the narrowest therapeutic window of any commonly-encountered psychoactive — recreational-to-unconscious gap is roughly 1.5-3 g vs 3-5 g (literally one extra "capful" of GBL or one mis-measured shot), and recreational-to-respiratory-arrest is 5-15× the active dose. Mixed with alcohol (multiplicative respiratory depression, no specific reversal agent), it is one of the highest case-fatality-rate club drugs in Western Europe (10.6% of acute drug toxicity presentations + 27% of critical-care admissions in Euro-DEN 2019 hospitals). Withdrawal from chronic use is among the most severe in clinical addiction medicine — fulminant autonomic crisis + delirium in >50% of untreated cases, ICU admission ~20-31% in observational cohorts, benzodiazepine-resistant in many cases (different receptor target — GHB acts at GHBR + GABA-B, BZDs at GABA-A), often requiring pharmaceutical GHB re-administration to abort. Sodium oxybate (Xyrem/Xywav/Lumryz) under REMS supervision IS A-tier evidence for narcolepsy/cataplexy/idiopathic-hypersomnia — but the safety envelope depends on the REMS architecture (single dispensing pharmacy, dose titration, monitoring, prohibition of alcohol/CNS-depressants, $177-212K/yr), none of which transfer to biohacker self-dosing. SKIP-PERMANENT applies to ALL non-narcoleptic archetypes; OPTIONAL only for diagnosed narcolepsy/cataplexy/IH under REMS. Verdict would change only if (a) a non-tolerizing GHB analog were discovered, or (b) Dylan develops narcolepsy with cataplexy and exhausts non-oxybate options — neither plausible.

Research pass: thorough
▸ Decision matrix by user profile Per-archetype
  • Dylan20-30, brain-priority, high cognitive workload (Dylan-archetype)
    SKIP-PERMANENT

    HIGH confidence. Narrow therapeutic window (1-2 g between fun and unconscious) is incompatible with any "occasional use" framing. Round-the-clock dosing required for any sustained subjective effect = dependence within weeks. Schedule I federal exposure + GBL UK/EU/AU bans. Daridorexant covers any sleep-architecture-preservation need; SSRIs/Selank/L-theanine cover anxiolysis; modafinil covers wake-promotion. No remaining use case. At 20 with zero substance baseline, introducing the most-overdose-prone CNS depressant in common use is a strict downside trade.

  • 30-50, executive maintenance
    SKIP-PERMANENT

    Same narrow-window calculus; additional alcohol-tolerance + social-drinking exposure raises real-world fatal-combination risk. Better Rx anxiolytics + sleep aids exist (SSRIs, hydroxyzine, daridorexant, low-dose trazodone).

  • 50+, mild cognitive decline
    SKIP-PERMANENT

    GABA-B agonism in older brains worsens memory + falls + cognitive function. Withdrawal in elderly is more dangerous (delirium superimposed on baseline cognitive vulnerability). The fall-risk plus respiratory-depression profile makes any non-medical use catastrophic. Even narcolepsy use in this archetype demands close cardiovascular + sodium-load monitoring (favors Xywav over Xyrem).

  • Anxiety-prone
    SKIP-PERMANENT

    Same reinforcement-driven dependence calculus as phenibut — the immediate anxiolytic effect is most reinforcing precisely where dependence forms fastest. Use SSRI/SNRI baseline + Selank/buspirone/hydroxyzine PRN; never GHB.

  • High athletic load, tested status
    SKIP-PERMANENT

    WADA banned (Section S2 / S6 depending on context); incompatible with training schedule (sedation + recovery interference); dose-error overdose risk catastrophic for training-schedule continuity. Even if WADA-irrelevant (Dylan's untested status), the dosing-window + dependence profile remains decisive.

  • Sleep-disordered (insomnia, non-narcolepsy)
    SKIP-PERMANENT

    Narcolepsy + IH are the only legitimate Rx indications. For garden-variety insomnia, daridorexant is the modern alternative (preserves sleep architecture, no dependence at therapeutic dose, no respiratory depression risk). Z-drugs / benzodiazepines / trazodone all dominate GHB for general insomnia. Sleep-architecture preservation argument is now competitive territory — DORAs (suvorexant, lemborexant, daridorexant) accomplish similar architecture-preserving sleep without GHB's lethality.

  • Sleep-disordered (narcolepsy with cataplexy / idiopathic hypersomnia, formal diagnosis)
    OPTIONAL

    only under REMS supervision. Sodium oxybate is first-line or co-first-line per AASM guidelines. Xywav preferred over Xyrem (lower sodium); Lumryz preferred over Xyrem if patient cannot tolerate middle-of-the-night dosing. Not Dylan's case — no narcolepsy diagnosis, no cataplexy, no IH symptoms.

  • Recovery-focused (post-injury, post-illness)
    SKIP-PERMANENT

    No role.

  • Strength/anabolic-focused
    SKIP-PERMANENT

    The 1990s GH-axis bodybuilding marketing was the proximate cause of FDA scheduling. Any GH-axis effect is dwarfed by dependence + overdose profile. Modern body-composition tools (creatine, protein, training) dominate completely.

  • Alcohol use disorder (specific Italian-context indication)
    CONDITIONAL

    CONSIDER under specialized AUD clinic supervision only — Italy-only. Not relevant for Dylan (zero alcohol baseline). For US AUD treatment, naltrexone + acamprosate + behavioral therapy + (rarely) French-style baclofen are the relevant alternatives. No archetype outside diagnosed narcolepsy/cataplexy/IH has GHB as anything other than SKIP-PERMANENT. Joins phenibut as one of the rare wiki entries with unanimous-skip across non-medical archetypes.

▸ Subjective experience (deep)

Low-recreational doses (0.5-1.5 g GHB sodium / 0.7-1.5 mL GBL):

  • Mild anxiolysis, sociability, mild euphoria
  • Talkative, lowered inhibitions, mild sensory enhancement (music, touch)
  • Comparable to low-dose alcohol (1-2 drinks) in social-disinhibition profile but without alcohol's cognitive impairment at this tier
  • Sexual disinhibition + lowered orgasm threshold (the chemsex use case)
  • Onset 15-30 min (GHB) / 5-15 min (GBL); duration 1-2 hr

Moderate-recreational doses (1.5-2.5 g GHB / 1.5-2.5 mL GBL):

  • "Drunk-like" euphoria + dissociation, more pronounced
  • Strong anxiolysis, social warmth, music + touch enhancement intensified
  • Mild ataxia, slurred speech beginning
  • Sexual + party-context use most common at this tier
  • Duration 2-3 hr; tapers into mild sleepiness

High-recreational / overdose threshold (2.5-4 g GHB / 2.5-4 mL GBL):

  • Sudden onset of overwhelming sedation
  • "G-nap" / "G-hole" — abrupt loss of consciousness, may be mistaken for sleep but ventilation can be compromised
  • Vomiting common (aspiration risk in unconscious user)
  • Amnesia for the period
  • Recovery typically within 1-3 hours unless co-ingestion (alcohol, GBL miscalculation) extends it

Clear overdose (5-7 g GHB / 5-7 mL GBL):

  • Coma, depressed reflexes, possible respiratory depression
  • Bradycardia, hypothermia
  • Vomiting in obtunded state → aspiration pneumonia risk
  • Often requires ED-level airway support; intubation in 10-57% of GHB poisonings (mean intubation duration 80-210 min in published series)

Lethal range (7-10+ g GHB or 5+ mL GBL, especially with alcohol):

  • Severe respiratory depression
  • Cardiovascular collapse, bradycardia, hypotension
  • Death from respiratory arrest or aspiration

The biphasic problem: the gap between "fun moderate dose" and "unconscious" is roughly 1-2 g — about a single capful of GBL or one mismeasured shot. There is no subjective sweet spot above the recreational tier; the next stop is unconsciousness. This is fundamentally different from alcohol (recreational-to-lethal ratio ~10-20×), benzodiazepines (huge therapeutic window when used alone), or modafinil (no acute lethality). GHB's narrow window is the central reason for the death toll.

Therapeutic-dose subjective experience (sodium oxybate Rx, narcolepsy patient):

  • Taken at bedtime (Xyrem twice-nightly q4hr, Xywav twice-nightly, Lumryz once-nightly extended-release)
  • Onset within 10-15 min of dose
  • Profound sleep within 15-30 min; sleep is "deep" (high SWS) — patients report not waking until alarm
  • No euphoria reported at therapeutic doses (4.5-9 g/night divided)
  • Morning grogginess possible at higher doses
  • Cataplexy reduction effect builds over weeks (not acute)
  • Critical: alcohol prohibited (multiplicative respiratory depression); other CNS depressants prohibited
▸ Tolerance + cycling deep dive
  • Tolerance buildup: extremely fast. Subjective tolerance within days of regular use. Functional dependence (withdrawal on cessation) within 1-3 weeks of round-the-clock dosing.
  • Recommended cycle: None. No cycling protocol survives contact with the dependence pharmacology in real-world recreational use. Chemsex + party-binge patterns reliably escalate to dependence given enough exposures.
  • Reset protocol: Inpatient pharmaceutical GHB tapering (European standard) — 12 doses/day every 2 hours of pharmaceutical-grade GHB with stepped reduction over 5-10 days, 85% completion rate. Benzodiazepine tapering is the alternative outside Europe — lower completion rate, higher symptom severity, frequent need for ICU-level care for refractory cases. Self-directed tapers fail because of the q2hr round-the-clock schedule + benzodiazepine resistance.
  • Cross-tolerance: With baclofen, phenibut, alcohol, benzodiazepines (incomplete but clinically meaningful — baclofen substitution has been used adjunctively in withdrawal). Fully cross-tolerant with sodium oxybate / Xyrem / Xywav (same active drug) and 1,4-butanediol (same active metabolite).
▸ Stacking deep dive

Synergistic with (recreational — DO NOT USE)

  • Alcohol: the dominant fatal combination. Multiplicative respiratory depression. >90% of GHB-related deaths in Australian forensic data involve co-ingestants, with alcohol the most common. Sodium oxybate Rx prescribing information explicitly prohibits alcohol use. Any concurrent alcohol intake with GHB is a death-risk pattern.
  • Other GABAergic depressants (benzodiazepines, Z-drugs, phenibut, baclofen, gabapentinoids): multiplicative respiratory depression + amnesia + dependence acceleration.
  • Opioids: multiplicative respiratory depression. Documented in death reports.
  • Methamphetamine, MDMA, cocaine (chemsex polypharmacy): the "stim + GHB" combination produces the chemsex-characteristic alternating euphoria-then-sleep pattern; stimulants mask GHB's sedative cue allowing further dosing → overdose risk amplified when stim wears off.

Avoid stacking with

  • Anything in the CNS-depressant family — alcohol, benzodiazepines, Z-drugs, opioids, baclofen, phenibut, gabapentin, pregabalin, ketamine, dextromethorphan at high doses.
  • Other oxybate / GHB-prodrug forms — Xyrem + recreational GHB = same drug, additive overdose risk.
  • MAOIs: theoretical concern via β-PEA / dopaminergic mechanisms; under-characterized but caution warranted.
  • Sodium-restricted patients (heart failure, hypertension): Xyrem's sodium load is clinically meaningful (Xywav exists specifically to address this).

Neutral / safe co-administration

Not relevant — verdict is don't take it at all outside REMS. The "neutral" category is moot in non-medical use.

▸ Drug interactions deep dive
  • CYP enzymes: GHB is not metabolized via CYP — primary elimination is via SSADH/SSAR (mitochondrial enzymes) into the TCA cycle. Drug-drug interactions are pharmacodynamic (additive CNS depression with depressants), not pharmacokinetic.
  • Alcohol: MAJOR pharmacodynamic interaction — multiplicative respiratory depression + competition at GABA-B + ethanol slows 1,4-BD → GHB conversion via ADH competition.
  • Valproic acid (sodium valproate): inhibits SSADH, the enzyme that clears GHB → succinate. Increases GHB exposure and prolongs duration. Clinically relevant in patients on valproate for seizures or mood stabilization.
  • Other GABAergic depressants: additive CNS + respiratory depression.
  • Renal/hepatic impairment: generally well-tolerated due to non-CYP metabolism, but accumulating GHB in severe hepatic failure has been documented.
▸ Pharmacogenomics
  • No actionable PGx for GHB response or dependence risk as of 2026. Endogenous metabolism via SSADH is genetically variable but clinically uncharacterized for response prediction.
  • SSADH deficiency (GABA-transaminase deficiency, OMIM 271980) — rare inborn error of metabolism producing GHB accumulation; affected children have neurodevelopmental delays + ataxia. Distinct from pharmacological GHB exposure but illustrates the metabolic-pathway relevance.
  • PON1 (paraoxonase) variants — relevant for GBL → GHB conversion rate. Slow PON1 metabolizers theoretically have delayed GBL onset, but no clinical data characterizes this for safety prediction.
  • Practical takeaway for Dylan: Genotype data (23andMe results June 2026 window) does not change verdict. The dose-error fatality risk is universal; PGx data cannot rescue the narrow therapeutic window.
▸ Sourcing deep dive
Path Vendor Cost Reliability Notes
Rx (REMS — narcolepsy/IH) — Xyrem Single dispensing pharmacy via Jazz REMS ~$160-180K/yr (cash); insurance copay variable High — pharmaceutical grade, certified Restricted-distribution REMS program; single specialty pharmacy; physician REMS certification required; patient-specific dispense records; alcohol/CNS-depressant-prohibition counseling mandatory. Not available outside narcolepsy/IH diagnosis.
Rx (REMS) — Xywav Same Jazz REMS ~$170-200K/yr High Lower-sodium formulation (preferred for cardiovascular comorbidity); narcolepsy + IH indications.
Rx (REMS) — Lumryz Avadel REMS $177-212K/yr (cash) High Once-nightly extended-release; eliminates middle-of-night dose requirement; orphan-drug exclusivity through 2031.
Italian Alcover (sodium oxybate for AUD) Italian pharmacy with Italian Rx €30-100/mo High Italy-only AUD indication. Not exportable to US in any practical sense.
Grey-market GBL (industrial solvent) Online "wheel cleaner" / "ink remover" / "nail polish remover" vendors $50-200 / liter Very low — DEA-monitored List I chemical, federal analogue-act prosecution, banned in UK/EU/AU; vendors face frequent law-enforcement action; product purity variable; mislabeling common; analytical-grade vs. industrial-grade matters for impurity profile DO NOT SOURCE. Sourcing GBL in the US is federally prosecutable under the Federal Analogue Act + List I chemical recordkeeping requirements. UK/EU/AU outright bans. Purity is uncontrolled — industrial GBL contains hepatotoxic impurities.
Grey-market GHB powder / liquid Clandestine synthesis + dark-market vendors Variable Very low — Schedule I felony; clandestine product purity variable DO NOT SOURCE. US Schedule I (post-2000); identical legal exposure to heroin or LSD in possession-amount terms.
Grey-market 1,4-butanediol (1,4-BD) Plastics / solvent vendors Variable Very low — analogue-act prosecutable; industrial purity issues DO NOT SOURCE. Treated as GHB by federal analogue act; purity uncontrolled.

Sourcing-difficulty rating: not-available for legitimate non-Rx pathway. Rx pathway is restricted-distribution REMS for narcolepsy/IH only. Grey-market sourcing is federally prosecutable (US) or outright illegal (UK/EU/AU) and carries impurity + dose-accuracy risks that compound the already-narrow therapeutic window.

Key point: Unlike phenibut (sourcing-solvable but verdict is don't), GHB/GBL is not even sourcing-solvable for biohacker use without serious legal exposure + product-purity uncertainty. The verdict of SKIP-PERMANENT is reinforced by sourcing infeasibility — you can't even get reliable product without taking on Schedule I felony risk.

▸ Biomarkers to track (deep)

Not applicable — verdict is not to take it. Listed for completeness in case of forced clinical exposure (Rx for narcolepsy or presentation in withdrawal/overdose):

  • If on Rx (Xyrem/Xywav/Lumryz):

    • Polysomnography baseline + post-titration for narcolepsy treatment monitoring
    • Multiple Sleep Latency Test (MSLT), Maintenance of Wakefulness Test (MWT), Epworth Sleepiness Scale (ESS) — efficacy outcomes
    • Sodium + electrolyte panel (Xyrem specifically — sodium load monitoring)
    • BP, HR — Xyrem-related hypertension monitoring
    • Mental health screening (depression, suicidal ideation are labeled risks)
    • Sleep apnea screening pre-treatment (contraindicated in untreated severe sleep apnea)

  • If presenting in acute toxicity (overdose):

    • GCS, respiratory rate, O2 sat — airway management decision
    • Co-ingestion screen (alcohol, opioid, benzodiazepine, stimulant) — polysubstance pattern
    • LC-MS/GC-MS GHB confirmation (immunoassay does NOT detect GHB)
    • CK, BUN, creatinine — rhabdomyolysis screen if prolonged immobility
    • LFTs — hepatotoxicity screen if chronic high-dose GBL exposure

  • If presenting in withdrawal:

    • Vital signs q1-2hr (HR, BP, temperature, RR, O2 sat) — autonomic crisis monitoring
    • Continuous cardiac telemetry during peak withdrawal (24-72 hr)
    • Electrolytes (Na, K, Mg, phosphate) — depletion common in autonomic crisis
    • LFTs — chronic-use hepatotoxicity baseline
    • LC-MS toxicology — confirms GHB (and rules out polysubstance)
    • CIWA-style adapted withdrawal score q4-6hr (CIWA-Ar applies imperfectly — GHB-specific scales like the SOWS-GHB are emerging)
    • Seizure precautions; EEG if concerning mental status
    • CK + creatinine — rhabdomyolysis screen
▸ Controversies / open debates Live debate

  1. "Xyrem proves GHB is safe, so biohacker dosing should be safe too." False — category error. Xyrem's safety record under REMS is real but rests on the entire architecture: single dispensing pharmacy, physician REMS certification, patient counseling, dose titration under medical supervision, mandatory alcohol prohibition, contraindication screening (sleep apnea, severe heart failure), narcolepsy/IH-only indication. Each of these elements meaningfully reduces the harm profile. None of them transfer to biohacker self-dosing of grey-market GHB or GBL. The pharmacological molecule is the same; the safety envelope is created by the surrounding clinical infrastructure, not by the molecule itself. Citing Xyrem's narcolepsy safety to defend recreational GHB use is exactly parallel to citing methadone-clinic safety to defend recreational heroin use.

  2. "Sleep architecture preservation is unique to GHB." Partially true 2010, contested 2026. GHB (sodium oxybate) is unique among older hypnotics in increasing slow-wave sleep (SWS, N3) — benzodiazepines and Z-drugs suppress SWS, eszopiclone partially preserves it, trazodone modestly preserves it, mirtazapine modestly preserves it. However: Dual orexin receptor antagonists (DORAs — suvorexant, lemborexant, daridorexant) preserve overall sleep architecture without specifically increasing SWS but without distorting it either. For most clinical insomnia indications, daridorexant is now the architecture-preserving option of choice without the dependence + lethality profile. The narcolepsy + cataplexy + idiopathic hypersomnia indications retain GHB's specific advantage (the SWS-enhancement appears to be the mechanism for cataplexy reduction, not just generic sleep promotion). Net 2026 framing: GHB's sleep-architecture argument has narrowed from "unique" to "specific to narcolepsy-class indications."

  3. "Endogenous = safe." False. GHB is endogenously produced at 1-4 µM concentrations in CNS tissue. Pharmacological doses produce 100-1000× tissue concentrations engaging receptors that endogenous concentrations do not (specifically GABA-B). The "endogenous metabolite" framing — heavily used in 1990s supplement-marketing — is biologically true but pharmacologically irrelevant for safety prediction. Adrenaline, dopamine, glutamate, and acetylcholine are also endogenous but lethal at pharmacological doses.

  4. "GBL is just a precursor, not a real drug." False — GBL is the more dangerous form in practice because of faster onset (5-15 min vs 15-30 min for GHB salt), smaller dosing volumes (~half the volume by mass for equivalent effect), higher BBB penetration, and industrial-solvent purity issues. GBL deaths in chemsex contexts are increasing in UK/EU; UK reclassified GBL from Class C to Class B in 2022 specifically in response.

  5. "Once-nightly Lumryz solves the dosing safety problem." Partially. Eliminating the q4hr middle-of-night second dose reduces dosing-error opportunities and improves patient adherence. But Lumryz is a Rx product under REMS, not a recreational alternative — and the per-dose narrow-window pharmacology is unchanged. Lumryz makes Rx narcolepsy use safer at the margin, but does not transform the recreational-use risk profile.

  6. "Sodium oxybate for AUD shows broader safety." Italy-only regulatory approval, contested broader application. The Italian Alcover program for alcohol use disorder demonstrates that GHB substitution can work analogous to methadone for opioid addiction, but only under specialized AUD-clinic supervision with q3-4hr round-the-clock dosing schedules. Not generalizable to other indications, and the dependence-and-substitution model implicitly acknowledges GHB's abuse liability rather than refuting it.

  7. Where my prior verdict might be wrong: Hard to imagine a scenario for non-narcoleptic users. If Dylan ever develops narcolepsy with cataplexy (extremely unlikely at 20 with no current symptoms), the verdict for that specific case flips to OPTIONAL under REMS — as one of the genuinely first-line indications. Discovery of a non-tolerizing GHB-receptor-selective agonist (engaging GHBR without GABA-B) could open targeted use without dependence pharmacology — but no such compound is in clinical development as of 2026 to public knowledge.

▸ Verdict change log
  • 2026-05-05 — Initial verdict: SKIP-PERMANENT, HIGH confidence. Rationale: narrowest therapeutic window of any commonly-encountered psychoactive (1-2 g between recreational and unconscious), Schedule I federal scheduling + GBL banned in UK/EU/AU, severe physical dependence + benzodiazepine-resistant withdrawal within weeks of round-the-clock dosing, polysubstance death pattern dominated by alcohol co-ingestion (>90% of GHB-related deaths in 2025 Australian forensic dataset). Sodium oxybate Rx (Xyrem/Xywav/Lumryz) is A-tier evidence for narcolepsy/cataplexy/IH but the safety envelope is REMS-architecture-dependent and does not transfer to non-medical use. What would change verdict: essentially nothing realistic for non-narcoleptic users. Discovery of a non-tolerizing GHBR-selective agonist could open targeted use without dependence; Dylan developing narcolepsy with cataplexy could open Rx-specific OPTIONAL — neither plausible.
▸ Open questions / gaps Open

  1. GHBR selectivity research — is a non-GABA-B-engaging GHBR agonist developable? Theoretically interesting (would isolate the high-affinity excitatory + low-dose stimulating effects from the GABA-B-driven sedation + dependence pharmacology) but no public evidence of active drug-discovery programs targeting GHBR specifically. If achievable, would restructure the entire risk profile.

  2. PGx for dependence risk + overdose susceptibility. Currently un-investigated. SSADH variants, PON1 variants, GABBR1/2 variants all theoretically modifiable factors but no clinical PGx panels predict response or risk.

  3. Long-term cognitive effects of pediatric narcolepsy treatment. Sodium oxybate is approved in 7+ year-olds with narcolepsy; long-term cognitive + neurodevelopmental data is accumulating but incomplete. Outside scope for Dylan but relevant for general clinical practice.

  4. Comparative efficacy of Xyrem vs Xywav vs Lumryz in head-to-head trials. Limited direct comparative trials; cross-trial efficacy comparisons are confounded by population + protocol differences. Real-world effectiveness data (REFRESH study, 2024) accumulating.

  5. GBL impurity profile + chronic-use hepatotoxicity quantification. Industrial GBL contains variable impurities; case reports of chronic GBL use → acute liver failure exist but population-level rate is uncharacterized. Relevant for grey-market harm-reduction but not for Dylan-context.

▸ Sources (full, with our context)
Back to compact view