Acute (single dose, opioid-naive, BZD-naive user)

• Onset 30–60 min oral, ~5–15 min sublingual or IV.
• Anxiolytic — "the anxiety just stops mattering" rather than the absence of anxious thoughts. Dissociative-quality calm.
• Sedating — sleepiness, slowed reaction time, ataxia at higher doses.
• Mildly euphoric, more so for short-acting / fast-onset agents (alprazolam, triazolam) — the "rush" reported by recreational users is more pronounced when blood levels rise rapidly. Diazepam IV gives a similar rush; oral diazepam does not (slower absorption).
• Anterograde amnesia at higher doses — the next-morning memory hole is the classic "Halcion blackout" / "Xanax bar" phenomenon. Triazolam was famously implicated in violent and bizarre behavior with no recollection.
• Disinhibition — paradoxical aggression and disinhibition occur in 1–10% of users and can be severe.
• Muscle relaxation — pleasant for some, unsteady for others.

Chronic (after weeks of daily use)

• Tolerance — "I need 1 mg now where 0.5 was working."
• Inter-dose anxiety — alprazolam users feel anxiety rise as the prior dose wears off (4–8 hours post-dose), reinforcing PRN re-dosing. This is the dependence loop in real time.
• Cognitive flattening — "brain fog," difficulty learning new material, slower processing. Often unnoticed by user, obvious on neuropsych testing.
• Emotional blunting — calmer, but also less able to feel pleasure (anhedonia signal).
• Sleep architecture distortion — REM and slow-wave sleep both suppressed; total sleep time may be longer but quality is worse.

Withdrawal (after weeks–months of daily use, abrupt or rapid cessation)

• Acute (days 1–14): rebound anxiety often worse than baseline, insomnia, tremor, sweating, tachycardia, headache, photophobia, hyperacusis, depersonalization, akathisia, perceptual disturbances. Seizures are possible — risk highest with short-acting agents (alprazolam most), users on high doses, polypharmacy, and prior seizure history. Seizures can be fatal. Delirium possible in heavy/long-duration users (similar phenomenology to alcohol DTs).
• Sub-acute (weeks 2–8): persistent anxiety, insomnia, mood lability, cognitive symptoms, GI distress, paraesthesia.
• Protracted (BIND, months 2–18+): intermittent waves of the above. ~10–15% of long-term users. Resolution slow, sometimes incomplete in published case series. The defining feature is symptoms appearing despite the drug having long since cleared the body.

The brutal asymmetry: feels good fast, withdraws bad slow.

• Tolerance buildup: fast. Sedation tolerance days–weeks, anxiolytic tolerance weeks–months, anticonvulsant tolerance months. Cognitive impairment shows little tolerance (you adapt subjectively but not objectively).
• Recommended cycle: there isn't one that's safe long-term. PRN-only use with self-imposed limits (<2× per week, <4 doses per month) is the maximum cycling discipline most users sustain, and even that drifts upward over months in real-world use.
• Reset protocol: there is no "reset." There is only taper. Ashton-Manual taper: substitute long-acting (diazepam) for short-acting first, then reduce by ~10% of current dose every 1–2 weeks, slower in the last 25%, total taper duration 6–18 months for chronic users. Cold-turkey from chronic high-dose is medically contraindicated (seizure risk).

Synergistic with (DANGEROUSLY)

• Opioids — respiratory depression, primary overdose-death mechanism. FDA Boxed Warning since 2016. Avoid co-prescribing.
• Alcohol — additive CNS depression and respiratory depression. Common cause of accidental death.
• Z-drugs (zolpidem, zopiclone, eszopiclone) — same GABA-A α1 site, additive sedation and respiratory depression. Do not stack.
• Barbiturates — additive at GABA-A; high lethality in combination.
• Other CNS depressants — first-generation antihistamines (diphenhydramine), tricyclic antidepressants, antipsychotics, gabapentinoids (gabapentin, pregabalin) — additive sedation and cognitive impairment.

Avoid stacking with (overlapping mechanism / wasteful)

• L-theanine — not dangerous but redundant; theanine is a much milder GABAergic/anxiolytic and the benzodiazepine swamps any theanine effect. Theanine is the better daily anxiolytic precisely because it has none of the dependence/withdrawal/cognitive-impairment liability.
• Magnesium glycinate — same logic; magnesium is the better daily lever for GABAergic tone without addiction.
• Picamilon, phenibut — phenibut especially is dangerous to co-administer (phenibut is a GABA-B + α2δ calcium channel ligand with its own severe withdrawal syndrome, and stacking compounds dependence pharmacology).

Neutral / safe co-administration (in single-dose acute medical contexts only — not endorsing chronic stacking)

• SSRIs, SNRIs (note: fluvoxamine and fluoxetine inhibit CYP3A4 and raise alprazolam/diazepam levels — clinically relevant)
• Most non-CNS-depressant medications

CYP enzymes (the major axis)

• Alprazolam, diazepam, midazolam, triazolam — primarily CYP3A4-metabolized. Strong CYP3A4 inhibitors (ketoconazole, itraconazole, clarithromycin, ritonavir, grapefruit juice, fluvoxamine) raise levels substantially. CYP3A4 inducers (rifampin, carbamazepine, St John's wort) lower levels.
• Diazepam — also CYP2C19. Polymorphism: poor metabolizers (~3% Caucasian, ~15–20% East Asian) have higher and longer-lasting exposure.
• Lorazepam, oxazepam, temazepam — direct phase-II glucuronidation (UGT), no significant CYP interaction. This is why they are preferred in elderly, hepatic impairment, and in patients on complex polypharmacy.
• Clonazepam — CYP3A4-metabolized but has minimal active metabolites.

Contraceptive interactions

• Generally none of clinical significance, though some case reports of reduced efficacy with chronic use of strong CYP3A4-inducing benzodiazepines (rare).

Other notable interactions

• Methadone + alprazolam — particularly high overdose-death rate; alprazolam combination implicated in roughly half of methadone-program overdose deaths in some cohorts.
• Buprenorphine + benzodiazepine — historically considered relatively safer than full-agonist opioid + benzodiazepine, but FDA explicitly stated in the 2017/2020 boxed-warning updates that this is NOT a reason to withhold buprenorphine if a patient is on a benzodiazepine — opioid use disorder treatment with buprenorphine outweighs the co-administration risk.

• CYP3A4 polymorphisms (CYP3A4*22 reduces enzyme activity) — affect alprazolam, midazolam, triazolam exposure. Clinical impact modest in most users.
• CYP2C19 polymorphisms — *2 and *3 alleles confer poor-metabolizer status (~3% Caucasian, ~15–20% East Asian, ~5% African). PMs of CYP2C19 have ~2× higher diazepam exposure and longer half-life. Dylan is Nordic/British ancestry, so PM probability is low (~3%) but not zero, and 23andMe results (June 2026) will resolve this.
• GABA-A receptor polymorphisms (GABRA1, GABRA2, GABRB3) — research-stage; weak signal for variability in benzodiazepine response, sedation, and possibly addiction liability. Not clinically actionable.
• UGT2B7 / UGT2B15 polymorphisms — affect lorazepam, oxazepam, temazepam clearance. Mostly subclinical.

For Dylan: not sourcing this, full stop. Even for the legitimate one-off PRN use case, the path is "talk to a prescriber for a genuine medical reason" not "stockpile from gray market." The risk of fentanyl-laced counterfeit pressed alprazolam is documented and lethal.

For chronic users (which Dylan is not and will not be):

• Baseline (before starting): cognitive battery (CNS Vital Signs, RAVLT, n-back), GAD-7, BAI/STAI, sleep architecture (Oura/Whoop deep + REM baseline), CMP (liver/kidney), CBC.
• During use: subjective anxiety, sleep quality, cognitive symptoms, dose creep documentation, GAD-7 quarterly.
• Post-cycle / post-taper: cognitive battery repeat at 1 month, 6 months, 12 months post-taper to characterize recovery / detect BIND.

For PRN single-dose use: none routine.