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

Donepezil (Aricept)

Extended Research
◈ Extended Research ◈

Our depth — beyond the mirror

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

▸ Our verdict SKIP-FOR-NOW HIGH

Massively-validated FDA drug for Alzheimer's, but the only halfway-decent healthy-young trial is Yesavage 2002 (n=18 pilots, single-domain, never replicated); the side-effect profile (GI, vivid dreams + insomnia from HS dosing, bradycardia/syncope, REM behavior disorder reports) is not justifiable for a 20yo with no cognitive deficit and a citicoline + alpha-GPC + (planned) modafinil + bromantane stack already covering the cholinergic + arousal axis. Verdict would only change if (a) a properly-powered healthy-young replication of Yesavage emerged with cleaner cognitive endpoints, or (b) Dylan developed a specific procedural-skill retention use case (rare).

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

    Healthy young AChEI use is poorly evidenced — Yesavage 2002 (n=18, mean age 52) and Grön 2005 (n=24) are the only halfway-relevant trials, both single-domain, never replicated. Side-effect profile (GI, sleep, bradycardia, mood, weight loss, REM disinhibition) is disproportionate to the marginal cognitive lift in someone with intact cholinergic neurons. Citicoline + alpha-GPC PRN + modafinil + bromantane covers the same axis with better felt effect, better evidence base in healthy young, and milder side-effect profile.

  • 20-30, MMA / combat-sport athlete
    SKIP

    bradycardia + REM behavior disorder + dream disinhibition + GI + weight loss are all anti-synergies with combat sport. Donepezil-induced RBD with a sparring partner sleeping next to you is a non-zero injury risk. Weight loss in a competitive athlete is a problem. Bradycardia interacting with cardiovascular load is unstudied.

  • 30-50, executive maintenance
    SKIP

    at baseline; WATCH-LIST if early subjective cognitive complaints + family history of AD + APOE ε4. Even then, lifestyle (sleep, exercise, diet, sleep apnea screen) and lighter-touch cholinergic support (citicoline, alpha-GPC, possibly low-dose huperzine) come first. Donepezil is not a primary prevention tool — no good evidence for asymptomatic prevention.

  • 50+, mild cognitive impairment (MCI)
    CONSIDER

    ADCS data shows transient delay-of-AD-conversion at 12 mo, lost by 36 mo. Off-label use is reasonable in selected patients — but neurology-level discussion, not self-prescribed.

  • 50+, mild-to-severe AD or DLB
    STRONG-CANDIDATE

    / standard of care. This is the FDA-approved indication. Donepezil + memantine is mainstream. Discuss with neurologist.

  • Anxiety-prone
    NOT INDICATED

    Donepezil is not anxiolytic; can worsen anxiety in some users via cholinergic activation + GI distress.

  • High athletic load, tested status
    NOT WADA-BANNED

    donepezil is not on the prohibited list. But physiologically anti-synergistic with combat / endurance sport for the bradycardia, weight loss, and REM-sleep concerns.

  • Sleep-disordered
    AVOID

    Donepezil is one of the most sleep-disrupting AChEIs (vivid dreams, insomnia, RBD). Use galantamine or rivastigmine if AChE inhibition is clinically required.

  • Recovery-focused (post-injury, post-illness)
    NEUTRAL

    no recovery benefit. Don't add.

  • Strength / anabolic-focused
    AVOID

    anorexia + weight loss are anti-anabolic.

  • Procedural-skill retention use case (pilots, surgeons, MMA technical seminar prep)
    —

    the only conceivable Yesavage-style application. Even here: small evidence base, side-effect cost, and easier alternatives (spaced retrieval, sleep optimization, modafinil for the learning session itself). Not recommended for Dylan.

  • Bipolar / psychotic-spectrum
    CAUTION

    Cholinergic activation can destabilize mood; donepezil-induced depression and rare psychiatric reactions are documented.

  • Lucid-dreaming hobbyist
    —

    off-label community use exists at 5-10 mg WBTB (waking back to bed) protocol. Not recommended — the side-effect profile (cardiac, GI, weight, RBD) is overkill for hobbyist dream work; galantamine 4-8 mg has better risk-reward in this niche.

▸ Subjective experience (deep)

Acute (single dose, naive user):

  • Onset: 3-4 hr to peak plasma. Subjective effect typically minimal day 1.
  • Day 2-7: GI side effects dominate — nausea, loose stool, occasional vomiting, especially in first 1-2 weeks. Often dose-limiting.
  • Sleep: vivid dreams, sometimes nightmares, by night 2-3. This is one of the most consistent subjective reports — not pleasant for most users.

Steady state (after 2-4 weeks):

  • GI typically improves; some users experience persistent mild nausea or anorexia → weight loss (common in elderly AD patients, less of a concern for Dylan if he were taking it, but still — chronic anorexia is real).
  • Cognitive subjective effect: subtle. In AD patients, family members notice better engagement, faster word-finding, less repetition. In healthy users, most report either nothing felt or a mild "background clarity" that is hard to distinguish from placebo. Notably does NOT produce the felt activation of modafinil, the focus of methylphenidate, or the motivation of bromantane — donepezil is a substrate-pathway drug, not a stimulant.
  • Sleep: vivid dreams persist for many; some develop insomnia (cholinergic REM activation conflicts with sleep onset). Some users report waking groggy from dream-heavy nights.
  • Mood: flat affect, mild dysphoria, irritability in a subset of healthy young experimenters — same "cholinergic over-tone depression" pattern documented for chronic high-dose alpha-GPC. Likely mechanism: chronic cholinergic > dopaminergic balance shift, possibly serotonergic involvement.

Compared to other AChEIs:

  • vs huperzine A: donepezil is smoother (no spikes), longer-acting, more consistent. Huperzine cycles out faster.
  • vs galantamine: donepezil lacks galantamine's nicotinic boost (so lower lucid-dreaming intensity in some users, but still real). Donepezil's GI is generally worse than galantamine's at equivalent CNS effect.
  • vs rivastigmine: donepezil has fewer peripheral cholinergic effects (no BChE inhibition) but rivastigmine is the only patch — better dose smoothing.
▸ Tolerance + cycling deep dive
  • Tolerance to cognitive effect: modest. In AD, the symptomatic benefit attenuates over years (because underlying disease progresses, not because tolerance to the drug develops at the receptor level). At a fixed dose, the cholinergic-pathway effect is reasonably stable.
  • Tolerance to side effects: GI side effects DO tolerate — most users improve substantially over 4-6 weeks. Vivid dreams, sleep disturbance, weight loss often persist.
  • Recommended cycle: none in clinical use — donepezil is dosed continuously in AD. Discontinuation produces rebound cognitive decline.
  • Reset protocol: if discontinuing after >3 months use, taper over 2-4 weeks to mitigate rebound. With ~70-hr half-life, the body essentially auto-tapers over the first 7-10 days regardless.
▸ Stacking deep dive

Synergistic with

  • Memantine (NMDA antagonist, FDA-approved for moderate-severe AD): ✅ Standard combination in moderate-severe AD; complementary mechanisms (cholinergic substrate + glutamatergic modulation). Not Dylan-relevant.
  • Cerebrolysin: ⚠️ Some Eastern European protocols add Cerebrolysin to AChE inhibitors in AD. Not Dylan-relevant unless Dylan develops cognitive decline (decades away).
  • Cognitive training / spaced retrieval / cardiovascular exercise: ✅ Behavioral cholinergic-axis support. Universally additive.
  • DHA / phosphatidylserine / citicoline: ⚠️ Theoretical complementarity (substrate + AChEI), but at clinical AChEI dose the bottleneck is no longer choline supply — synaptic ACh is already maximized by enzyme inhibition. In healthy young users, citicoline alone covers the choline-substrate role without the AChEI's costs.

Avoid stacking with

  • Other AChE inhibitors (huperzine A, galantamine, rivastigmine, alpha-GPC at high chronic dose): ❌ Cholinergic excess risk. Additive AChE inhibition + raised substrate = nausea, sweating, bradycardia, tremor, REM hyper-disinhibition. Pick one cholinergic intervention.
  • Beta-blockers (propranolol, metoprolol): ❌ Additive bradycardia. Donepezil + propranolol can produce symptomatic bradycardia and AV block. If both are needed (anxiety + cognitive support), do so under cardiology supervision with ECG monitoring. Dylan's planned propranolol PRN for sales-call anxiety — this is a direct stacking conflict if he were on chronic donepezil.
  • Anticholinergic drugs (oxybutynin, TCAs, diphenhydramine, scopolamine, first-gen antihistamines): ❌ Pharmacodynamic antagonism — anticholinergics blunt donepezil's effect. Stupid combination.
  • Neuromuscular blockers (succinylcholine, vecuronium): ⚠️ Donepezil can prolong succinylcholine action (AChE inhibition affects metabolism). Surgical relevance — disclose donepezil before anesthesia.
  • CYP3A4 strong inhibitors (ketoconazole, ritonavir, clarithromycin): ⚠️ Increased donepezil exposure; consider dose reduction.
  • CYP3A4 strong inducers (rifampin, carbamazepine, St. John's Wort): ⚠️ Reduced donepezil exposure; clinical effect may diminish.
  • CYP2D6 strong inhibitors (paroxetine, fluoxetine, bupropion): ⚠️ Increased exposure. Dylan's V5 backup includes bupropion — if he were ever on donepezil + bupropion, expect higher donepezil levels.
  • NSAIDs (ibuprofen, naproxen): ⚠️ Additive GI bleed risk via cholinergic gastric acid stimulation.
  • Pro-arrhythmic drugs (Class I/III antiarrhythmics, QT-prolonging antibiotics): ⚠️ Additive cardiac risk.

Neutral / safe co-administration

  • DHA, magnesium, vitamin D, vitamin K, NAC, curcumin, rhodiola, theanine, ashwagandha — all stack-safe at standard doses.
  • Melatonin — neutral, may help with donepezil-induced insomnia.
  • Modafinil — no major interaction (different mechanisms; modafinil is mild CYP3A4 inducer, theoretically slightly reduces donepezil exposure but clinically minor).
▸ Drug interactions deep dive
  • CYP2D6 substrate (major). Inhibitors (paroxetine, fluoxetine, bupropion, quinidine, diphenhydramine) raise donepezil exposure. CYP2D6 poor metabolizers (PMs) have ~2× plasma levels at the same dose.
  • CYP3A4 substrate (major). Inhibitors (ketoconazole, itraconazole, ritonavir, clarithromycin, grapefruit juice in large quantities) raise exposure. Inducers (rifampin, carbamazepine, phenytoin, St. John's Wort, modafinil-mild) reduce exposure.
  • Cholinergic synergy: with succinylcholine, neostigmine, pyridostigmine, other AChEIs — cholinergic excess.
  • Beta-blockers, calcium channel blockers (non-DHP, e.g., verapamil, diltiazem), digoxin, cholinergic agents: additive bradycardia.
  • NSAIDs: additive GI bleed risk.
  • Anesthetics (succinylcholine, others): prolonged neuromuscular block.
  • Antipsychotics: ⚠️ small worsening of EPS / parkinsonism in some users on combination (cholinergic-dopaminergic balance shift). Also ↑ neuroleptic malignant syndrome risk (rare).
  • Hormonal contraceptives: no significant interaction.
▸ Pharmacogenomics
  • CYP2D6:
    • Poor metabolizers (PMs, ~7-10% of Whites): ~2× plasma donepezil; higher side-effect rate. Dylan's Nordic/British ancestry: ~7-10% PM prevalence. A 23andMe-derived CYP2D6 phenotype call would inform whether he'd be a PM (relevant if donepezil were ever clinically warranted later in life).
    • Ultra-rapid metabolizers (UMs): lower plasma levels; may need higher doses. ~1-2% of Whites; higher in some Mediterranean populations.
    • Intermediate metabolizers (IMs): between PM and EM; modest effect.
  • CYP3A4 / CYP3A5: less PGx-relevant than 2D6 because most variants are loss-of-function with limited penetrance. CYP3A5*3/*3 (most Whites) is functionally low-3A5; metabolism essentially via 3A4.
  • APOE ε4: carriers may show differential cholinergic responsiveness; mixed data on AChEI response in ε4 vs non-ε4 AD patients. Relevant for Dylan post-23andMe (June 2026) only insofar as it informs decades-out cognitive-decline planning.
  • CHRM, CHRN polymorphisms (muscarinic, nicotinic receptors): theoretical modulators, limited clinical-prediction utility.
  • BCHE polymorphisms: less relevant for donepezil specifically (no BChE activity), more relevant for rivastigmine.
▸ Sourcing deep dive
Path Vendor Cost Reliability Notes
US Rx Generic donepezil 10 mg × 30 (CVS, Walgreens, etc.) ~$15-30/mo with insurance, ~$10-25/mo with GoodRx High Cheapest legal path. Requires Rx from PCP, neurologist, or telehealth. Telehealth Rx for donepezil without an AD diagnosis is essentially impossible — this is not modafinil-style telehealth-friendly.
US Rx (specialty) Aricept brand 10 mg × 30 ~$300+/mo High No clinical reason to use brand over generic.
Indian generic 1mg.com, generic donepezil 10 mg × 30 ~$5-10/mo Medium-high Available without Rx in India; international shipping legal-gray. Lower cost than US generic but with import-customs friction; not worth the friction for a SKIP-FOR-NOW compound.
Gray-market nootropic vendor Various ~$20-40/mo Variable Unnecessary for legitimately-Rx compound at this generic price; signal-to-noise here is bad — buyer-beware.
Off-label US Rx Concierge / longevity-clinic prescriber ~$200-500 visit + Rx cost Variable Some longevity clinics will prescribe donepezil off-label for "cognitive enhancement" or MCI. Not advised for Dylan — cost + workup + ongoing monitoring overhead vs nil real benefit at his profile.

Recommendation for Dylan: do not source. This is a SKIP-FOR-NOW compound. Existing V4 + planned V5 (citicoline 500 mg, alpha-GPC PRN, modafinil, bromantane, alcar, plus Cerebrolysin cycles) covers the cholinergic + cognitive-arousal axis without the side-effect profile.

▸ Biomarkers to track (deep)

Baseline (before starting; not relevant for Dylan unless future use)

  • Resting heart rate + ECG (rhythm, PR interval, QTc) — bradycardia and conduction-disease screen
  • Weight — track for chronic anorexia
  • CBC, CMP, LFTs — general baseline; LFTs because rare hepatotoxicity
  • CYP2D6 genotype — informs PM/IM/UM status (extractable from 23andMe data)
  • Sleep architecture / RBD screen if any history of acting-out dreams
  • MMSE / MoCA / cognitive testing baseline if AD/MCI workup; not relevant for healthy young

During use (if used)

  • Resting HR / orthostatic BP at week 4, 12, then quarterly
  • ECG at 12 weeks if any cardiac symptoms or risk factors
  • Weight monthly
  • Mood / sleep diary for first 8 weeks
  • Cognitive tracking (subjective + standardized if AD)
  • LFTs at 6 months
  • MMSE / cognitive measures every 6 months in AD

Post-discontinuation

  • Monitor for cognitive rebound (AD context) over 4-8 weeks
  • Sleep / mood normalization typically within 2-3 weeks (drug elimination ~2-3 weeks given 70-hr half-life)
▸ Controversies / open debates Live debate
  • Yesavage 2002 — what does it actually mean for healthy young adults? The pilot study is real, statistically clean, but n=18 with mean age 52 (i.e., not 20-somethings, contrary to occasional nootropic-community misreporting). Whether the result generalizes to 20yo brains (where cholinergic signaling is more robust at baseline and ceiling effects are more likely) is untested. My read: signal probably attenuates substantially in young healthy adults; the Grön 2005 fMRI work in n=24 healthy young adults is a partial replication for episodic memory but not for procedural retention specifically.
  • Healthy-adult evidence base is thinner than donepezil's reputation suggests. Most "donepezil enhances cognition in healthy adults" claims in the popular nootropic literature trace back to Yesavage 2002 + Grön 2005 + a handful of smaller mixed/null trials. The asymmetry between massive AD evidence and tiny healthy-adult evidence is striking, especially for a drug that has been on the market 30 years.
  • Disease-modification debate: donepezil is symptomatic, not disease-modifying — no consistent reduction in amyloid burden, tau, or neurodegeneration markers. Some 2020s data on slower hippocampal atrophy on donepezil remains contested. Clinically, treat as symptomatic.
  • MCI prevention: ADCS (Petersen 2005) showed 12-month delay-to-AD-conversion that disappeared by 36 months. Mechanistically interesting but clinically modest; FDA declined to approve for MCI partly on this basis. Several 2020s commentators argue earlier intervention would have shown more benefit; impossible to test now.
  • Bradycardia / sudden death pharmacovigilance signal: real but small; most cases in elderly with comorbidity. For healthy young adults: theoretical but mechanistically credible. Combined with beta-blocker (Dylan's planned propranolol PRN), the additive bradycardia would be a real concern.
  • REM sleep behavior disorder reports: small but consistent signal in pharmacovigilance. Mechanism: cholinergic disinhibition of REM atonia. Particularly relevant for an MMA athlete who shares a bed; non-zero injury risk.
  • Cholinergic-induced depression in healthy users: anecdotal but consistent across long-term donepezil and chronic high-dose alpha-GPC users. Likely a real cholinergic > dopaminergic chronic-tone phenomenon, plausible but not RCT-proven.
  • 23 mg ER (Aricept 23) vs 10 mg: marketed for moderate-severe AD with marginal benefit at substantially higher side-effect cost; widely viewed as a marginal product. Most clinicians stop at 10 mg.
  • Donepezil vs galantamine vs rivastigmine for AD: modest differences. Donepezil: longest half-life, highest selectivity. Galantamine: nicotinic potentiation may add to procedural and dream-related effects (and lucid-dreaming popularity). Rivastigmine: BChE+AChE, patch available (best dose-smoothing). For nootropic use case in healthy young, galantamine is the lower-side-effect, lower-cardiac-risk, lower-Rx-burden alternative.
▸ Verdict change log
  • 2026-05-06 — Initial verdict: SKIP-FOR-NOW. Confidence: HIGH. Reasoning: massive A-tier AD evidence, but vanishingly thin healthy-young evidence (Yesavage 2002 n=18 mid-career pilots + Grön 2005 n=24 — single-domain, never replicated, mean ages above target population). Side-effect profile (GI, vivid dreams + insomnia, bradycardia interacting with planned propranolol PRN, weight loss in MMA athlete, RBD reports, cholinergic-induced mood flattening) is disproportionate to the marginal cognitive benefit in someone whose cholinergic neurons are intact and who already has citicoline + alpha-GPC PRN + planned modafinil + bromantane covering the cognitive-arousal axis. Verdict would change if: (a) a properly-powered (n>100) healthy-young RCT replicates Yesavage in a meaningful cognitive domain (working memory, episodic memory, executive function); (b) Dylan develops a specific procedural-retention use case where the marginal benefit justifies the cost; (c) Dylan ages into MCI / cognitive decline (decades out).
▸ Open questions / gaps Open
  • Is there a properly-powered healthy-young replication of Yesavage anywhere in the literature? Not as of 2026-05. Multiple small (~n=20-30) trials, mixed/null on most domains except episodic memory (Grön 2005).
  • Does donepezil produce measurable cognitive benefit in 20-somethings with intact cholinergic systems, or is the effect floor-only / ceiling-bounded? Untested with adequate power.
  • Does the lucid-dreaming community's experience map onto procedural-skill retention (i.e., is REM enhancement the actual mechanism behind Yesavage's effect)? Plausible but unproven.
  • Subgroup: APOE ε4 carriers under 30 — pre-symptomatic cholinergic support a useful intervention? No data, would be ethically and logistically challenging.
  • Long-term (>5 yr) safety in healthy young adults: essentially none. All long-term data is in elderly AD/DLB.
  • Does the cardiac signal generalize to healthy young adults with normal conduction? Probably small absolute risk, but additive with beta-blockers and unstudied in athletic / high-cardiovascular-load populations.
  • Comparative head-to-head donepezil vs modafinil for sustained cognitive workload in healthy young adults: doesn't exist. Different mechanisms; modafinil is the better-evidenced healthy-young option by orders of magnitude.
  • Whether chronic donepezil produces measurable cholinergic-induced anhedonia or affect blunting in young healthy users: anecdotal-only.
▸ Sources (full, with our context)
Back to compact view