◈ Nootpedia
Compact view
Research pass: thorough Peptide · Intranasal WATCH-LIST MEDIUM

Intranasal Insulin

Extended Research
◈ Extended Research ◈

Our depth — beyond the mirror

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

▸ Our verdict WATCH-LIST MEDIUM

Real signal in MCI/AD (Craft 2012) and healthy adults (Benedict 2004/2007) — but SNIFF 2020 missed primary endpoint largely due to a documented delivery-device failure mid-trial (the ViaNase II device performed differently than the original ViaNase I), confounding interpretation. Mechanism is well-characterized and side-effect profile is exceptionally clean for a chronic CNS-active intervention. For Dylan: not actionable until (a) 23andMe APOE status is in (APOE4 carriers show stronger cognitive response signals), (b) a cleaner delivery path exists than off-label compounded insulin + research-grade atomizer, (c) some direct healthy-young evidence beyond the small Benedict cohorts. Verdict would upgrade to OPTIONAL-ADD if APOE4 + Kurve VP3 or comparable device becomes accessible. Verdict would downgrade to SKIP-FOR-NOW if SNIFF-2 fails replication in 2027.

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

    Real signal in healthy adults (Benedict series) but small effect size relative to lower-friction options (Adamax, Semax, citicoline, modafinil). Sourcing complexity is high. Strongest argument is long-term hippocampal preservation + APOE4 risk mitigation if applicable. Wait for 23andMe APOE status and revisit. If APOE4 carrier → upgrade to OPTIONAL-ADD. If APOE non-4 → remain WATCH-LIST or downgrade.

  • 30-50, executive maintenance
    OPTIONAL-ADD

    Brain insulin sensitivity declines with age; midlife is the window where IN insulin's preventive case strengthens. Modest evidence, low side effect ceiling, sourcing solvable for motivated users.

  • 50+, mild cognitive decline
    STRONG-CANDIDATE

    Direct target population for Craft 2012 + Brünner 2015. APOE4 carriers especially. Strong rationale; sourcing path through Bredesen-trained clinicians or functional medicine.

  • 50+ with diagnosed MCI or early AD
    STRONG-CANDIDATE

    → near-PRIMARY (with clinical supervision). Most evidence-supported population. Should be discussed with treating clinician.

  • Anxiety-prone
    NEUTRAL

    Mild anxiolytic / cortisol-dampening effect documented in healthy adults. Not a primary anxiety treatment but mild positive layer.

  • High athletic load, tested status
    OPTIONAL-ADD

    WADA: insulin is on the prohibited list (S4.5 — hormone and metabolic modulators) for athletes. Tested athletes should NOT use IN insulin. Untested (Dylan): irrelevant.

  • Sleep-disordered
    NEUTRAL

    No documented sleep effect.

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

    Not a primary recovery tool; brain-side benefit only.

  • Strength/anabolic-focused
    NEUTRAL

    Tiny systemic absorption is irrelevant for muscle anabolism. IN insulin is NOT a stealth anabolic.

  • Diabetic / pre-diabetic
    CAUTION

    Cognitive benefit signal is real but interaction with diabetic medication and metabolic state is more complex. Clinical supervision required.

  • APOE4 carriers (any age)
    OPTIONAL-ADD

    Strongest theoretical and empirical case for long-term IN insulin as APOE4-specific prevention. Most-discussed in Bredesen-protocol context.

▸ Subjective experience (deep)

Onset: 15-30 minutes after intranasal dose.

Peak: 30-60 minutes. Acute subjective duration is short — 1-2 hours of perceptible cognitive shift. Memory consolidation effects (the actual mechanism of benefit) are not directly subjective and accumulate over days/weeks.

Characteristic effects (per Benedict, Hallschmid, and limited self-experimenters):

  • Subtle mood lift — described as "calm-positive," not stimulating
  • Slight reduction in appetite / food cravings (food-reward modulation)
  • Improved memory access — anecdotally "names come faster," "I retain what I just read"
  • Mild reduction in stress reactivity
  • Notably absent: stimulant signature, jitter, sleep impact, energy surge

Honest variability:

  • Effects are quantitatively smaller and qualitatively subtler than peptide nootropics like Semax. Most healthy users will not "feel something" acutely. The benefit is detected on memory tests, mood scales, and FDG-PET — not in subjective bright signal.
  • This is a classic "you'll see it on the data, not in the mirror" intervention.
  • For Dylan-archetype (already cognitively optimized 20yo): expected effect size is small. Strongest argument for inclusion is brain preservation / hippocampal support over years, not acute cognitive lift.
▸ Tolerance + cycling deep dive
  • Tolerance buildup: Unknown for healthy users. Trials run up to 12 months continuous daily use without documented tolerance to cognitive endpoints. Theoretical hypothalamic receptor downregulation possible but not demonstrated.
  • Recommended cycle: No established cycling protocol. Trials are continuous daily. Conservative biohacker approach: 8-12 weeks on, 2-4 weeks off — but no evidence base for this.
  • Reset protocol: N/A — no documented dependence or rebound.
▸ Stacking deep dive

Synergistic with

  • citicoline (CDP-choline) — Cholinergic precursor complements insulin's effect on synaptic plasticity. Both upregulate hippocampal function via different pathways. Already in Dylan's V4 (500 mg/day).
  • omega-3 DHA — Membrane fluidity supports insulin receptor function; well-established neurotrophic synergy. Already in Dylan's V4 (2 g DHA/day).
  • Mg L-threonate — Hippocampal-targeted Mg supports NMDA function; complementary to insulin's plasticity signaling. Already in Dylan's V4.
  • semax / n-acetyl-semax-amidate / adamax — Different mechanism (BDNF/NGF transcriptional upregulation vs insulin receptor signaling). Theoretically synergistic for hippocampal support; no published combination data.
  • cerebrolysin — Both target neurotrophic / synaptic plasticity pathways. Theoretical synergy.
  • oxytocin (intranasal) — Different receptor system but similar nose-to-brain delivery considerations; some users alternate or stack for separate effects (insulin = memory, oxytocin = social/mood).

Avoid stacking with

  • Systemic insulin therapy (diabetic users only) — additive glycemic effect. Not relevant for Dylan.
  • High-dose intranasal peptides on the same morning — practical concern: only one nose-to-brain protocol can target the olfactory cleft cleanly per session. If stacking with Semax/Adamax, alternate nostrils OR alternate days OR space ≥30 min apart with full nasal recovery between.
  • Other intranasal corticosteroids or decongestants — alter mucosal physiology, may impair olfactory deposition. Avoid concurrent use during IN insulin protocol.

Neutral / safe co-administration

  • Dylan's V4 base stack (NAC, magnesium, fish oil, PS, rhodiola, theanine, B-complex)
  • Creatine
  • Caffeine
  • Modafinil
  • Bromantane
  • ALCAR, Apigenin, Astaxanthin
▸ Drug interactions deep dive
  • No CYP interactions — peptide hormone metabolized by peptidases, not CYP enzymes.
  • Theoretical: additive with systemic insulin — irrelevant for Dylan (no diabetes).
  • Theoretical: additive with insulin sensitizers (metformin, berberine, GLP-1 agonists) — at typical IN insulin doses, systemic absorption is too small for meaningful interaction. Not a practical concern.
  • No documented interactions with hormonal contraceptives, antidepressants, antipsychotics, stimulants, or other typical CNS drugs.
▸ Pharmacogenomics
  • APOE genotype (rs429358 + rs7412) is the single most important variant.
    • APOE3/3 (most common): typical response to IN insulin in trials.
    • APOE3/4 or APOE4/4: differential response — depending on study, either enhanced acute cognitive response (Reger 2008 acute) or attenuated chronic response (Craft 2012 subgroup). The interaction is real but direction depends on context (acute vs chronic, healthy vs MCI).
    • APOE2 carriers: under-represented in trials, response unclear.
    • Practical implication: Dylan's 23andMe results (June 2026) include APOE genotype. Don't act on IN insulin until known.
  • Insulin receptor variants (INSR polymorphisms) — theoretically relevant but not characterized in IN insulin response.
  • TOMM40 + APOE haplotype (rs10524523) — affects AD risk and cognitive trajectory; relevance to IN insulin response unstudied but theoretically aligned with APOE story.
  • CYP variants: irrelevant (insulin is not CYP-metabolized).
▸ Sourcing deep dive
Path Vendor Cost Reliability Notes
US Rx insulin (preferred) GoodRx / Costco / WalMart for Humulin R or NovoLog $25-50 / 10mL vial U-100 (3-6 month supply at 20 IU/day) High Requires US Rx. Telehealth path: hard — most prescribers won't write insulin off-label for cognition. Functional medicine / Bredesen-trained prescribers more likely.
OTC insulin (state-dependent) WalMart ReliOn Humulin R (R brand only, not analogs) ~$25 / 10mL vial U-100 High Available OTC in: IN, LA, ME, MS, MO, OH, OR, PA, RI, TN, WY, WI. NOT available OTC in CA or most blue states. Verify your state's specific pharmacist policy — ReliOn brand specifically is the OTC product.
Compounded IN insulin formulation Empower Pharmacy, Olympia Pharmacy, Hallandale Pharmacy ~$80-150 / month Medium-High Compounded with metered nasal atomizer attached. Requires US Rx via partnered telehealth provider. Cleanest path if you can find a willing prescriber.
Precision intranasal atomizer Aptar Pharma OptiMist (research grade) ~$200-400 device Medium Research-only access; not retail.
Precision intranasal atomizer Kurve ViaNase / VP3 Not commercially available Research SNIFF trial device. Manufactured by Kurve Therapeutics for clinical research only.
Affordable atomizer (compromise) Teleflex MAD Nasal (Mucosal Atomization Device) $5-10 per device, multi-pack on Amazon Medium Clinical emergency atomizer. Designed for systemic absorption (e.g., naloxone) not specifically olfactory deposition, but provides a fine droplet spray and is the most accessible option for self-experimenters. Best practical option for non-clinical biohackers but with caveat that olfactory deposition is unverified.
DIY pipette / dropper (do not recommend) $1 Low Drops will deposit on lower turbinate / drain to throat. Wastes most of the dose, gets some systemic absorption, minimal CNS delivery. Don't bother.

Sourcing reality for Dylan

  • The hard part is not insulin — it's the device. Insulin itself is cheap and accessible (Walmart ReliOn $25/vial OTC in 12 states; otherwise telehealth Rx). The bottleneck is the precision atomizer that actually targets the olfactory cleft.
  • Best practical path: MAD Nasal atomizer + ReliOn or compounded IN insulin formulation. Acknowledge that olfactory deposition is unverified — you're getting "good enough" delivery, not research-grade.
  • Best research-grade path: Compounding pharmacy (Empower, Olympia) with prescribed IN insulin formulation including device. Requires sympathetic prescriber. Cost ~$80-150/mo.
  • Functional medicine clinics offering Bredesen protocol sometimes include IN insulin — could be a path if Dylan ever pursues clinical-supervised cognitive optimization.
  • Why this is hard: SNIFF trial itself failed in part because the device industry hasn't solved olfactory-targeted nasal delivery for general use. This is a real engineering / regulatory bottleneck, not a sourcing convenience problem.
▸ Biomarkers to track (deep)

Baseline (before starting)

  • APOE genotype (rs429358, rs7412) — non-negotiable; single most important variant
  • Fasting glucose + HbA1c — confirm normal metabolic state
  • Fasting insulin + HOMA-IR — baseline insulin sensitivity
  • Working memory: digit span, dual n-back baseline
  • Declarative memory: paragraph recall (Logical Memory subtest from WMS-IV style)
  • Word-list recall (15 words, immediate + 30-min delayed)
  • Mood VAS daily for 2 weeks
  • Subjective focus / cognitive clarity VAS daily for 2 weeks
  • Optional: serum BDNF baseline

During use

  • Same VAS daily (mood, focus, cognitive clarity)
  • Working memory weekly
  • Word-list recall weekly
  • Nasal mucosa state (irritation, congestion frequency)
  • CGM if available — confirm no meaningful glycemic effect
  • Body weight (food-reward effect could shift intake; Hallschmid signal is mild)

Post-cycle (if cycled)

  • Repeat baseline cognitive measures 2-4 weeks post-cessation
  • Compare on vs off delta — actual decision-quality data
▸ Controversies / open debates Live debate

  1. SNIFF 2020 interpretation. Field is split between "IN insulin failed in Phase 3" (skeptics) and "SNIFF was a device-failure trial, not a mechanism failure" (Craft camp). I weight the second interpretation more heavily because the device-change confound is documented in the published paper, the pre-device-change subgroup showed signal, and the mechanism is well-supported by Phase 1/2 + healthy-adult data. But honest acknowledgment: a "real" Phase 3 with consistent device hasn't been completed yet.

  2. Effect size in healthy young users. Benedict 2004 + 2007 are positive but small (n=20-40) and use 4×/day dosing impractical for daily life. Whether 1× or 2× daily produces meaningful cognitive benefit in already-optimized 20yo is unclear. Honest expectation: small effect, may not be subjectively detectable, primary value is brain preservation over years.

  3. APOE4 interaction direction. Reger 2008 showed APOE4 carriers had stronger acute response; Craft 2012 showed APOE4 carriers had weaker chronic MCI response. These could be reconciled (acute receptor sensitivity vs chronic compensation) but is not fully resolved. APOE4 status modifies expected response but the direction is context-dependent.

  4. Brain insulin resistance hypothesis. "Type 3 diabetes" framing is popular but not universally accepted. If brain insulin resistance is the core AD pathophysiology, IN insulin is a near-causal treatment. If it's a downstream marker, IN insulin is symptomatic relief. The honest answer is probably "partial cause + downstream marker" — meaningful but not the only mechanism.

  5. Cycling: theater or necessary? Trials ran continuous daily for 4-12 months without documented tolerance. Conservative biohacker cycling (8 weeks on, 2-4 weeks off) has no evidence base — neither for nor against.

  6. Sublingual / buccal alternatives? Some functional medicine protocols use sublingual insulin formulations as an "easier delivery" alternative. Mechanism is different (sublingual = systemic absorption with some retrograde CNS exposure via vascular routes, not olfactory). Evidence base is thinner. Not recommended over IN.

  7. Do we even need Suzanne Craft's specific protocol? Some self-experimenters use much higher doses (60-80 IU) or much lower (10 IU). Dose-response is poorly characterized outside the 20-40 IU range used in trials. The 40 IU = inferior-to-20 IU finding from Craft 2012 suggests possible inverted-U; experimenters going higher may be missing the peak.

▸ Verdict change log
  • 2026-05-06 — Initial verdict: WATCH-LIST for Dylan. Real mechanism and real cognitive signal in MCI/AD and healthy adults, but (a) effect size in already-optimized 20yo unclear, (b) sourcing is hard given off-label insulin Rx + research-grade atomizer requirements, (c) APOE genotype unknown until June 2026 23andMe results, (d) lower-friction higher-evidence options (Adamax, Semax, citicoline, modafinil, cerebrolysin) cover similar terrain in his V5 stack. Verdict would upgrade to OPTIONAL-ADD if APOE4 carrier + accessible Kurve VP3 / Empower compounded path. Verdict would downgrade to SKIP-FOR-NOW if SNIFF-2 fails replication in 2027 or if Adamax + Semax combination delivers measurable signal first.
▸ Open questions / gaps Open
  1. No direct healthy-young dose-response trial. Benedict series uses one dose schedule (40 IU 4×/day) in n=20-40. Optimal dose for healthy young brain-priority user is extrapolated, not evidenced.
  2. No long-term (>12 month) safety in healthy users. Theoretical concern about hypothalamic receptor adaptation is unresolved.
  3. SNIFF-2 outcome (expected 2026-2027) is the field's pivotal data point. If positive with consistent device → IN insulin moves from WATCH-LIST to OPTIONAL-ADD broadly. If negative → field reconsiders.
  4. No head-to-head with peptide nootropics (Semax, Adamax) or with cerebrolysin. All comparison is speculation.
  5. Optimal device for non-clinical use. Practical engineering gap — research-grade atomizers are not commercially available; commercial atomizers are not validated for olfactory deposition.
  6. APOE4 interaction direction in healthy young users. Mostly extrapolated from older / MCI cohorts. Direct healthy-young APOE-stratified trial doesn't exist.
  7. Sex differences. Hallschmid food-reward work suggests women > men response on appetite endpoints. Whether sex effects extend to memory/mood endpoints is unclear.
▸ Sources (full, with our context)

Primary literature — Craft / Wake Forest

Primary literature — Hallschmid / Lübeck

Primary literature — Reger et al

Mechanism / nose-to-brain transport

Meta-analysis + reviews

Device / delivery technology

APOE / pharmacogenomics

Regulatory / safety

Sourcing references

Back to compact view