Compact view
Research pass: thorough Supplement · Capsule OPTIONAL-ADD MEDIUM

Carnosic Acid

Extended Research
Extended Research

Our depth — beyond the mirror

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

Our verdict OPTIONAL-ADD MEDIUM

A-tier preclinical neuroprotection across multiple labs and models (5xFAD AD, Parkinson's, ischemia, subarachnoid hemorrhage), elegant proelectrophilic mechanism that activates only in damaged tissue, FDA GRAS, very few side effects at supplement doses, cheap (~$15-30/mo), and stack-clean with V4 curcumin + V5 astaxanthin + NAC. **The catch: zero published human cognitive RCTs exist.** Lipton's 2025 diAcCA paper from Scripps is the highest-profile recent translational push, but human trials haven't started. For Dylan: cheap insurance compatible with the MMA subconcussive-impact thesis, but should be sized as a low-conviction additive layer behind astaxanthin/cerebrolysin/idebenone, not as a primary brain-protection move. Verdict would upgrade to STRONG-CANDIDATE if (a) diAcCA enters and clears Phase 1, (b) NRF2 wild-type genotype confirmed in Dylan, or (c) a clean human cognitive/oxidative-stress RCT lands positive. Would downgrade to SKIP-FOR-NOW if a meaningful hepatotoxicity signal emerges in healthy long-term users at >200 mg/day.

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

    (MEDIUM confidence). Mechanism aligns well — BBB-crossing Nrf2 inducer, anti-inflammatory, mitochondrial-supportive, complementary to V4 curcumin + V5 astaxanthin + idebenone, stack-safe with everything in V4/V5. But: zero published human cognitive RCTs in any age group, and the felt subjective effect at this dose is essentially nil. For Dylan specifically: add as a V5 layer-3 candidate after the V5 core (modafinil/bromantane/Adamax/cerebrolysin/ALCAR/astaxanthin) is established and after the June 2026 23andMe NFE2L2 readout. If NFE2L2 wild-type or heterozygous: cheap insurance for the MMA subconcussive thesis. If homozygous A/A loss-of-function: deprioritize and rely on NAC + curcumin for the same Nrf2 territory. Not a needle-mover; cheap insurance only.

  • 30-50, executive maintenance
    OPTIONAL-ADD

    (MEDIUM confidence). Same logic. The metabolic-syndrome / hepatic-steatosis preclinical signal becomes more relevant in this demographic. Cheap and stack-safe.

  • 50+, mild cognitive decline / neurodegenerative concern
    STRONG-CANDIDATE

    (MEDIUM-to-HIGH confidence). This is the demographic where the preclinical Alzheimer's data and the diAcCA / Lipton 2025 narrative are most directly relevant. Even without human RCTs, the proelectrophilic mechanism + GRAS safety + cheap cost + stack-safe profile justifies daily use as a low-risk neuroprotective add. Particularly compelling in NFE2L2 wild-type + APOE4 carriers (mechanism multiplier). 120-180 mg/day pure CA or equivalent rosemary extract, with breakfast, indefinitely. Pair with curcumin + astaxanthin + omega-3.

  • Anxiety-prone
    NEUTRAL

    No anxiogenic or anxiolytic effect documented. Not a tool for or against anxiety.

  • High athletic load, tested status (MMA, Olympic-style sport)
    OPTIONAL-ADD

    Not on WADA or NCAA prohibited lists (food-derived antioxidant). Antioxidant + anti-inflammatory mechanism relevant for high training volume; mild metabolic-rate / mitochondrial-biogenesis signal in adipocyte models may have minor relevance to body comp goals. No anabolic effect. Pairs with creatine + omega-3 + astaxanthin for layered recovery support. Won't move performance numbers acutely.

  • Sleep-disordered
    NEUTRAL

    No direct sleep effect documented. AM dosing is preferred (with food + fat); evening dosing has been reported anecdotally to cause mild stimulation in a minority.

  • Recovery-focused (post-injury, post-illness, post-stroke)
    STRONG-CANDIDATE

    Anti-inflammatory + mitochondrial-supportive + Nrf2-driven antioxidant defense + GRAS safety. The 2019 subarachnoid hemorrhage data + 2024-2025 ischemia work is the most directly relevant preclinical body. Pair with BPC-157/TB-500 protocols (different mechanism, no interaction).

  • Strength / anabolic-focused
    NEUTRAL

    No anabolic, hormonal, or strength-related effect. Anti-adipogenic preclinical signal is weight-loss-direction, not strength-direction. Won't move strength numbers.

  • Long-COVID / chronic neuroinflammation
    STRONG-CANDIDATE

    preclinically (MEDIUM clinical confidence). The 2022 de Oliveira review (PMC8772720) explicitly framed CA as a candidate for long-COVID via NLRP3 + Nrf2 mechanism. No clinical trials yet, but the mechanism stack matches the pathophysiology well.

Subjective experience (deep)

Subtle. Cumulative. Not a felt nootropic. This is the most "rust-proofing" compound in the entire stack universe — closer in subjective profile to astaxanthin or vitamin E than to anything stimulant- or anxiolytic-class.

  • Onset: No acute felt effect. Plasma peaks around 1-3 hours post-dose; tissue accumulation is gradual; clinical-mechanism endpoints (Nrf2 gene induction) measurable in animal models within days but the human subjective experience is essentially nil over weeks-to-months.
  • Peak / plateau: After 4-8 weeks of consistent dosing, observable changes (if any) are: marginal improvements in oxidative-stress biomarkers (MDA, oxidized LDL), possibly mild reductions in hsCRP, possibly a slight reduction in post-workout muscle soreness if MMA volume is high. Most users notice nothing subjectively and rely on the mechanism + bloodwork to justify continuation.
  • Taper: No withdrawal. Effects fade gradually (weeks) as Nrf2-driven gene-expression changes normalize.
  • What it does NOT feel like: Not a stimulant, not a mood-lifter, not a focus enhancer, not a sleep aid. It is fundamentally invisible while it works.

For Dylan specifically: this is firmly in the "insurance compound" category. The marginal subjective benefit at 60-180 mg/day is realistically ~0. The case for adding it is mechanism + cheap + stack-safe, not "this will feel different."

Tolerance + cycling deep dive
  • Tolerance buildup: None reported clinically or anecdotally. Mechanism is gene-induction-based with negative feedback regulation; chronic dosing maintains a steady-state Nrf2-driven expression elevation rather than pushing receptor downregulation.
  • Recommended cycle: None required. Daily continuous use is the norm in all preclinical chronic studies (typically 3 months in 5xFAD; longer in metabolic models).
  • Reset protocol: Not applicable. If discontinuation is desired, no taper required; effects fade over 2-4 weeks.
Stacking deep dive

Synergistic with

  • curcumin (Dylan's V4 Doctor's Best Curcumin Phytosome 500 mg): Strongly synergistic and the cleanest pairing. Both are Nrf2 activators with similar mechanism — curcumin is a direct electrophile via its α,β-unsaturated diketone and a Keap1 cysteine modifier; CA is the proelectrophilic version. They act on overlapping but non-identical Keap1 cysteine sets (Satoh's work suggests partial overlap). Both are anti-inflammatory via NF-κB suppression. Both are fat-soluble. Both BBB-cross. Layered Nrf2 induction with redundant pathways — reduces single-point-of-failure risk in the Nrf2 pathway. Take together at breakfast with fish oil. The Phytosome formulation handles curcumin's bioavailability problem; CA needs the same fat vehicle.
  • astaxanthin (Dylan's V5 plan, 12 mg AM): Strongly synergistic. Astaxanthin is a membrane-spanning structural antioxidant (polyene chain quenches peroxyl radicals across the bilayer interior); CA is a mobile lipid-soluble antioxidant + Nrf2 inducer. They protect different parts of the membrane network and engage different upstream signaling tiers (astaxanthin also activates Nrf2 but more weakly than CA's proelectrophilic flip). Layered membrane antioxidant + Nrf2 redundancy. Co-administer at breakfast with fat.
  • NAC (Dylan's V4 Swanson 1200 mg/day): Strongly synergistic. NAC is a glutathione precursor (provides cysteine substrate); CA induces glutathione synthesis (via Nrf2-driven GCLC). NAC supplies the building blocks; CA turns up the factory. The combination drives glutathione pool restoration faster than either alone. No interaction; safe co-administration. Already in V4.
  • idebenone (Dylan's V5 contingency, 90-180 mg/day): Mechanistically complementary. Idebenone is a BBB-crossing electron carrier that activates Nrf2 via NQO1; CA is a BBB-crossing proelectrophilic Keap1 modifier. Both end at Nrf2-driven gene induction but enter the pathway at different nodes. Both BBB-cross; both lipid-soluble; both fat-meal absorption. No documented competition.
  • cerebrolysin (Dylan's V5 plan, 5 mL IM × 10-20 days q3mo): Mechanistically complementary, not redundant. Cerebrolysin is a neurotrophic peptide cocktail (BDNF/NGF mimetics); CA is the antioxidant gene-induction layer. Cerebrolysin is the regenerative-protection cycle layer; CA is the daily oxidative-protection floor. Particularly useful for Dylan's MMA subconcussive thesis — cerebrolysin supports plasticity and recovery from impact insult, CA reduces the oxidative cascade that determines how badly an impact damages a neuron in the first place.
  • omega-3 / DHA (Dylan's V4 Carlson DHA Gems): Provides the lipid vehicle for absorption + the membrane substrate that CA helps protect from peroxidation. DHA is the most peroxidation-susceptible fatty acid in human tissue; CA's Nrf2-driven antioxidant battery directly reduces DHA peroxidation. Co-administer at breakfast.
  • vitamin C, vitamin E, alpha-lipoic acid, CoQ10: All members of the broader endogenous antioxidant network that CA upregulates. Layered protection at different membrane zones and aqueous compartments. No documented antagonism.
  • sulforaphane (broccoli sprouts, BroccoMax, etc.): Both are Nrf2 activators but via different mechanisms (sulforaphane is a direct electrophile; CA is proelectrophilic). Theoretically additive, although both alone push Nrf2 strongly enough that the marginal benefit of stacking is unclear. No clinical interaction data.
  • modafinil, bromantane, Adamax/Semax, ALCAR, taurine, apigenin (Dylan's V5 cognitive layer): No mechanism overlap with CA; no documented interactions; safe co-administration.

Avoid stacking with

  • High-dose acetaminophen / chronic heavy alcohol: Hepatotoxicity stacking concern at high CA doses (>300 mg/day). Not Dylan's situation.
  • Narrow-therapeutic-index CYP3A4 substrates at very high CA doses (>500 mg/day): cyclosporine, tacrolimus, certain statins, hormonal contraceptives, some antiarrhythmics. Theoretical only at supplement doses; flag for high-dose users.
  • Other strong Nrf2 activators at supraphysiological doses combined — sulforaphane + bardoxolone + dimethyl fumarate stacking with high-dose CA could theoretically push Nrf2 too hard (chronic Nrf2 hyperactivation has its own concerns: cancer-cell survival effects, immune dampening). Not relevant at standard doses.

Neutral / safe co-administration

  • All V4 stack items: NAC, citicoline, magnesium, phosphatidylserine, rhodiola, theanine, glycine/tryptophan, D3+K2, beta-alanine, creatine, vitamin C.
  • All V5 cognitive additions: modafinil, bromantane, Adamax/Semax, taurine, apigenin, astaxanthin, ALCAR.
  • BPC-157 / TB-500 healing peptides — different mechanism, no interaction documented.
Drug interactions deep dive
  • CYP3A4: Mild in vitro induction + inhibition (substrate-dependent) at micromolar concentrations. Clinical relevance at 60-180 mg/day supplement doses is likely minimal — no documented drug-level interactions in human studies. Flag for narrow-therapeutic-index CYP3A4 substrates only at high doses (>300 mg/day pure CA).
  • CYP2B6: Mild in vitro induction. Minimal clinical relevance at supplement doses.
  • CYP2C9: Mild in vitro inhibition. Theoretical interaction with warfarin (CYP2C9 substrate) at high CA doses; not documented clinically.
  • Hepatotoxic co-medications: Additive hepatic load — flag if Dylan ever uses high-dose acetaminophen, isotretinoin, methotrexate, or alcohol at CA doses >200-300 mg/day. Not his current situation.
  • Anticoagulants / antiplatelets: No documented clinical interaction. Theoretical concern via CYP2C9 (warfarin) at very high doses only.
  • Hormonal contraceptives: Theoretical CYP3A4-induction concern at high doses. Not relevant to Dylan; flag for any female user on OCPs.
  • Iron supplementation: Phenolic compounds bind iron in the gut and can reduce non-heme iron absorption. Take CA / rosemary extract at least 2 hours apart from iron supplements if iron status is borderline.
Pharmacogenomics

This is one of the more interesting pharmacogenomic cases in the supplement space — CA's efficacy depends on the entire Nrf2/Keap1 pathway being intact, and there are several documented polymorphisms that affect that pathway. Dylan's 23andMe results in June 2026 should clarify several of them.

  • NFE2L2 promoter polymorphisms (rs6721961 C>A, rs35652124 G>A). Both are in the NRF2 (NFE2L2) gene promoter region and both reduce Nrf2 transcriptional activity (loss-of-function variants).

    • rs6721961: A allele is associated with high oxidative stress, low antioxidant capacity, insulin resistance, increased diabetes/hypertension risk. Caucasian frequency: ~10% AA homozygous, ~35% CA heterozygous, ~55% CC wild-type.
    • rs35652124: A allele linked to lower Nrf2 expression and more severe inflammatory activity in alcoholic liver disease. Frequencies vary by population.
    • Implication for CA response: carriers of the A alleles at either SNP have reduced baseline Nrf2 induction capacity. The proelectrophilic mechanism still works (Keap1 cysteine alkylation is the same), but the downstream Nrf2 transcriptional output is dampened. Likely smaller effect size in heterozygous carriers; potentially poor responders in homozygous A/A carriers.
    • 23andMe v5 chip coverage of these specific SNPs is variable — may need to derive from imputation or order a follow-up panel (e.g., Promethease analysis of raw 23andMe data, or a dedicated Nrf2-pathway panel).
    • Action for Dylan: When 23andMe lands (~June 2026), pull rs6721961 and rs35652124 status. If wild-type at both: CA case strengthens. If heterozygous: still likely a responder, slightly attenuated. If homozygous loss-of-function at either: deprioritize CA, consider stacking with NAC + glutathione + sulforaphane to maximize what Nrf2 capacity remains.
  • KEAP1 variants. KEAP1 is the cysteine-rich protein that CA alkylates. Loss-of-function KEAP1 variants (rare, mostly in cancer contexts) would actually enhance baseline Nrf2 activity and reduce the marginal benefit of CA. Gain-of-function KEAP1 variants would dampen Nrf2 activation across all electrophiles including CA. Not commonly genotyped; not on 23andMe v5.

  • NQO1 rs1800566 (C609T). Same SNP relevant to idebenone. NQO1 is one of CA's downstream Nrf2 target genes — the T allele (loss-of-function) reduces baseline NQO1 activity, but CA upregulates NQO1 expression so the relative benefit may actually be larger in T-allele carriers (more headroom for induction). Speculative.

  • GSTM1 / GSTT1 deletions. Glutathione-S-transferase null genotypes are common (~50% of Caucasians for GSTM1-null) and modestly increase oxidative stress baseline. CA's Nrf2 induction upregulates GST expression, so GST-null carriers may benefit relatively more from CA supplementation. Speculative; no controlled data.

  • APOE4 carriers. Speculative — Nrf2 dysfunction is one mechanism in APOE4-driven oxidative stress in AD. CA's mechanism is plausibly more useful in APOE4 carriers, but no human stratified data exists. If Dylan's 23andMe returns APOE4: rationale strengthens marginally.

Bottom line: CA is one of a handful of supplements where 23andMe results have a plausible bearing on response prediction. Pull the NFE2L2 promoter SNPs after 23andMe results land before committing to long-term use. The compound is cheap enough that an empirical 8-12 week trial is also defensible regardless of genotype.

Sourcing deep dive
Path Vendor Cost Reliability Notes
OTC Pure Bulk Carnosic Acid 90% (powder) $25-40 / 10 g ($5-15/mo at 60-180 mg/day) medium-high Pure 90% CA powder, cheapest /mg of pure CA. Requires home encapsulation or measured-dose dosing. Best for biohackers comfortable with bulk powder.
OTC Doublewood Rosemary Extract (standardized to ≥20% carnosic acid, 500 mg caps) $20-30/120 caps ($10-15/mo at 600-900 mg extract = 120-180 mg CA) high Doublewood is reliable for this category; transparent COA program; established US vendor. Top pick for Dylan if he wants encapsulated convenience.
OTC Nature Restore Rosemary Extract 20% Carnosic Acid (500 mg, 90 caps) $15-25/bottle ($8-15/mo) medium-high Amazon-available, US-manufactured, non-GMO. Good budget option.
OTC Life Extension Carnosol/Carnosic Acid Rosemary Extract $25-35/60 caps ($15-25/mo) high Life Extension is reliable across the board; standardized formulation; likely dual-standardized to both CA and carnosol. Premium.
OTC Nootropics Depot Rosemary Extract 20% CA (NootropicsDepot.com) $15-25/mo high Nootropics Depot's standard third-party-testing program; reliable.
OTC NOW Foods Rosemary Extract (iHerb) $8-15/mo medium-high USP-style budget pick; CA standardization may be lower (~5-10%) — verify label.
OTC Fingredients Pure Carnosic Acid 90% (specialty) varies medium Pure CA powder option; less well-known vendor.
Bulk / industrial Jeeva Organic, Kingherbs, Naturex (food-additive grade rosemary extract) per-kg pricing for B2B varies Industrial rosemary extract (Type I/II per food regulation, standardized to CA + carnosol). Not retail-friendly but worth noting that the CA in supplements ultimately traces back to this supply chain.
AVOID "Rosemary essential oil" sold as cognitive supplement varies low Essential oil contains volatile monoterpenes (1,8-cineole, α-pinene, camphor) — NO carnosic acid in essential oil. Different chemistry, different mechanism. Don't confuse the categories.

Dylan's recommendation: Doublewood Rosemary Extract 20% CA (500 mg caps), 1-2 caps/day with breakfast = 100-200 mg CA daily, ~$10-15/mo on Amazon. Fits the existing Amazon V4 channel; reliable vendor; convenient encapsulation. Alternative: Pure Bulk 90% CA powder if he wants the cheapest /mg-pure-CA option and is comfortable with powder dosing (~$5-10/mo).

Biomarkers to track (deep)
  • Baseline (before starting):

    • ALT, AST, GGT — hepatic enzymes (the relevant watch panel at any sustained CA dose).
    • Lipid panel (TC, LDL, HDL, TG) — for tracking the metabolic / anti-adipogenic preclinical signal.
    • hsCRP — inflammation marker; CA's anti-inflammatory mechanism predicts modest decrease.
    • Optional / specialty: MDA (malondialdehyde), oxidized LDL, total antioxidant capacity (TAC), glutathione (reduced + oxidized) — better oxidative-stress proxies; usually only via specialty panels.
    • NFE2L2 rs6721961 + rs35652124 — Nrf2 promoter genotype; pharmacogenomically informative.
    • NQO1 rs1800566 — downstream Nrf2 target; co-relevant with idebenone decisions.
    • APOE genotype — speculative response modifier.
    • Subjective: cognitive-fatigue rating at end of 6-12 hr workday (1-10); post-training muscle soreness rating; digital-eye-strain rating.
  • During use (every 6-12 months):

    • ALT/AST/GGT at week 8-12 if dosing >200 mg/day pure CA chronically; otherwise routine semiannual.
    • hsCRP, lipid panel — looking for modest improvements at 12-24 weeks.
    • MDA / oxidized LDL / TAC if specialty panel available — looking for delta after 12 weeks.
    • Subjective ratings (rough delta tracking).
  • Post-cycle (if cycled): N/A — no cycling.

For Dylan specifically: tie this into the June 2026 baseline panel. ALT/AST/GGT are already standard liver-panel components. NFE2L2 SNPs derive from 23andMe (verify chip coverage; flag for follow-up imputation panel if needed).

Controversies / open debates Live debate
  1. "Lipton 2025 will go to human trials soon" — wait-and-see. The Scripps press release is enthusiastic about fast-tracking diAcCA via the GRAS pathway, but as of May 2026 no IND filing or Phase 1 trial registration has been announced publicly. Pharmaceutical translation of preclinical leads is slow even with GRAS substrates; expect 2-5 years before any meaningful clinical readout. Don't price in human efficacy data that doesn't exist yet.

  2. "Rosemary improves memory" misattribution. The Moss et al. studies (2003, 2012) showing inhaled rosemary essential oil improves memory test scores are frequently cited as "rosemary cognitive evidence" in support of CA supplementation. This is a category error. Essential oil contains volatile monoterpenes (1,8-cineole, α-pinene, camphor) — there is essentially zero carnosic acid in essential oil. CA is a non-volatile diterpene that lives in the leaf's lipid fraction, not the volatile fraction. Essential-oil cognitive evidence does not transfer to CA-standardized rosemary extract supplements. Flag this for accuracy in the V6 wiki and any user-facing copy.

  3. Human translation of preclinical data is genuinely thin. Despite strong, replicated preclinical neuroprotection across multiple labs and disease models (Schubert, Lipton, Satoh, multiple Parkinson's labs), there are zero published placebo-controlled human cognitive or AD trials of CA or standardized CA-rich rosemary extract. The supplement is sold on mechanism + GRAS safety + Lipton's recent paper, not on clinical efficacy. Buyer should know this.

  4. Bioavailability data is mostly rat. The 14-65% F% numbers come from rat oral PK studies. Human PK data is sparse and mostly derived from rosemary-extract studies that included multiple polyphenols. The Lipton group's diAcCA pro-drug specifically addresses CA's bioavailability/stability problem, but pure CA's behavior in humans is not as well-characterized as for, e.g., curcumin or astaxanthin.

  5. Standardization variability. "Rosemary extract" supplements range from 5% to 90% carnosic acid by weight. Without checking the label and ideally a third-party COA, the actual CA dose in a given supplement is often unclear. Always look for "standardized to X% carnosic acid" with a specified percentage. The 7% / 20% / 25% / 60% / 90% products are the most common tiers.

  6. In-vitro hepatotoxicity vs. human GRAS status. Carnosic acid shows dose-dependent hepatotoxicity in vitro (EC50 ~95 μM in human hepatocytes; Aruoma 2012) but is FDA GRAS as a food preservative and has no human hepatotoxicity reports at supplement doses. The disconnect is dose-relevance: the in-vitro toxicity is at concentrations far above plasma levels achievable from oral supplementation. The pragmatic interpretation is: 60-180 mg/day is safe; 300+ mg/day pure CA chronically deserves ALT/AST monitoring; 500+ mg/day chronically is in poorly-characterized territory.

  7. "Pro-electrophilic = uniquely targeted" — partly marketing. The PED concept is real and clever, but in practice CA does not exclusively activate in damaged tissue — there is some baseline Nrf2 induction in healthy tissue too (oxidative stress is a continuum, not a binary). The framing of "only active in damaged tissue" is a useful approximation but not a literal truth. The therapeutic-window advantage over direct electrophiles is real but modest.

  8. "Carnosic acid vs carnosol" confusion. Carnosol is an oxidized derivative of CA that's also bioactive and shares many of CA's properties. Many "rosemary extract" supplements are dual-standardized to CA + carnosol. Mechanistically they overlap but are not identical (carnosol has somewhat different Keap1 cysteine targeting). For most practical purposes, treating them as equivalent is fine.

  9. Whether the proelectrophilic mechanism survives at supplement doses. Most mechanism papers use cell-culture concentrations in the low-μM range. Whether human plasma + brain CA concentrations from 60-180 mg/day oral dosing actually reach the threshold for the quinoid switch in damaged tissue is not directly demonstrated. Animal PK data suggests yes; human data is thin.

Verdict change log
  • 2026-05-05 — Initial verdict: OPTIONAL-ADD (MEDIUM confidence). Justification: A-tier preclinical neuroprotection (Schubert/Lipton lineage; Satoh 2008 mechanism; Lipton 2025 5xFAD diAcCA; multiple stroke/SAH/Parkinson's models); B-tier metabolic preclinical; C-tier general antioxidant in healthy humans; zero published human cognitive RCTs; FDA GRAS safety; cheap (~$15-30/mo); stack-clean with V4 curcumin + V5 astaxanthin + NAC + idebenone + cerebrolysin; mechanism aligned with Dylan's brain-priority + MMA subconcussive-impact thesis. Plan: defer first trial until V5 core is stable (week 12+) and 23andMe NFE2L2 readout is in (~June 2026). At that point: 120-180 mg/day pure CA with breakfast/lunch from Doublewood or Pure Bulk, no monitoring required at this dose tier, decide on continuation at 12 weeks based on subjective + bloodwork delta. Verdict would upgrade to STRONG-CANDIDATE if (a) diAcCA enters and clears human Phase 1, (b) NFE2L2 wild-type confirmed in Dylan, or (c) clean human cognitive/oxidative-stress RCT lands positive. Would downgrade to SKIP-FOR-NOW if hepatotoxicity signal emerges in healthy long-term users at >200 mg/day or NFE2L2 homozygous loss-of-function confirmed without rescue stack benefit.
Open questions / gaps Open
  1. No published human cognitive RCT of CA or standardized rosemary extract. Zero. This is the single largest evidence gap. Mechanism is strong, preclinical is strong, human cognitive data is absent. A 12-week placebo-controlled RCT in healthy adults at 120-180 mg/day with cognitive endpoints (working memory, processing speed, attention) + oxidative-stress biomarkers would be high-leverage.
  2. diAcCA Phase 1 timing. When (and whether) the Lipton/Scripps diAcCA pro-drug enters human trials. IND filing not yet announced as of May 2026.
  3. NFE2L2 genotype-stratified response data. No published study stratifies CA response by NFE2L2 promoter SNP status. Would make CA a precision-medicine compound rather than an empirical guess.
  4. Effect on biomarkers of repeated subconcussive impact (NfL, S100B, GFAP) in contact-sport athletes. High-leverage open question for Dylan specifically. Not yet studied. Same gap as astaxanthin and idebenone.
  5. Long-term safety data in healthy 60-180 mg/day users at 5+ years. Thin database; supplement-community track record is reassuring but not formal.
  6. Tissue saturation kinetics in human brain at supplement doses. Whether 60-180 mg/day CA achieves preclinical-effect tissue concentrations in human cortex/hippocampus is unconfirmed; rat data extrapolated.
  7. Whether CA + curcumin + astaxanthin layered Nrf2 induction is genuinely additive or saturating. No combinatorial human study exists. Mechanistically additive; clinically unverified.
  8. Whether the proelectrophilic "damage-targeted" mechanism survives at supplement doses in human tissue. Plausible from rat PK + cell mechanism, not directly demonstrated.
  9. CA in MMA / boxing / contact-sport contexts specifically. No direct study. The Dylan-archetype use case is mechanism-justified, not empirical.
Sources (full, with our context)

Pivotal recent (2025) clinical / translational

Foundational mechanism

Disease-model preclinical

Pharmacokinetics + bioavailability

Safety / toxicity

Pharmacogenomics

Metabolic / anti-obesity preclinical

Sourcing

Encyclopedia cross-reference

  • ../NOOTROPICS-ENCYCLOPEDIA-2026-05-05.md Section 29 — original "Cheap optional V5 antioxidant add" verdict; this compound file deepens and refines it with the 2025 Lipton/diAcCA story + pharmacogenomic detail + sourcing matrix.

Cross-link compounds

  • astaxanthin.md — companion membrane antioxidant; strongly synergistic at breakfast; Nrf2 redundancy.
  • idebenone.md — BBB-crossing mitochondrial Nrf2 activator at a different pathway node; complementary stack.
  • cerebrolysin.md — neurotrophic peptide cycle; mechanism-complementary protection layer for the MMA subconcussive thesis.
  • curcumin.md — closest mechanistic relative in V4 (direct-electrophile Nrf2 inducer); CA = proelectrophilic version of the same axis; layered Nrf2 induction.
Back to compact view