CogniMary ingredients, one by one
Fifteen ingredients. Four of them have a number on the label. The other eleven share one. Here is what each was studied at, what the research found, and how much of it can be in the bottle.
CogniMary names fifteen ingredients. Folate, vitamin B12, magnesium and chromium are quantified individually. Cinnamon, ginkgo, bilberry, bacopa, green tea, berberine, L-theanine, alpha lipoic acid, GPC choline, lutein and zeaxanthin share a 635 mg proprietary blend.
Of the eleven, six cannot reach the dose used in the research behind them, and the two that comfortably can — lutein and zeaxanthin — are the two whose large cognition trial found nothing.
The fifteen, at a glance
The four quantified nutrients are covered on the supplement facts page, because they are label questions rather than blend questions. Everything below is about the eleven.
Where the ceilings come from →The four quantified nutrients →
Cinnamon extract
The heaviest ingredient in the blend, and the one with the least to do with memory.
- Printed position #1 of 11
- Ceiling 635.0 mg
- Studied at 1,000 mg a day
- Studied for blood sugar and blood lipids, not memory
Cinnamon sits first on the printed list, which means it is the heaviest thing in the 635 mg. Its ceiling is therefore the whole blend, and even that is below the 1,000 mg used in the trial we read.
That trial was a randomised, double-blind, controlled study of Ceylon cinnamon extract in 150 adults over twelve weeks. It measured lipids and glucose, not cognition. LDL cholesterol fell by about 6 mg/dL and the change was not statistically significant. Fasting blood sugar fell by about 8.6 mg/dL and that change was, with a stronger effect in participants who had type 2 diabetes.
So cinnamon is a defensible ingredient in a metabolic formula. In a formula sold for memory and focus, it is the heaviest component and the least connected to the claim.
Ginkgo biloba
The most studied ingredient on this label, and the one with the most discouraging large trial behind it.
- Printed position #2 of 11
- Ceiling 317.5 mg
- Studied at 120 to 240 mg a day
- Studied for cognition, with the largest trial finding nothing
Ginkgo sits second, so its ceiling is 317.5 mg — comfortably above the 120 to 240 mg used in the cognition literature. It is one of the few ingredients here that could be present at a researched dose.
What the research found is another matter. The Ginkgo Evaluation of Memory study randomised just over 3,000 adults aged 72 to 96 to 120 mg of standardised extract twice daily or placebo, and followed them for years. Its published conclusion is blunt: ginkgo at that dose “did not result in less cognitive decline in older adults with normal cognition or with mild cognitive impairment”.
The 2009 Cochrane review of ginkgo for cognitive impairment and dementia reached a similar place from a different direction, describing the evidence of predictable and clinically significant benefit as inconsistent and unreliable.
We include ginkgo's positive framing nowhere on this site, because the largest and best-funded trial of it found nothing. A seller is entitled to disagree with that reading. A buyer is entitled to know it exists.
Bilberry extract
Named twice on the panel, with two different plants.
- Printed position #3 of 11
- Ceiling 211.7 mg
- Studied at 160 to 480 mg a day of standardised extract
- Studied for night vision, and not much else with good evidence
The panel prints Bilberry (Vaccinium cyanococcus fruit) Extract. Those two names do not go together. Bilberry is Vaccinium myrtillus, a European species. Cyanococcus is not a species at all; it is the section of the genus Vaccinium that contains the North American blueberries.
The two are distinguishable, and not only in a laboratory. In bilberry the anthocyanin pigment runs through both skin and flesh; in blueberry it is concentrated in the skin, leaving the flesh pale. That difference is used to detect blueberry material passed off as bilberry extract.
We cannot tell from a label which is in the capsule, and we are not going to assume the worse answer. What we can say is that the panel gives one plant's common name and another plant's botanical address, and that an identity test would settle it.
On evidence: the best-studied use of bilberry is night vision, and even there the trial record is contested. Human data on bilberry for cognition is scarce. Doses in the literature run from 160 to 480 mg of standardised extract; the ceiling at position three is 211.7 mg, so the bottom of the range fits.
Bacopa monnieri
The ingredient whose standardisation the label gets exactly right, and whose dose it cannot reach.
- Printed position #4 of 11
- Ceiling 158.8 mg
- Studied at 300 mg a day for at least twelve weeks
- Studied for speed of attention, with memory effects called inconclusive
The panel prints [NLT 20% Bacosides] — not less than twenty percent bacosides. That is the standardisation used in the trials, and printing it is a genuinely good sign. It means somebody in the formulation chain knew what the research used.
The meta-analysis pooled randomised controlled trials totalling 437 participants. Its reference dose was 300 mg a day and its inclusion criterion was twelve weeks or more, because bacopa is not an acute ingredient. It found an improvement in speed of attention — choice reaction time improved by around 10.6 milliseconds at the 300 mg dose — and described the memory findings as inconclusive because the trials varied too much.
Bacopa sits fourth here. Its ceiling is 158.8 mg, a little over half the dose the meta-analysis is built on. The right standardisation at the wrong amount is a frustrating combination, because it is so nearly right.
Green tea extract
The only ingredient on this label where the firm published number is a safety ceiling rather than a dose.
- Printed position #5 of 11
- Ceiling 127.0 mg
- Studied at no cognition dose we could verify
- Studied for many things; the firm number here is a safety ceiling, not a dose
The panel prints [NLT 50% Epigallocatechin gallate], so at least half the green tea extract by weight is EGCG. Since the extract's own weight is not disclosed, half of an unknown is still an unknown.
We could not verify a cognition dose for EGCG from a primary source we were able to open, so this ingredient gets no verdict on the ceiling chart rather than a guessed one. What we could verify is a safety position.
The European Food Safety Authority reviewed green tea catechins in 2018 and found that doses of 800 mg of EGCG a day or above, taken as a food supplement, produced a statistically significant rise in serum transaminases — a marker of liver stress. EFSA drew a distinction between supplement boluses and traditional tea infusions, judging the latter safe at typical consumption.
Green tea sits fifth, ceiling 127 mg, so the EGCG in a serving is at most about 64 mg and almost certainly less. That is a long way below the level EFSA flagged. The point is not that this product is dangerous; it is that a label which does not disclose an EGCG amount leaves a buyer unable to add it to whatever else they take.
Berberine HCl
A metabolic ingredient in a cognitive formula, at a fourteenth of its trial dose.
- Printed position #6 of 11
- Ceiling 105.8 mg
- Studied at 500 mg three times a day, so 1,500 mg a day
- Studied for blood sugar and cholesterol; no human cognition evidence
Berberine is studied at 500 mg three times a day — 1,500 mg daily — in trials of glycaemic control and lipid lowering. It is a serious ingredient with a substantial metabolic literature behind it.
For cognition, the Alzheimer's Drug Discovery Foundation's researcher summary is explicit: there is no clinical evidence that berberine is neuroprotective in humans or that it prevents or treats cognitive decline, and there were no planned or ongoing human trials in that area. The cognitive claims that circulate rest on mouse models.
We are citing a secondary source there rather than the underlying papers, because the primary articles were behind access checks we could not clear, and we would rather say so than pretend otherwise.
Berberine sits sixth. Its ceiling is 105.8 mg, about seven percent of the daily trial dose. It is also an ingredient with a broad interaction profile, which is covered on the side effects page.
L-theanine
One of the few ingredients here with a clean acute effect, at under half the dose that produces it.
- Printed position #7 of 11
- Ceiling 90.7 mg
- Studied at 200 mg in a single dose
- Studied for attention, most clearly in people who score high on trait anxiety
Theanine is the amino acid in tea that gets credited for the calm part of a cup. The attention work uses a single 200 mg dose, and the clearest result we read came from a study of visual attention and reaction time at exactly that amount.
The finding has an important condition attached. The improvement appeared in participants who scored high on trait anxiety, and not in the low-anxiety group. That is a real effect in a defined population rather than a general cognitive enhancement, and it is the sort of nuance that disappears from a marketing bullet.
Theanine sits seventh, ceiling 90.7 mg. Less than half of the studied dose, in a product taken daily rather than acutely before a task, which is not how the research was designed either.
Alpha lipoic acid
Spelled “Alpa” on the panel, and supported by no randomised trial in dementia at all.
- Printed position #8 of 11
- Ceiling 79.4 mg
- Studied at 600 mg a day
- Studied for diabetic nerve pain; for cognition, Cochrane found no trials at all
The label prints Alpa Lipoic Acid (ALA). It is a typographical error rather than a different compound, and we have kept it in the transcription because the panel is the record. It is a small thing, and small things on a printed panel are worth noticing.
The Cochrane review of alpha lipoic acid for dementia is unusual: it found no randomised, double-blind, placebo-controlled trials to include. Its conclusion is that in the absence of such trials no evidence exists to explore any potential effect, and that ALA cannot be recommended for people with dementia.
The one cognition-specific study we could open was a 2023 open-label pilot in fifteen older adults without cognitive impairment, at 600 mg a day for twelve weeks. It found no positive effect on cognition or mood. Fifteen people, unblinded, uncontrolled — weak evidence by design, and it points the same way as the Cochrane review.
ALA has a stronger record in diabetic peripheral neuropathy, which is a different condition and not what this product is sold for. Position eight gives it a 79.4 mg ceiling against a 600 mg literature dose.
GPC choline
The most encouraging single trial on this label, at roughly a ninth of its dose.
- Printed position #9 of 11
- Ceiling 70.6 mg
- Studied at 600 mg a day
- Studied for mild cognitive impairment, in one recent controlled trial
Alpha-GPC, or choline alfoscerate, is the ingredient here with the best recent result. A 2024 randomised, double-blind, placebo-controlled trial gave 600 mg a day for twelve weeks to a hundred people with amnestic mild cognitive impairment. The ADAS-cog score improved by 2.34 points more than placebo, which was statistically significant, and it was well tolerated.
Two caveats belong next to that. It is one trial, and we could not find a meta-analysis or replication confirming it. And it was run in people with diagnosed mild cognitive impairment, which is not the same population as a healthy adult buying a supplement for focus.
GPC choline sits ninth. Its ceiling is 70.6 mg, a little under a ninth of the trial dose. Of all the gaps on this label, this is the one we find most frustrating, because the ingredient had something to show.
Lutein and zeaxanthin
The two that fit comfortably, and the two whose large cognition trial found nothing.
- Printed position #10 of 11
- Ceiling 63.5 mg
- Studied at 10 mg a day in AREDS2
- Studied for eye health; the cognition arm of AREDS2 found nothing
Lutein and zeaxanthin are carotenoids from marigold flowers, and they are dosed in single-figure milligrams rather than hundreds. AREDS2, the large eye-health trial, used 10 mg of lutein and 2 mg of zeaxanthin a day.
Positions ten and eleven give ceilings of 63.5 mg and 57.7 mg. Both are far above the AREDS2 amounts, so these are the two ingredients on this label that can straightforwardly be present at a researched dose. That is worth saying plainly.
What the research found is the harder half. AREDS2 included a cognitive-function substudy across thousands of participants followed for about five years, and the cognition scores of every subgroup declined at a similar rate. No combination of the supplements tested made a difference. The National Eye Institute's own summary says exactly that.
So the two ingredients that can reach their dose are the two whose large trial was null for the thing this product is sold for. Both remain reasonable ingredients for eye health, which is not the claim on the front of the bottle.
What this adds up to
Read end to end, the eleven fall into four groups, and it is worth naming them.
Right ingredient, wrong amount
Bacopa and GPC choline both have real trial support and both sit at positions where the ceiling is well under the trial dose. These are the near misses.
Bacopa, GPC choline
Right amount, null result
Lutein and zeaxanthin can be present at their AREDS2 doses. The AREDS2 cognition substudy found no difference.
Lutein, zeaxanthin
Wrong category
Cinnamon and berberine are metabolic ingredients. Berberine has no human cognition evidence at all, and cinnamon's trial measured glucose and lipids.
Cinnamon, berberine
Open questions
Green tea has no verified cognition dose. Bilberry's species is unresolved on the panel itself. Ginkgo could be at its dose, and its largest trial was null.
Green tea, bilberry, ginkgo
That is a formula with real ingredients in it, assembled in a way that makes it very hard to expect any of them to do what their studies did.
Order CogniMary from the official website
Sixty capsules, thirty days, and a panel you can read before you buy. Start with one bottle so you can see the price and the guarantee term before you commit to six.
No price was published to this website. The figure is on the checkout page.
References
Every source below was opened and read. Nothing on this site cites a study we did not read, and no identifier here was written from memory.
- U.S. Food and Drug AdministrationDaily Value on the Nutrition and Supplement Facts Labels — FDA.gov · source
- U.S. Food and Drug AdministrationConverting Units of Measure for Folate, Niacin, and Vitamin E on the Nutrition and Supplement Facts Labels: Guidance for Industry — FDA.gov · source
- U.S. Government Publishing Office21 CFR 101.36 — Nutrition labeling of dietary supplements — eCFR.gov · source
- NIH Office of Dietary SupplementsFolate — Health Professional Fact Sheet — ods.od.nih.gov · source
- NIH Office of Dietary SupplementsVitamin B12 — Health Professional Fact Sheet — ods.od.nih.gov · source
- NIH Office of Dietary SupplementsMagnesium — Health Professional Fact Sheet — ods.od.nih.gov · source
- NIH Office of Dietary SupplementsChromium — Health Professional Fact Sheet — ods.od.nih.gov · source
- Lindberg, Zobitz, Poindexter & Pak — 1990Magnesium bioavailability from magnesium citrate and magnesium oxide — Journal of the American College of Nutrition · PMID 2407766 · source
- Walker, Marakis, Christie & Byng — 2003Mg citrate found more bioavailable than other Mg preparations in a randomised, double-blind study — Magnesium Research · PMID 14596323 · source
- Kappeler, Heimbeck, Herpich et al. — 2017Higher bioavailability of magnesium citrate as compared to magnesium oxide — BMC Nutrition · DOI 10.1186/s40795-016-0121-3 · source
- Malouf, Grimley Evans & Areosa Sastre — 2003Folic acid with or without vitamin B12 for cognition and dementia — Cochrane Database of Systematic Reviews · CD004514 · source
- Markun, Gravestock, Jäger et al. — 2021Effects of Vitamin B12 Supplementation on Cognitive Function, Depressive Symptoms, and Fatigue: A Systematic Review, Meta-Analysis, and Meta-Regression — Nutrients · DOI 10.3390/nu13030923 · source
- Chen et al. — 2024Magnesium and Cognitive Health in Adults: A Systematic Review and Meta-Analysis — Advances in Nutrition · PMC11362647 · source
- Snitz, O'Meara, Carlson et al. — 2009Ginkgo biloba for Preventing Cognitive Decline in Older Adults: A Randomized Trial — JAMA · PMID 20040554 · source
- Birks & Grimley Evans — 2009Ginkgo biloba for cognitive impairment and dementia — Cochrane Database of Systematic Reviews · CD003120.pub3 · source
- Kongkeaw, Dilokthornsakul, Thanarangsarit et al. — 2014Meta-analysis of randomized controlled trials on cognitive effects of Bacopa monnieri extract — Journal of Ethnopharmacology · PMID 24252493 · source
- Higashiyama, Htay, Ozeki et al. — 2011Effects of l-theanine on attention and reaction time response — Journal of Functional Foods · DOI 10.1016/j.jff.2011.03.009 · source
- Klugman, Sauer, Tabet & Howard — 2004Alpha lipoic acid for dementia — Cochrane Database of Systematic Reviews · CD004244.pub2 · source
- Fiorentini, Cantatore, Ferrari et al. — 2023Cognitive and Mood Effect of Alpha-Lipoic Acid Supplementation in a Nonclinical Elder Sample: An Open-Label Pilot Study — International Journal of Environmental Research and Public Health · PMID 36767724 · source
- Lee, Kim, Park et al. — 2024Efficacy and safety of choline alphoscerate for amnestic mild cognitive impairment: a randomized double-blind placebo-controlled trial — BMC Geriatrics · PMID 39300341 · source
- European Food Safety Authority — 2018Scientific opinion on the safety of green tea catechins — EFSA Journal · DOI 10.2903/j.efsa.2018.5239 · source
- Alzheimer's Drug Discovery FoundationBerberine — Cognitive Vitality For Researchers — alzdiscovery.org · source
- Ranasinghe et al. — 2025Effects of Cinnamomum zeylanicum extract on lipid profile, glucose levels and its safety in adults: a randomized, double-blind, controlled trial — PLOS ONE · DOI 10.1371/journal.pone.0317904 · source
- Chu, Cheung, Lau & BenzieBilberry (Vaccinium myrtillus L.), in Herbal Medicine: Biomolecular and Clinical Aspects — NCBI Bookshelf · NBK92770 · source
- Chew, Clemons, Agrón et al. — 2015Effect of Omega-3 Fatty Acids, Lutein/Zeaxanthin, or Other Nutrient Supplementation on Cognitive Function: The AREDS2 Randomized Clinical Trial — JAMA · PMID 26305649 · source
- U.S. Food and Drug AdministrationFDA Calls on Certain Firms to Stop Producing and Issuing Misleading ‘FDA Registration Certificates’ — FDA.gov · source
- U.S. Food and Drug AdministrationQuestions and Answers on Dietary Supplements — FDA.gov · source