Cocoa flavanols
Also called cocoa extract, flavan-3-ols. Flavanols (a subclass of flavonoids).
Bottom line
Cocoa flavanols are a subclass of flavonoids found in cocoa [17]. In short-term trials in mainly healthy adults they slightly lowered blood pressure compared with control, though not significantly in people with normal blood pressure [11]. The large Mars Edge-supported COSMOS trials in older adults found no overall memory or cognition benefit [3][5]. Trials reported nausea and gastrointestinal complaints [11].
- In the Mars Edge-supported COSMOS-Mind trial of 2,262 older adults, 500 mg a day of cocoa flavanols for 3 years had no effect on global cognition compared with placebo [3].
- In the Mars Edge-supported COSMOS-Web trial of 3,562 older adults, cocoa flavanols did not improve memory overall; a benefit appeared only in people with lower diet quality or lower habitual flavanol intake [5].
- A Cochrane review of 35 trials found flavanol-rich cocoa lowered blood pressure by about 2 mmHg in mainly healthy adults in the short term, with moderate-quality evidence [11].
- Single drinks of 415-623 mg cocoa flavanols did not improve working memory or cognitive control in placebo-controlled crossover trials in young adults [7][8][10].
- In the Cochrane review, gastrointestinal complaints and nausea were reported by 1% of people taking cocoa and 0.4% of controls [11].
New to Cocoa flavanols? Start with What are cocoa flavanols? A plain-English guide.
Evidence by outcome
| Outcome | Grade | Basis | Modifiers | Rationale |
|---|---|---|---|---|
| Blood pressure | B Moderate | Meta-analysis with high (unexplained) heterogeneity: Cochrane review of 35 RCTs, moderate-quality evidence | Short-term only | Why B grade for Blood pressure: explainedApproved of the Blood pressure grade |
| Working memory | C Limited, no effect found | C, direction none: small acute placebo-controlled crossover trials from one research group were null; N needs a meta-analysis or at least 2 independent, adequately sized RCTs | Short-term onlySingle research group | Why C grade for Working memory: explainedApproved of the Working memory grade |
| Executive function | C Limited | Conservative reading of conflicting evidence (citation-rules 9): pooled meta-analysis results positive, largest long-term trials null | Why C grade for Executive function: explainedApproved of the Executive function grade | |
| Global cognition | C Limited, no effect found | C, direction none: two COSMOS substudies in older adults (COSMOS-Mind n=2,262, 3 years; COSMOS-Clinic n=573, 2 years) were null on global cognition, their primary outcome, but they share one parent trial and are not independent | Industry-funded | Why C grade for Global cognition: explainedApproved of the Global cognition grade |
| Episodic memory | C Limited, no effect found | C, direction none: only COSMOS-Web was null on a primary memory endpoint; COSMOS-Mind and COSMOS-Clinic tested memory as a secondary outcome, and all three are substudies of one parent trial, so N is not available | Industry-funded | Why C grade for Episodic memory: explainedApproved of the Episodic memory grade |
| Vascular function | C Limited | C cap: both meta-analyses pool healthy and clinical populations with no healthy-only estimate; high heterogeneity and risk-of-bias concerns | Industry-fundedShort-term only | Why C grade for Vascular function: explainedApproved of the Vascular function grade |
Grades describe evidence for each outcome, not a recommendation. Open a grade for its rationale and dated grade history. How we grade
The COSMOS trials were supported by an investigator-initiated grant from Mars Edge, which included infrastructure support and the donation of study pills [5], and the main COSMOS report states that neither Mars Edge nor Pfizer (which supplied the multivitamin) had a role in design, analysis or manuscript preparation, apart from independently completed blinded assays supported by Mars Edge [15]. Our evidence grading method explains each grade, and the evidence matrix compares compounds.
Do cocoa flavanols improve memory?
Not in older adults overall, according to the largest trials. In COSMOS-Web, 3,562 older adults were randomly assigned to cocoa extract (500 mg of cocoa flavanols per day) or placebo for 3 years, and the prespecified primary endpoint, an improvement in memory in all participants after 1 year, was not statistically significant [5]. Two of that Mars Edge-supported study's authors are employed by Mars, Incorporated, and two others received an unrestricted research grant from Mars [5].
The same trial reported that the flavanol intervention restored memory among participants in the lower thirds of habitual diet quality or habitual flavanol intake, a subgroup finding [5]. A second COSMOS analysis, COSMOS-Mind, found no effect on an episodic memory composite after 3 years (mean change z-score 0.03, 95% CI -0.04 to 0.09) [3]. In COSMOS-Clinic, 573 older adults had detailed in-person testing, and 2 years of cocoa extract had no significant effect on episodic memory compared with placebo [4]. Both are part of the Mars Edge-supported COSMOS study [3][4].
Earlier, smaller work pointed the other way. In a 12-week randomized trial of healthy, sedentary adults aged 50-69 (37 completers), a 900 mg daily dose of cocoa flavanols improved performance on a memory task linked to one part of the hippocampus, compared with a low-flavanol diet, but had no effect on delayed retention; this trial was part-funded by an unrestricted grant from Mars, Inc. [6]. A 2026 meta-analysis found the results of trials on working and verbal memory were inconsistent [1].
Do cocoa flavanols improve overall cognition?
Not in the long-term trials of older adults. COSMOS-Mind, a 3-year randomized trial that tested cognition by telephone in 2,262 participants (mean age 73), found cocoa extract containing 500 mg a day of flavanols had no effect on global cognition compared with placebo (mean z-score 0.03, 95% CI -0.02 to 0.08); it was part of the Mars Edge-supported COSMOS study [3]. COSMOS-Clinic, also Mars Edge-supported, found no significant effect on 2-year change in global cognition (mean difference -0.01 standard deviation units) [4]. A 2026 meta-analysis of 16 randomized trials found no benefit on the Mini-Mental State Examination, a short screening test of overall cognition [1].
Do cocoa flavanols improve executive function and processing speed?
Evidence is mixed. A 2026 GRADE-assessed meta-analysis of 16 randomized trials with 2,847 participants found that regular dark chocolate or cocoa intake improved results on the Trail Making Test part A (weighted mean difference -4.51; 95% CI -8.41 to -0.60) and part B (-13.1; 95% CI -17.73 to -8.47) and on a verbal fluency test (3.15; 95% CI 0.71 to 5.6) [1]. Only eight of the 16 studies found a significant effect on at least one cognitive domain, and subgroup analysis suggested larger effects at 500 mg or more of cocoa flavanols a day and 8 weeks or longer [1]. The authors rated the quality of evidence as moderate and high; the review was funded by Iran University of Medical Sciences [1].
A 2023 meta-analysis of seven trials in healthy adults found chocolate reduced executive-function test time (weighted mean difference -11.77; 95% CI -22.49 to -1.05), but the authors could not run subgroup analyses because of too few trials and significant heterogeneity [2]. Against both, the Mars Edge-supported COSMOS-Mind trial found no effect on an executive function composite over 3 years [3], and COSMOS-Clinic, also Mars Edge-supported, found no significant 2-year effect on executive function and attention [4].
Does a single dose sharpen attention or working memory?
In recent placebo-controlled crossover trials in young adults, no. In two experiments with 48 and 32 healthy young adults, 415 mg of cocoa flavanols did not improve visual working memory recall [7]. In a trial of 48 young adults comparing placebo, 415 mg of cocoa flavanols, 215 mg of caffeine and both together, neither substance nor the combination affected accuracy or reaction times, with cocoa powders donated by Barry Callebaut [8]. In 36 healthy adults, neither a 415 mg nor a 623 mg drink improved cognitive control or response inhibition [10].
One pre-registered crossover trial in 48 university students found cocoa flavanols made visual search faster but did not help temporal attention, with cocoa powders donated by Barry Callebaut [9]. All four trials share authors from one research group, so independent replication would strengthen this conclusion [7][8][9][10].
Do cocoa flavanols affect blood pressure?
Trials found a small, short-term effect. A Cochrane review of 35 randomized trials with 1,804 mainly healthy participants found flavanol-rich cocoa products lowered blood pressure compared with control in trials lasting 2 to 18 weeks (systolic mean difference -1.76 mmHg, 95% CI -3.09 to -0.43; diastolic -1.76 mmHg, 95% CI -2.57 to -0.94) [11]. The reviewers downgraded the evidence from high to moderate quality because of heterogeneity they could not explain [11].
The effect may depend on starting blood pressure. There was no significant difference in people with normal blood pressure (23 treatment comparisons, 1,063 participants), although the test for differences between subgroups was of borderline significance [11]. Unblinded trials tended to show larger reductions than double-blind trials, though this difference was not statistically significant, and excluding trials with authors employed by industry sponsors reduced the effect slightly, which the reviewers took as a sign of some reporting bias [11]. Two of the Cochrane authors were investigators on trials included in the review [11].
A 2026 meta-analysis of 25 trials (1,167 participants) found similar small reductions (systolic -2.734 mmHg, diastolic -1.848 mmHg); it was funded by Meiji Co., Ltd., a maker of cocoa-containing foods, five authors are Meiji employees, and two work for Yakujihou Marketing Jimusho Co., Ltd. [12]. In a 12-week randomized trial by Mars, Inc. researchers, up to 2,000 mg of cocoa flavanols a day did not change blood pressure in healthy adults with normal blood pressure compared with a flavanol-free control [14]. In a long-term, Mars Edge-supported COSMOS analysis of 8,905 older adults without high blood pressure at the start, cocoa extract for a median 3.4 years did not significantly change the rate of new high blood pressure (hazard ratio 0.96, 95% CI 0.88 to 1.05), although subgroup results differed by starting blood pressure [24].
Do cocoa flavanols improve blood vessel function?
Pooled trials suggest a modest improvement in flow-mediated dilation, a measure of how well an artery widens when blood flow increases, but the evidence has quality problems. A 2019 dose-response meta-analysis of 15 articles (18 intervention arms) by a team including authors at the Nestlé Research Centre found flow-mediated dilation improved by 1.17% (95% CI 0.76% to 1.57%) compared with control, with an inverted U-shaped dose response [13]. The authors reported substantial variation in results that they could not explain and significant risk-of-bias concerns in a large majority of studies [13].
The Meiji-funded 2026 meta-analysis also found improved flow-mediated dilation (mean difference 1.314%) and pulse wave velocity, a measure of arterial stiffness, but rated certainty for vascular outcomes as limited [12].
What the evidence does not show
The large, long-term trials do not show that cocoa flavanols improve memory or overall thinking in older adults in general [3][4][5]. Possible benefits appeared only in subgroups of the Mars Edge-supported COSMOS trials: people with lower diet quality or flavanol intake in COSMOS-Web [5], and people with poorer diet quality in COSMOS-Clinic, in an analysis not corrected for multiple testing [4]. One small 12-week trial part-funded by Mars, Inc. was also positive [6]. In Great Britain, claims that cocoa supports positive mood or a normal blood pressure were assessed and are not authorized [19].
Doses used in studies
Ranges researchers used, not dosing advice. Bars share one scale. Read our disclaimer.
Timing and form
A review by authors including Mars, Inc. researchers reports that in studies using labeled (-)-epicatechin, it reaches peak plasma levels of its metabolites about 1 hour after ingestion, and that urinary excretion indicates 95% is absorbed [16]. A second review reports that blood levels are influenced by food matrix, processing, amount, age, sex and genetic differences [17].
Forms studied include cocoa drinks in the acute trials [8][10], capsules of cocoa extract in COSMOS [15] and dark chocolate or cocoa products in the pooled trials [1][11]. FDA's qualified health claims apply to high flavanol cocoa powder and do not apply to regular cocoa powder or chocolate [18].
What it is
Cocoa flavanols are a subclass of flavonoids, and cocoa is rich in them [17]. Cocoa is made from the bean of the cacao tree, Theobroma cacao, and flavanols are also found in tea, grapes, wine and other foods, according to the COSMOS trial team [15]. Cocoa extract also contains the methylxanthines theobromine and caffeine [15].
Processing matters. The COSMOS authors note that cocoa powder is alkalized to improve solubility, alter color and modify flavor, and that food-based cocoa is highly susceptible to variation in flavanol, theobromine and other bioactive content [15].
How it may work (mechanism, not effect): flavanols found in cocoa have been shown to increase the formation of endothelial nitric oxide, which promotes vasodilation, the widening of blood vessels [11].
Safety and interactions
Safety flag. In the Cochrane review, gastrointestinal complaints and nausea were reported by 1% of participants taking cocoa and 0.4% in control groups [11]. In COSMOS, a Mars Edge-supported trial of 21,442 older US adults, the authors reported no safety concerns over a median 3.6 years; nausea was slightly more common with cocoa extract [15]. In a trial by Mars, Inc. researchers, up to 2,000 mg a day for 12 weeks was well tolerated by healthy adults [14]. As of November 12, 2026, the NIH Office of Dietary Supplements (as archived November 1, 2026) and LiverTox (as archived November 5, 2026) have no cocoa entry, and our sources give no safety information for pregnancy, breastfeeding or children. Small trials of chocolate in pregnant women were not reviewed for this page.
Who sources say should be careful.
- People who are very sensitive to caffeine. The COSMOS trial excluded them because its cocoa extract contained theobromine and caffeine [15].
- People taking antiplatelet medicines, such as clopidogrel. In one small uncontrolled pilot study, the effect of clopidogrel on a platelet function test increased after a week of dark chocolate [23].
As of November 12, 2026, the NIH Office of Dietary Supplements (as archived November 1, 2026) and LiverTox (as archived November 5, 2026) have no cocoa entry, and our sources give no safety information for pregnancy, breastfeeding or children. Small trials of chocolate in pregnant women were not reviewed for this page. These groups or those with a health condition should seek advice from a clinician.
In the trials reviewed by Cochrane, cocoa products were highly tolerable: gastrointestinal complaints and nausea were reported by 1% of participants in the cocoa groups and 0.4% in control groups, with moderate-quality evidence [11]. In the Mars Edge-supported COSMOS trial, 21,442 older US adults took 500 mg of cocoa flavanols a day or placebo for a median 3.6 years, and the authors reported no safety concerns [15]. In that trial, people taking cocoa extract were slightly more likely to report nausea (HR 1.06; 95% CI 1.02 to 1.11) [15].
At higher intakes, a trial by Mars, Inc. researchers gave healthy adults aged 35-55 up to 2,000 mg a day for 12 weeks; adverse events were mild and did not differ significantly from the flavanol-free control group [14]. The Cochrane reviewers noted that long-term trials are still needed to assess potential adverse effects of chronic intake [11].
As of November 11, 2026 we found no regulator upper intake for cocoa flavanols on the FDA, GB register, Food Standards Agency and Health Canada pages we checked. For one authorized novel food, an extract of defatted cocoa powder, the UK Food Standards Agency requires labels telling consumers not to consume more than 600 mg of polyphenols, corresponding to 1.1 g of the extract, per day [21].
Interactions sources report
As of November 11, 2026 we found no controlled drug-interaction trials of cocoa flavanol supplements (PubMed and Europe PMC search, November 11, 2026). For antiplatelet medicines, such as clopidogrel and aspirin, one uncontrolled 1-week pilot study gave 30 g a day of 65% cocoa dark chocolate to 20 people with stable coronary artery disease taking both [23]. The effect of clopidogrel on a platelet function test increased, the effect of aspirin did not change significantly, and no serious adverse events occurred; the study was funded by The University of the West Indies, and a cocoa estate donated the cocoa beans [23]. The extract used in COSMOS contained about 50 mg of theobromine and 15 mg of caffeine a day, and the trial excluded people who reported extreme sensitivity to caffeine [15].
If you take any of these medicines, talk to your doctor or pharmacist before combining them with cocoa flavanol supplements.
Regulatory status
| Market | Status | Notes |
|---|---|---|
| US | FDA does not intend to object to certain qualified health claims for cocoa flavanols in high flavanol cocoa powder (at least 4% naturally conserved cocoa flavanols) on conventional foods, and states the evidence is very limited; the claims do not apply to chocolate. Supplements: not verified as of November 11, 2026. The FDA source we checked covers conventional foods only, and we could not confirm the status of cocoa flavanol extracts as dietary supplements. | |
| UK | GB register authorizes the claim 'Cocoa flavanols help maintain the elasticity of blood vessels, which contributes to normal blood flow.' with a 200 mg daily intake condition; a normal blood pressure claim is not authorized. The register marks this claim as authorized on the basis of proprietary data. One defatted cocoa powder extract is an authorized novel food in England, Scotland and Wales. | |
| Canada | Not verified as of November 11, 2026. We checked the Health Canada NHPID single-ingredient monograph list and found no monograph for cocoa or cocoa flavanols; we could not confirm the licensing status of cocoa flavanol products. | |
| Australia | Theobroma cacao is a permissible active ingredient, excipient or homeopathic ingredient, with caffeine limits for oral medicines; cocoa powder is also listed (Therapeutic Goods (Permissible Ingredients) Determination (No. 3) 2026). ARTG product listings and FSANZ food status: not verified as of November 11, 2026. | |
| EU | Regulation (EU) 2015/539 extended the blood vessel elasticity claim to high-flavanol cocoa extract, authorized on April 21, 2015 restricted to Barry Callebaut Belgium NV for five years. Current status: not verified as of November 11, 2026. We checked the legislation.gov.uk copy of the regulation; EUR-Lex did not return the text. |
FAQ
Do cocoa flavanols improve memory?
Not overall, in the largest trials. In the Mars Edge-supported COSMOS-Web trial of 3,562 older adults, 500 mg a day did not improve memory in all participants at 1 year, although memory improved in people with lower diet quality or lower habitual flavanol intake [5]. COSMOS-Mind and COSMOS-Clinic, also Mars Edge-supported, found no effect on episodic memory [3][4].
Is dark chocolate the same as a cocoa flavanol supplement?
Not necessarily. The COSMOS authors note that processing steps such as alkalizing change cocoa's bioactive content, and their trial used a cocoa extract supplement rather than cocoa foods [15]. FDA's qualified health claims for high flavanol cocoa powder do not apply to chocolate [18].
Do cocoa flavanols lower blood pressure?
Trials found a small short-term effect. A Cochrane review of 35 trials reported about 2 mmHg lower blood pressure than control in mainly healthy adults, but the difference was not significant in people with normal blood pressure [11]. In Great Britain a normal blood pressure claim for cocoa flavanols is not authorized [19].
Do cocoa flavanol supplements contain caffeine?
The cocoa extract used in COSMOS contained about 50 mg of theobromine and 15 mg of caffeine a day [15]. Anyone sensitive to caffeine or taking medicines should ask a pharmacist or clinician.
References
- Arabpour Z, Arabpour Z, Jazayeri S. Effect of Chronic Dark Chocolate or Cocoa Intake on Cognitive Function: A Grading of Recommendations Assessment, Development, and Evaluation (GRADE)-Assessed Systematic Review and Meta-Analysis of Randomized Controlled Trials. Nutr Rev. 2026.
- Shateri Z, Kooshki A, Hormoznejad R, et al. Effects of chocolate on cognitive function in healthy adults: A systematic review and meta-analysis on clinical trials. Phytother Res. 2023.
- Baker LD, Manson JE, Rapp SR, et al. Effects of cocoa extract and a multivitamin on cognitive function: A randomized clinical trial. Alzheimers Dement. 2023.
- Vyas CM, Manson JE, Sesso HD, et al. Effect of cocoa extract supplementation on cognitive function: results from the clinic subcohort of the COSMOS trial. Am J Clin Nutr. 2024.
- Brickman AM, Yeung LK, Alschuler DM, et al. Dietary flavanols restore hippocampal-dependent memory in older adults with lower diet quality and lower habitual flavanol consumption. Proc Natl Acad Sci U S A. 2023.
- Brickman AM, Khan UA, Provenzano FA, et al. Enhancing dentate gyrus function with dietary flavanols improves cognition in older adults. Nat Neurosci. 2014.
- Altınok A, Karabay A, Akyürek EG. Acute effects of cocoa flavanols on visual working memory: maintenance and updating. Eur J Nutr. 2022.
- Akyürek EG, Altınok A, Karabay A. Concurrent consumption of cocoa flavanols and caffeine does not acutely modulate working memory and attention. Eur J Nutr. 2024.
- Karabay A, Saija JD, Field DT, et al. The acute effects of cocoa flavanols on temporal and spatial attention. Psychopharmacology (Berl). 2018.
- Altinok A, Moruzzi M, Akyürek EG. Acute effects of cocoa flavanols on cognitive control and response inhibition. Nutrition. 2025.
- Ried K, Fakler P, Stocks NP. Effect of cocoa on blood pressure. Cochrane Database Syst Rev. 2017.
- Sakuda M, Ito K, Adaniya Y, et al. Impact of cocoa flavanols on blood pressure and vascular function: an updated systematic review and meta-analysis of randomized controlled trials. J Nutr Health Aging. 2026.
- Sun Y, Zimmermann D, De Castro CA, et al. Dose-response relationship between cocoa flavanols and human endothelial function: a systematic review and meta-analysis of randomized trials. Food Funct. 2019.
- Ottaviani JI, Balz M, Kimball J, et al. Safety and efficacy of cocoa flavanol intake in healthy adults: a randomized, controlled, double-masked trial. Am J Clin Nutr. 2015.
- Sesso HD, Manson JE, Aragaki AK, et al. Effect of cocoa flavanol supplementation for the prevention of cardiovascular disease events: the COcoa Supplement and Multivitamin Outcomes Study (COSMOS) randomized clinical trial. Am J Clin Nutr. 2022.
- Borges G, Ottaviani JI, van der Hooft JJJ, et al. Absorption, metabolism, distribution and excretion of (-)-epicatechin: A review of recent findings. Mol Aspects Med. 2018.
- Cifuentes-Gomez T, Rodriguez-Mateos A, Gonzalez-Salvador I, et al. Factors Affecting the Absorption, Metabolism, and Excretion of Cocoa Flavanols in Humans. J Agric Food Chem. 2015.
- Seecheran NA, Sukha D, Grimaldos K, et al. Effect of cocoa (Theobroma cacao L.) on platelet function testing profiles in patients with coronary artery disease: ECLAIR pilot study. Open Heart. 2022.
- Hamaya R, Li S, Lau J, et al. Long-Term Effect of Cocoa Extract Supplementation on Incident Hypertension. Hypertension. 2025.
- U.S. Food and Drug Administration. FDA Announces Qualified Health Claim for Cocoa Flavanols in High Flavanol Cocoa Powder and Reduced Risk of Cardiovascular Disease. Updated February 3, 2023.
- Department of Health and Social Care (GOV.UK). Great Britain nutrition and health claims (NHC) register (spreadsheet, 19 May 2026; cocoa rows). Updated May 19, 2026.
- legislation.gov.uk. Commission Regulation (EU) 2015/539 (cocoa flavanols health claim).
- Food Standards Agency. Novel food authorisation NOVEL-54: Extract of defatted cocoa powder. Updated April 1, 2025.
- Health Canada (Natural and Non-prescription Health Products Directorate). Natural Health Products Ingredients Database: Single ingredient monographs. Updated March 1, 2019.
- Australian Government, Department of Health, Disability and Ageing (Federal Register of Legislation). Therapeutic Goods (Permissible Ingredients) Determination (No. 3) 2026 (F2026L01382), volumes 1, 2 and 5. Updated October 13, 2026.