In one pass Cocoa beans contain molecules called flavanols that help the lining of blood vessels release a little more nitric oxide and let the vessels relax slightly.
Educational content, not medical advice — consult a clinician.
Cocoa beans contain molecules called flavanols that help the lining of blood vessels release a little more nitric oxide and let the vessels relax slightly. That part is real, and the chapter on how flavanols relax blood vessels explains it. But the effect is small, and the trials used purified capsules. The bar in your hand packages flavanols together with a lot of sugar, fat, and calories. Pinning the small benefit of the first onto the whole candy bar is the core trick behind "healthy chocolate."
This story separates the genuinely useful component in cocoa from the candy bar it comes packed in: how strong the evidence is, how large the effect is, and why most chocolate on the shelf has already processed that benefit away.
Myth · The claims bundled into healthy chocolate
In recent years, headlines have loved to say that dark chocolate protects the heart, lowers blood pressure, lifts your mood, is an antioxidant, and even protects the brain. So chocolate went from a sweet to a "functional food" you can eat with a clear conscience, especially bars labeled high-cacao, dark, or raw cacao, which sell for more.
This story bundles three claims together, and once you pull them apart the conclusions differ:
Good for the heart, lowers blood pressure: purified, high-dose flavanols lower blood pressure a little, but in the largest trial, total cardiovascular events did not fall significantly.Rich in antioxidants, fights aging: antioxidant activity in a test tube does not mean eating it slows aging, and there is no reliable evidence for that in people.A healthy sweet you can eat freely: the benefit is counted in milligrams, the sugar and fat in grams, and most of the flavanols are lost in processing anyway.
Chapter 2
What flavanols are · mostly lost in processing
To talk about "healthy chocolate," you first need to know which molecule the benefit comes from, and how fragile it is.
The part of cocoa that helps blood vessels is a group called flavanols (flavan-3-ols), and the most studied one is epicatechin. All the evidence that "cocoa protects blood vessels" is about these molecules, not cocoa butter and not sugar.
The trouble is that they are lost in large amounts on the long road from bean to bar. Fermentation, roasting, and especially alkalization (also called Dutch processing), which makes cocoa powder darker and smoother, all break them down. A study that measured cocoa powders on the market found that natural cocoa powder had the most flavanols, that the heavier the alkalization the bigger the loss, and that heavy alkalization removed most of them.
That leads to a counterintuitive reality: the darkest, smoothest kind is often exactly the kind that has lost the most, and neither "dark" nor "high cacao" on the label guarantees the flavanols are still there.
Mechanism · Why smoother means fewer flavanols
Bitter, astringent, sour: the very tastes that processing works hardest to strip out are the signal that flavanols are still there. Once you understand why, you can roughly judge by mouth.
First, what alkalization does chemically. Alkalizing means adding alkali to cocoa to raise the pH. Flavanol molecules carry several phenolic hydroxyl groups (the —OH groups on a benzene ring), and in alkaline conditions those groups lose their hydrogen easily. Without it, the molecule becomes much more reactive: it oxidizes as soon as it meets oxygen, then links up with two or three others into large brown pigments. So getting darker and losing flavanols are not two separate things; they are two ends of the same reaction. Part of the deep brown you see used to be flavanols.
Next, why astringency works as an indicator. Astringency is not a taste but a sense of touch. In the mouth, flavanols grab proline-rich proteins in saliva. Saliva normally uses these proteins to lay a slippery film over the inside of the mouth; once grabbed, the proteins clump and settle out as flakes, and the film breaks up. Tongue and palate lose their lubrication and rub against each other, and that tight, puckering feeling is the result. So the astringency in your mouth is close to a direct reading of whether these molecules are still there.
That lets you run the shelf judgment in reverse:
Jet-black cocoa products that melt on the tongue with no astringency at all have most likely been heavily alkalized, with little flavanol leftNatural cocoa powder that is reddish-brown, smells sour, and tastes astringent and bitter is the kind that kept the most"Dark," "high cacao," and "raw cacao" on the package describe the recipe or the process; not one of those words promises the active compound That is also the lowest-effort practical takeaway in this story: instead of comparing labels, compare how astringent the first bite is.
Chapter 3
How much the big trial found
Cocoa flavanols are among the better-studied "superfood" ingredients, but reading the evidence means watching two things: the direction of the effect and its size.
First, be clear about what the trials handed out. Participants were never given chocolate. They were given flavanols purified from cocoa and packed into capsules: no sugar, no cocoa butter, and a daily amount far above what you could get from a snack bar. Every conclusion below is about these capsules.
Cardiovascular events: the largest randomized trial so far, COSMOS, compared these capsules with a placebo in more than twenty thousand older adults. The primary endpoint, total cardiovascular events, did not fall significantly. Cardiovascular death, a secondary measure, did come out lower, but a signal on a secondary measure is only a lead; it cannot be read as proof that cocoa protects the heart.
Blood pressure: real, but small. The pooled randomized trials were mostly small and short, and flavanol-rich cocoa lowered blood pressure only slightly, by a few millimeters of mercury at most. That is far less than weight loss, cutting salt, exercise, or blood-pressure medication.
Evidence · What the vessel-dilation test measures
Besides cardiovascular events and blood pressure, there is a third kind of result: endothelial function, meaning how well the lining of the blood vessels works. In healthy people, higher doses of cocoa flavanols improved the vessels' ability to widen; again, these were purified high-dose flavanols, not just any bar of chocolate. This layer deserves its own look, because it is the easiest to read as a stronger conclusion than it really is.
How this measure is taken. A cuff is wrapped around the arm and inflated to stop blood flow for a few minutes, then released all at once. The rush of returning blood scrapes hard across the vessel lining, and a healthy lining responds by releasing signaling molecules that make the artery relax and widen on the spot. Ultrasound measures how much it widened: that is flow-mediated dilation (FMD). So it does not measure whether the vessel is diseased; it measures whether the lining's relaxation pathway still responds.
What kind of evidence this is. It is a process marker. A more responsive pathway is a good thing, and it does show that flavanols acted where the mechanism predicted. But between "a more responsive pathway" and "one fewer heart attack, a few more years of life" lie many steps, and only trials with can test them. The largest trial did exactly that, and on the hard endpoint it found no overall benefit.
So the evidence in this story sums up like this: lowering blood pressure is a real but very small effect; lowering cardiovascular death is only a signal on a secondary measure; better endothelial function is a process marker; and all three used high-dose purified flavanols, none of which amounts to eating more chocolate.
Keep one general reading habit too: the more a study leans on a process marker, the more you should ask whether the later result you actually care about was measured, and what it showed. Marketing loves to quote process markers, because they more easily produce impressive numbers.
Chapter 4
How flavanols relax blood vessels
The small real benefit of cocoa flavanols lands on the inner lining of your blood vessels.
The innermost layer of a blood vessel is made of endothelial cells, which produce a signaling molecule called nitric oxide (). NO relaxes the smooth muscle in the vessel wall, so the vessel widens, blood pressure drops a little, and blood flows more easily. Epicatechin is thought to raise the amount of NO available in the lining, which produces that measurable but small widening of the vessels and drop in blood pressure.
Once the mechanism is clear, you can see where the marketing oversteps:
A real mechanism does not mean a big effect: the drop in blood pressure from a bar of chocolate is a matter of millimeters of mercury, nowhere near enough to replace blood-pressure medication"Antioxidant = anti-aging" is an overreach: antioxidant activity in a test tube does not mean eating it slows aging, and reliable evidence in people is missingThe trials used purified, high-dose flavanols with no sugar or fat; the chocolate you eat is the opposite So this is a real but gentle pathway in the blood vessels. It can make them a little more comfortable; it is a nice extra, not a rescue.
A CLOSER LOOK
Cocoa acts on signals in the vessel lining
The pathway is real but modest; purified trial flavanols are not the same as a whole chocolate bar.
1NO is a signaling molecule
Epicatechin is thought to increase available NO in the vessel lining; NO relaxes smooth muscle and the vessel widens.
2A mechanism with limits on scale
The measurable effect is small and cannot replace blood-pressure medication. Trials of purified high-dose flavanols do not directly represent chocolate with its sugar and fat.
Illustration for understanding; not to scale. Saved figures include explanations and sources.
Mechanism · How epicatechin raises available NO
"Epicatechin raises the available in the vessel lining" is the claim, but how it does so is the real link in this chain. NO in the lining has a production side and a loss side, and epicatechin is thought to nudge both.
The production side. The enzyme that makes NO in endothelial cells is endothelial nitric oxide synthase (eNOS). It usually sits half idle and only runs at full speed once it is phosphorylated (a phosphate group is hung on the enzyme, like flipping a switch from standby to on). In cell experiments, epicatechin pushes this step, so the same batch of enzyme releases more NO per unit of time.
The loss side. After NO is made, it has to travel some distance to reach the smooth muscle. If along the way it meets a superoxide anion (a reactive oxygen species carrying an extra electron that grabs at whatever it meets), the two react almost instantly and the NO is gone before it arrives. People with inflamed vessels, chronically high blood sugar, or a smoking habit have more superoxide, and that is how NO gets eaten halfway. The older explanation was that flavanols in the blood directly mop up superoxide on NO's behalf. But after you eat them, flavanol levels in the blood are very low, far too low to take on superoxide head to head, so that explanation no longer holds. More likely, they turn down the enzyme in the vessel wall that makes superoxide ( oxidase): less is made at the source, so less NO is lost on the way.
Keep track of where the evidence sits. The details of both the production and the loss side come mainly from cell and animal experiments. What human trials measure is the outcome end: better widening of the vessels and slightly lower blood pressure.
Put together, here is what shows on the outside. More NO made and less lost means more usable NO around the lining; the smooth muscle relaxes and the vessel widens a little. The same heartbeat pushing the same amount of blood through a slightly wider tube means slightly lower pressure. That is where the few-millimeters-of-mercury effect comes from.
This also marks a line that often gets crossed. What it adjusts is vascular tone, a knob that can be loosened or tightened at any moment. It does not touch plaque that has already grown into the vessel wall. Turn the knob a little and the blood-pressure reading moves a little; the plaque stays where it is. So improving vessel function and reversing vessel disease are two completely different sentences.
One more claim gets cleared up along the way. Even where "antioxidant" comes into this, it refers to something specific and located: the vessel wall making a little less superoxide. It is not the vague "scavenges free radicals, so it slows aging." The first can say where it happens, what the result is, and how it is measured; the second inflates a local mechanism into a whole-body promise.
Chapter 5
Sugar, fat, calories and portion
When you eat chocolate as a "health food," the biggest problem is not the flavanols but everything they come packed with.
A typical bar of milk chocolate is roughly half sugar, with most of the rest fat, and hardly any flavanols left; it usually carries around 500 to 600 kcal per 100 g. Dark chocolate is more bitter and has somewhat less sugar, but it is no lighter in calories.
The comparison that matters most is trial dose versus snack dose. The trials handed out purified flavanol capsules, 500 mg a day in COSMOS, with no sugar and no cocoa butter. Try to get the same amount from a Dark Chocolate & Cocoa bar off the shelf and the sugar and fat come along for the ride: the benefit is counted in milligrams, the cost in grams.
The honest place for it: dark chocolate is a sweet that is slightly better than many sweets, and fine as a treat. But it belongs in a different category from food worth eating for your health.
Myth · How the health halo flips the math
Two more comparisons deserve to be laid out on their own.
First: high cacao content does not mean high flavanols. The percentage on the package is the share of cocoa solids in the bar, meaning cocoa powder plus cocoa butter. It is a recipe number, not an active-ingredient content. It says nothing about how heavily the beans were fermented and roasted, or whether the cocoa powder was alkalized. So two bars with the same cacao percentage can differ a lot in actual flavanols, and nothing on the label will tell you. The useful clue is in your mouth: the bar that tastes astringent at first and bitter at the finish probably kept more.
Second: the health halo can flip the math. Once something is labeled "healthy," people's sense of portion may loosen: less hesitation before reaching for a second square, and less chance of remembering they already had some today. So the "healthy" label cuts guilt on one side and may enlarge the amount actually eaten on the other, and the net sugar and calories can even end up higher than with no label. Research on this is mixed, but the direction is worth guarding against: what the label rewrites is your sense of whether this counts as a lot.
Together, the two explain a pattern that seems contradictory: the more a Dark Chocolate & Cocoa bar's marketing stresses its benefits, the smaller the reader's net gain may be, possibly even negative. The benefit is counted in milligrams and may not even be in the bar you bought; the cost is grams of sugar and fat, and it may grow along with the thought "this is healthy."
So the most practical defense against the health halo is not memorizing which bar is better. It is reclassifying it: count its sugar and calories as a sweet first, then weigh the small extra benefit. Do it in the other order and the math is already wrong.
Chapter 6
What to do · a treat, not a supplement
The judgments to take away:
Claims that do not hold: chocolate is a health food, you can eat as much as you like, it slows aging. The benefit of cocoa flavanols is real but small, and most of it is lost in processing.
Partly true but inflated: purified, high-dose cocoa flavanols lower blood pressure slightly and improve the function of the vessel lining, but that comes from capsules, not from the candy bar.
Low-risk ways to handle it:
Eat it if you want it, as a treat: a small square of 70% to 85% dark chocolate, eaten slowly, is satisfying and relatively low in sugarIf flavanols are really what you are after, unsweetened natural cocoa powder is the more solid sourceDo not let "it's healthy" talk you into eating more; for steadier blood pressure, rely on less salt, weight loss, exercise, and your doctor's plan
In practice · How to pick cocoa powder and drink it
If you decide to skip the sugar and go straight for the flavanols, the cocoa-powder route is worth spelling out.
When picking, look for one word. Prefer cocoa powder labeled non-alkalized or natural. On the other hand, a package that says alkalized (also called Dutch processed), or only says dark cocoa powder, is most likely the alkalized kind. Alkalizing processes the flavanols away, and the blacker the color and the smoother the texture, the more has usually been lost. So the rule here is the reverse of picking chocolate: the powder that is reddish-brown, smells sour, and tastes astringent when you mix it is the kind that kept more.
When drinking it, do not add the sugar back. Cocoa powder itself has almost no sugar, and a cup is very low in calories. But once hot cocoa follows the traditional recipe, with sugar and milk stirred in, it turns back into a sweet drink. To keep the point of this route, mix it with water or unsweetened milk and accept the bitterness, or just sprinkle the powder into oats, yogurt, or coffee and let other foods provide the texture.
Do not treat it as medicine. After the switch to cocoa powder, the benefit is still that small, few-millimeters-of-mercury effect. It is worth doing because the cost is close to zero (no sugar, no fat, and no extra money), not because the benefit got bigger. Whether something is worth doing depends on benefit minus cost, not on the benefit alone, and this whole story has been teaching that arithmetic.
Two more things come along with cocoa. One is cadmium and lead: cocoa powder and chocolate sometimes contain cadmium and lead from the soil and the wider environment, and products with more cocoa tend to carry more. That is not a reason to give it up, but it is one more reason to keep it at an occasional small square rather than a daily health supplement eaten by the handful, and to give children even less. The other is caffeine and theobromine (a stimulant molecule that acts more gently than caffeine): the darker the product, the more it contains. If you are sensitive to caffeine or sleep poorly, treat it like coffee and do not eat it close to bedtime.
abt-2018-cocoa-cadmium-lead
References · 4
Sesso, H. D., Manson, J. E., Aragaki, A. K., Rist, P. M., Johnson, L. G., Friedenberg, G., et al. (2022). Effect of cocoa flavanol supplementation for the prevention of cardiovascular disease events: the COcoa Supplement and Multivitamin Outcomes Study (COSMOS) randomized clinical trial. The American Journal of Clinical Nutrition, 115(6), 1490-1500. In 21,442 older adults, a 500 mg/day cocoa-flavanol supplement did not significantly reduce total cardiovascular events (primary endpoint) but reduced cardiovascular death (secondary) by 27% (HR 0.73). 10.1093/ajcn/nqac055
Miller, K. B., Hurst, W. J., Payne, M. J., Stuart, D. A., Apgar, J., Sweigart, D. S., & Ou, B. (2008). Impact of alkalization on the antioxidant and flavanol content of commercial cocoa powders. Journal of Agricultural and Food Chemistry, 56(18), 8527-8533. Natural (non-alkalized) cocoa powder had the highest flavanols and antioxidant capacity; alkalization (Dutch processing) sharply reduced flavanols, with heavy alkalization losing the majority. 10.1021/jf801670p
Ried, K., Fakler, P., & Stocks, N. P. (2017). Effect of cocoa on blood pressure. Cochrane Database of Systematic Reviews, 4, CD008893. Meta-analysis of ~40 RCT comparisons (~1,804 participants) found flavanol-rich cocoa lowered blood pressure by a small amount (systolic ~1.8 mmHg), attributed to flavanol-driven endothelial nitric oxide. Abstract: 35 trials (40 comparisons, 1,804 mainly healthy participants) lasting 2 to 18 weeks (mean 9); SBP -1.76 mmHg (-3.09 to -0.43) and DBP -1.76 mmHg; about -4 mmHg systolic in people with hypertension and no significant difference in normotensive people; moderate-quality evidence for a 'small (2 mmHg)' effect 'in the short term'. The abstract does not test the nitric-oxide mechanism (abstract, PMID 28439881). 10.1002/14651858.CD008893.pub3
Sansone, R., Rodriguez-Mateos, A., Heuel, J., Falk, D., Schuler, D., Wagstaff, R., et al. (2015). Cocoa flavanol intake improves endothelial function and Framingham Risk Score in healthy men and women: a randomised, controlled, double-masked trial (the Flaviola Health Study). British Journal of Nutrition, 114(8), 1246-1255. Twice-daily cocoa flavanols (~450 mg each) improved flow-mediated dilation and lowered estimated cardiovascular risk in healthy adults, using purified high-dose flavanols rather than chocolate. 10.1017/S0007114515002822