Dark Chocolate Nutrition: Why Cacao Percentage Doesn't Tell the Whole Story

Pick up a Hershey’s Special Dark bar. It looks like serious dark chocolate. It’s dark brown, almost black, and the packaging implies health. It’s also likely to have fewer flavanols than a lighter-colored bar made from non-alkalized cocoa that costs less at the same grocery store.

That’s the dutching problem, and it’s why cacao percentage is mostly a marketing number.

The bioactives in dark chocolate that have generated real research interest are flavanols, specifically (-)-epicatechin and (+)-catechin. Fresh cacao beans are extraordinarily rich in them. But chocolate is not fresh cacao. Between the bean and the bar, flavanols get destroyed at almost every step. Fermentation, roasting, and particularly alkalization all reduce flavanol content. By the time you buy a bar, you have no reliable way to know what’s left unless the flavanol content is certified on the label.

Nutritional Profile

The numbers below are per 28g (1 oz) of 70-85% dark chocolate, based on USDA food composition data and Katz et al. (2011, Antioxidants and Redox Signaling, PMID: 21470061):

That flavanol range isn’t sloppy reporting. It reflects how dramatically processing conditions affect the final number. A well-made, non-alkalized 70% bar can have 600mg of flavanols per ounce. A Dutch-process bar in the same cacao percentage range might have under 50mg. Same label, completely different chemistry.

The magnesium and iron numbers are worth noting. Dark chocolate is a real source of both. At 16% DV for magnesium per ounce, it’s one of the better non-leafy-green sources most people will actually eat. See magnesium in the body for why that mineral’s relationship with absorption is more complicated than these numbers suggest.

The Dutching Problem: Why Cacao Percentage Misleads

Dutch-process cocoa uses an alkali solution (typically potassium carbonate) to neutralize the natural acidity of cacao. It produces a darker color, a smoother flavor, and better solubility in liquids. It also destroys 60-90% of the flavanols, depending on the conditions used (Katz et al., 2011).

The destruction is chemical. Flavanols are polyphenols with multiple hydroxyl groups that are sensitive to pH. Alkalization shifts the pH dramatically, breaking down the epicatechin and catechin structures. The darker color you see is actually a product of those same reactions, which means darkness is correlated with flavanol loss, not flavanol preservation.

Natural (non-alkalized) cocoa keeps the original acidity. It’s lighter in color, more bitter, and has a fruitier flavor profile. It also retains far more of the flavanol content from the original bean.

You can spot Dutch-process on a label. Ingredients that say “cocoa processed with alkali” or specify “Dutch-process cocoa” are alkalized. Natural cocoa will usually say “natural cocoa” or just “cocoa” without the alkali qualifier. The phrase to watch for is “processed with alkali.” If you see it, the flavanol content is likely low regardless of what the cacao percentage says.

Fermentation and roasting also matter, though consumers have no visibility into either. Cacao beans must be fermented after harvest to develop the flavor precursors that eventually become chocolate’s characteristic taste. Longer, hotter fermentation degrades more flavanols but produces better flavor. High-temperature roasting reduces flavanol content further. Most commercial chocolate doesn’t disclose roasting temperatures or fermentation protocols. This is why certified flavanol content is the only reliable measure.

What Flavanols Actually Do

Think of the flavanol research like a drug trial with a poorly standardized drug. Most of the observational studies on chocolate and health suffer from the same problem: they ask participants how much chocolate they eat, not how many flavanols they consumed. A person eating 1 oz of Dutch-process dark chocolate per day and a person eating 1 oz of non-alkalized dark chocolate per day are in the same data category but consuming vastly different amounts of the bioactive compound being studied.

The best evidence comes from trials that used standardized flavanol supplements, which removed the processing variable entirely.

The COSMOS-Cocoa trial (Sesso et al., 2022, New England Journal of Medicine, PMID: 36351090) is the landmark study. It randomized 21,442 adults to either 500mg of cocoa flavanols per day (as a supplement) or placebo, for an average of 3.6 years. In the full population, total cardiovascular events were not reduced. But in participants who had prior cardiovascular disease at enrollment, cocoa flavanol supplementation was associated with a 27% reduction in cardiovascular mortality.

That’s an important distinction. The COSMOS trial does not show that dark chocolate prevents heart disease in healthy people. It shows that concentrated, standardized flavanol supplementation was associated with reduced cardiovascular mortality specifically in people who already had cardiovascular disease. Conflating these findings with “eat more dark chocolate” misrepresents what the evidence shows.

On blood pressure, the evidence is cleaner and applies more broadly. A Cochrane review by Ried et al. (2012, Cochrane Database of Systematic Reviews, PMID: 22895979) pooled short-term randomized trials and found that flavanol-rich cocoa reduced systolic blood pressure by 2-3 mmHg and diastolic blood pressure by 1-2 mmHg compared to control. These are modest effects. They’re real effects in the short-term data, consistent with the vasodilatory mechanism of epicatechin (which increases nitric oxide production in endothelial cells), but they’re not a substitute for blood pressure management in someone with hypertension. For more on the polyphenol mechanisms involved, that page covers how this class of compounds interacts with vascular function.

Theobromine vs Caffeine

An ounce of dark chocolate has about 12mg of caffeine. A cup of coffee has about 95mg. If you’re tracking caffeine intake, dark chocolate barely moves the needle.

What chocolate does have is 130mg of theobromine per ounce. Theobromine is in the same chemical family as caffeine (both are methylxanthines), but it behaves differently. Its half-life is roughly 7 hours versus caffeine’s 5. It crosses the blood-brain barrier less effectively, so its central nervous system stimulation is milder. It acts more on smooth muscle, which is why it’s been studied as a bronchodilator, and it has a different receptor binding profile overall.

The practical result: chocolate produces a gentler, longer-lasting stimulant effect than coffee. If you eat dark chocolate in the evening and then wonder why you’re lying awake at midnight, theobromine is the likely culprit, not caffeine. Its longer half-life means it’s still active hours after you’ve eaten it.

This distinction also matters for pets. Dogs and cats metabolize theobromine very slowly. What’s a minor stimulant dose for a human can be toxic for a dog because of how much longer the compound stays in their system. The toxicity is real and can be severe.

The Heavy Metal Caveat

Consumer Reports tested 28 dark chocolate bars in 2022 and 2023 and found that several contained lead and cadmium at levels that could pose health risks with daily consumption. Both metals are naturally present in cacao, and cadmium in particular is absorbed from soil by the cocoa plant. Processing doesn’t remove them.

This doesn’t mean dark chocolate is unsafe to eat occasionally. EFSA and FDA have both established tolerable intake levels, and 1-2 oz per day from most bars stays below those thresholds for most adults. But it’s a real constraint on the “eat as much dark chocolate as you want” framing that sometimes follows articles about its benefits.

Heavy daily consumption of dark chocolate is not supported by the evidence in the first place. The COSMOS trial used 500mg of flavanols as a supplement, not whole chocolate. Achieving 500mg of flavanols from chocolate bars would require eating a lot of chocolate from bars with uncertain flavanol content. A certified flavanol supplement like CocoaVia, which lists flavanol content on the label and has been used in clinical trials, is the only reliable way to hit that dose. From a whole-food standpoint, 1-2 oz of a non-alkalized bar per day is a reasonable position that captures whatever flavanol benefit is achievable from the food itself without stacking up unnecessary heavy metal exposure.

For more on how the broader category of polyphenols behaves in the body, or on the connection between diet and cardiovascular inflammation markers, those pages extend the context here.

This page is for general nutrition education. It’s not medical advice. Talk to your doctor or registered dietitian about dietary changes that affect your health.