Chocolate: the percentage does not tell you what is inside

Dark chocolate is the one item on this site sold simultaneously as a treat, a health food and a performance aid.

Two of those three have something behind them, the label tells you almost nothing about which product delivers, and the contamination question has an answer that runs opposite to what most people assume.

What cocoa actually contains

The compounds that matter are flavanols, principally epicatechin, and they live in the cocoa solids rather than the cocoa butter.

Flavanols support nitric oxide production in the endothelium, which relaxes blood vessels and improves blood flow. The same pathway described under dietary nitrate, arrived at by a different route.

Dark chocolate carries roughly 2 to 3 times more flavanol-rich cocoa solids than milk chocolate, which is the entire basis of the distinction between them.

The intake associated with modest health effects sits around 200 to 900 milligrams of flavanols daily. And the first problem: almost no product tells you how many flavanols it contains.

The percentage is not the answer

The number on the front describes cocoa content by mass. It does not describe flavanol content, and the two come apart.

A fighter buying the highest number on the shelf on the assumption that more cocoa means more of the active compound has bought a bitterer product rather than a stronger one.

The performance evidence

Smaller and narrower than the marketing suggests, and real within its limits.

In moderately trained young men, 40 grams of dark chocolate daily for 14 days, against white chocolate as control, produced a statistically significant increase of around 17 percent in distance covered.

The protocol matters here. That figure came from a 2-minute all-out effort performed after 20 minutes of cycling at around 80 percent of the first ventilatory threshold, not from a standalone sprint. Oxygen cost during the steady portion also fell.

Separate work found 40 grams daily for 2 weeks reduced oxidative stress markers after prolonged exhaustive exercise and increased mobilisation of free fatty acids afterwards.

Two caveats to hold. The population is moderately trained rather than elite, the standard limitation that has appeared under nitrate, green tea and tart cherry. And the exercise is moderate-intensity cycling with a sprint finish, which is closer to combat sport than a time trial and still not the same thing.

And the question that keeps returning

Reduced oxidative stress markers sounds unambiguously good, and it is the same claim that raised a question under berries and under vitamin C.

The oxidative and inflammatory response to hard training is partly the signal that drives adaptation. An intervention that reliably blunts it may blunt part of what the training was for.

For cocoa flavanols this is theoretical rather than demonstrated, in the same position as tart cherry and unlike high-dose vitamin C and E.

The periodisation answer is identical: if you use cocoa deliberately as an intervention, the windows are competition, travel and congested schedules rather than the middle of a build block.

If you are eating a square in the evening because you want it, none of this applies. The doses in the research are not the doses of enjoyment.

The heavy metal question, properly

Cocoa takes up cadmium and lead from soil, and the amounts depend on where the beans were grown.

A 2024 analysis in Frontiers in Nutrition examined 72 cocoa-containing products purchased in the United States across four sampling years between 2014 and 2022.

That gap between the two rows is the story. Proposition 65 is among the most conservative thresholds in existence, and 97.2 percent of products fell below the FDA level.

Median concentrations of every metal tested were lower than even the Prop 65 levels. The exceedances were driven substantially by outlier products rather than by the category.

The authors' own conclusion: contamination in more than half the products tested may not pose appreciable risk for the average person consuming a single serving, while particular products, or more than one serving daily alongside other dietary sources, could add up to exceed those thresholds.

And the finding nobody expects

Organic products were significantly more likely to show higher levels of both cadmium and lead.

That is the opposite of what most people buying organic chocolate believe they are getting, and the explanation is straightforward: these metals come from soil rather than from agricultural chemicals, and organic certification governs inputs rather than soil geology.

Cacao grown in volcanic soils, including several regions marketed as premium origins, tends to carry more cadmium for the same reason.

Which produces an uncomfortable conclusion. The premium, high-percentage, organic, single-origin bar is not obviously the safer purchase, and may be the worse one on this specific measure.

What a fighter should do with that

The exposure that matters is cumulative rather than acute. Nobody is harmed by a chocolate bar.

The relevant question is whether you eat cocoa products daily for years, and a professional athlete who has used dark chocolate as a nightly habit through multiple camps is in a different exposure category from somebody eating it occasionally.

• Vary the brand and the origin rather than buying the same bar for years, since the contamination is product-specific and driven by outliers.
• Count cocoa powder, hot chocolate, chocolate protein powder and chocolate desserts in the same total rather than counting only bars.
• Keep it to a serving rather than several, which is where the study's own caveat bites.

For anybody genuinely concerned, independent testing organisations publish product-by-product results, and that is more useful than any label.

The theobromine point

Chocolate contains both caffeine and theobromine, and the second gets ignored.

Theobromine is also the reason chocolate is toxic to dogs, and dark chocolate carries considerably more of it than milk.

Two more things that matter in the last week

Tannins and constipation. Dark chocolate is high in astringent tannins, which slow intestinal transit. In a week where gut content is being emptied deliberately, a food that holds material in the intestine is adding weight you were trying to remove, along with the bloating that comes with it.

Tyramine and headaches. Cocoa contains tyramine, and in a fighter who is dehydrated and in a calorie deficit it is a recognised trigger for vasomotor headaches. A fighter who gets a bad head in fight week and has eaten dark chocolate has a candidate he probably has not considered.

The fight week answer

Out, and for six reasons rather than one.

There is no version of this that survives the last 3 days, and no preparation that changes it.

After the weigh-in and after the fight

After the scale, chocolate milk is a legitimate option and worth separating from chocolate.

It delivers carbohydrate, protein and fluid together, and milk performed well on fluid retention. The chocolate is flavour and sugar rather than the point.

After the fight it is food you want and there is no argument against it. That evening is not a nutrition problem.

Milk, white and the rest

Milk chocolate carries a fraction of the cocoa solids and therefore a fraction of the flavanols, with more sugar and milk fat. Nothing on this page about performance transfers to it. It is confectionery, and treating it as a lesser version of a health food is a category error rather than a matter of degree.

White chocolate contains cocoa butter and no cocoa solids at all: no flavanols, and no heavy metals from that source. Its use in the research above is precisely as a control.

Cocoa powder is the most concentrated flavanol source per gram and also the most concentrated source of the metals, and the alkalisation question applies to it more than to bars, because most supermarket cocoa powder is Dutch processed.

Chocolate protein powders and athlete bars are a separate category. They are supplements with the batch-testing requirement attached, and their cocoa content is usually too low for any of this to apply.

And the iron, which is a real point

Cocoa is a meaningful source of non-haem iron, and dark chocolate carries a useful amount per 100 grams.

For a fighter eating little red meat, or a female athlete managing iron status, that is worth knowing rather than dismissing.

Two qualifications. It is non-haem iron, absorbed less efficiently, and it arrives alongside polyphenols that inhibit absorption in the same way described under green tea.

So a contributing source rather than a solution, and the vitamin C timing point applies here as everywhere else non-haem iron appears.

Where it sits by phase

Three things the label will not tell you

Dark chocolate carries real compounds with modest effects at doses most people do not reach, in a category where the front of the packet describes mass rather than potency.

The performance evidence exists, is small, comes from moderately trained people, and shares an adaptation question with several other antioxidant interventions here.

The contamination issue is real, driven by outliers rather than the category, and worse in organic and volcanic-soil products than in the ones you would expect.