Un neurologue alerte sur les chocolats noirs très corsés : une étape de leur fabrication risque de diminuer les antioxydants présents dans la tablette

An intensely dark chocolate bar can seem like the obvious choice for someone who enjoys cocoa and hopes to get more of its beneficial plant compounds. The percentage is prominent, the flavour is bold, and the colour looks reassuringly rich. Yet a processing decision made long before the bar reaches the shelf can complicate that picture. Treating cocoa with an alkali—known as alkalisation or Dutch processing—can make it darker and mellower while substantially reducing its flavanol content.

The caution has been voiced in a discussion of chocolate and brain health featuring Gabriele Gratton, a psychology professor whose work concerns neuroscience. He noted that Dutch processing and roasting can reduce the flavanols present in cocoa, so a chocolate bought in a shop will not necessarily contain the amounts of these compounds that people associate with research on cocoa. That is a reason to read claims about dark chocolate carefully, not a reason to fear an occasional square.

What cocoa percentage leaves out

The percentage printed on a dark chocolate wrapper describes its cocoa content, not a laboratory measurement of flavanols. Cocoa solids contain these plant compounds, but their quantity can change during production. Two bars with the same advertised cocoa percentage may therefore differ in the flavanols they deliver. Colour and bitterness are imperfect clues, too: a darker appearance does not prove that more flavanols remain.

That distinction is easy to miss while shopping. A bar labelled 85% cocoa may contain more cocoa-derived ingredients than a sweeter alternative, but the number alone says nothing precise about how the cocoa was fermented, roasted or alkalised. Nor does a strong taste act as a reliable flavanol test. It tells you whether you like eating the chocolate; it does not reveal its chemical profile.

This does not make high-cocoa chocolate meaningless. Dark chocolate generally contains more cocoa solids than milk chocolate and can be a source of flavanols. The point is narrower: “more cocoa” and “more preserved flavanols” are related ideas, not interchangeable promises. Processing belongs in the conversation alongside the percentage on the front of the pack.

The step behind the warning

Alkalisation is an optional stage in cocoa processing. A manufacturer treats cocoa material with an alkaline substance to change its acidity. The treatment can alter colour and flavour and help cocoa powder disperse in water. Depending on the production method, it may be applied to cocoa nibs, cocoa mass or pressed cocoa cake. That is why describing it solely as a treatment of finished chocolate would be misleading.

There are understandable reasons to use it. Cocoa begins with sharp, bitter and astringent notes that are not to everyone’s taste. Processing helps develop the familiar flavour and texture people expect from chocolate. Alkalisation can contribute a smoother taste and a deep colour, but those attractive qualities come with a potential trade-off: some flavanols are lost or changed.

It is not the only stage that matters. Fermentation, drying and roasting can also affect cocoa’s polyphenols. A scientific review found that roasting generally reduces several of these compounds, although outcomes vary with conditions and with the compound measured. The final bar reflects an entire chain of choices, not one switch flipped on or off at the factory.

How much can be lost?

A study of commercial cocoa powders offers a clear illustration. Researchers measured an average total flavanol content of 34.6 milligrams per gram in natural, non-alkalised powders. The averages were 13.8 milligrams per gram for lightly alkalised powders, 7.8 for moderately alkalised powders and 3.9 for heavily alkalised powders. In the samples studied, flavanol levels fell as the degree of alkalisation increased.

Those are figures for cocoa powders, not a conversion chart for chocolate bars. A finished bar contains a mixture of ingredients and has undergone additional processing; the study cannot tell you how many milligrams of flavanols are in any particular square. It does, however, show why a very dark colour should not be treated as proof of high antioxidant content. The colour can be influenced by the same treatment that reduces flavanols.

“Antioxidants” also covers a broad set of compounds and laboratory measurements. Cocoa flavanols, including catechin, epicatechin and related compounds, are central to much of the health discussion. A lower measured flavanol level is a meaningful change to the food’s composition, but it does not let us calculate an individual person’s health outcome from a wrapper or a shade of brown.

What about brain health?

Interest in cocoa and the brain has a plausible starting point: researchers study how flavanols interact with blood vessels and other biological pathways. But a plausible mechanism is not the same as proof that eating a commercial dark chocolate bar will protect memory or prevent cognitive decline. The amount of flavanols used in a study may differ greatly from the unknown amount in a shop-bought product.

One useful reality check comes from the COSMOS trial. In a group of 573 older adults assessed in person, daily cocoa extract providing 500 milligrams of flavanols did not produce a statistically significant overall improvement in global cognition over two years compared with placebo. The researchers reported a possible signal among participants whose diets were poorer at the start, but said that finding needed further study. The intervention was a measured extract, not an ordinary chocolate bar.

Cardiovascular claims deserve similar care. The US Food and Drug Administration has said that the credible evidence for a reduced cardiovascular risk claim involving specially formulated, high-flavanol cocoa powder is very limited. Its qualified claim does not apply to regular cocoa powder or chocolate. Neither an impressive cocoa percentage nor the presence of flavanols turns a dessert into a proven preventive treatment.

Choosing chocolate without guesswork

If you want dark chocolate primarily because you enjoy it, choose a flavour you like. There is no need to turn every square into a nutritional calculation. If preserving cocoa flavanols is part of your aim, look beyond the front-of-pack percentage. A manufacturer’s statement that its cocoa is non-alkalised is more relevant to this particular issue than a dramatic colour or a claim that the bar is “extra dark.” Even then, without a measured flavanol amount, you cannot know its precise content.

For cocoa powder, the distinction may be easier to spot because products are sometimes described as natural or Dutch-processed. Natural, non-alkalised cocoa generally retains more flavanols than cocoa treated with alkali. That does not mean it must replace Dutch-processed cocoa in every recipe: the two have different flavour and preparation qualities. It simply lets you make a more informed choice about which characteristic matters most to you.

Portion size remains relevant regardless of processing. Dark chocolate is an enjoyable food that also contains substantial energy and fat, so eating much more of it in pursuit of antioxidants is not a sound shortcut. The practical lesson from the neurologist’s warning is modest but useful: savour the bar for what it is, and be sceptical of the assumption that the darkest-looking chocolate automatically offers the most flavanols.

Repères en français