News · Brain & Mental Health
Dark chocolate cut 13 seconds off a standard attention test across 16 randomized trials
Cocoa research has produced a decade of enthusiastic small studies and no settled answer. A GRADE-assessed pooling of 16 trials and 2,847 people found the clearest gains on tests of executive function and processing speed.
- Dark chocolate cut 13 seconds off the harder half of a standard attention test.
- The easier half improved too, by about four and a half seconds.
- Half the trials found an effect somewhere, and not always in the same place.
- The gains cluster in executive function and processing speed, not memory.
- 16 trials and 2,847 people, formally graded; a faster test is not a protected brain.
Chocolate and the brain has been one of the more embarrassing corners of nutrition science. Small trials, enthusiastic press releases, an industry with an interest, and no accumulation.
Writing in Nutrition Reviews, researchers went back over it properly. Cocoa and dark chocolate are abundant dietary sources of polyphenols and may improve cognitive performance, and previous studies have been controversial, so they pooled 16 trials covering 2,847 participants and graded the result.
Chronic dark chocolate or cocoa intake significantly improved results of the Trail Making Test, by 4.5 seconds on the simpler version and 13 seconds on the harder one.
What the test actually asks you to do
The Trail Making Test is a piece of paper and a stopwatch, and it has been used in clinics for seventy years.
Part A is a page of numbered circles that you connect in order, as fast as you can. It measures visual scanning and processing speed, and most healthy adults finish it in well under a minute.
Part B alternates: 1, then A, then 2, then B, and so on. Holding two sequences in mind and switching between them is what neuropsychologists call executive function, and it is the ability that degrades first in normal aging and in most kinds of cognitive decline.
The improvement was three times larger on part B, which is the harder and more informative half.
What eight out of sixteen means
Eight studies of the 16 included studies found a significant effect on at least 1 of the cognitive function domains.
That sentence is worth reading carefully, because it is doing two things. It says half the trials found something, which is more than chance would deliver. It also says the something was not the same thing each time: different trials used different tests and found effects in different places.
Pooling rescues part of that. When the same measure is used across enough studies, as the Trail Making Test was, a consistent direction emerges from noise that looks like disagreement at the level of individual papers.
The authors’ summary reflects both halves: chronic dark chocolate or cocoa intake has potential benefits on some dimensions of cognitive function, including executive function, attention, processing speed, language, and rote memory.
Why grading the evidence matters here
Cocoa research is unusually prone to enthusiastic reporting, and the GRADE system exists to slow that down.
It formally downgrades evidence for small samples, inconsistency between studies, indirect outcomes and imprecision, and rates what is left. Applying it to a literature like this is more useful than another positive trial, because the problem was never a shortage of positive trials.
What survives grading is narrower than the headlines of the last decade but it is not nothing: a measurable effect on timed executive tasks, in randomized comparisons, pooled across studies that mostly did not agree on what else to measure.
What this does not establish
A faster Trail Making Test is not a protected brain.
As you grow older, it may take longer to remember things, and cognitive test scores in a trial lasting weeks say nothing about whether that trajectory changes over decades. No trial here followed anyone to dementia, and none could.
The dose question is also unresolved. The trials used cocoa preparations and dark chocolate at concentrations chosen by researchers, not by shoppers, and the flavanol content of commercial chocolate varies enormously between products and is rarely printed on the packet.
The sentence to keep
Dark chocolate contains sugar and fat in quantities that a cognitive benefit measured in seconds does not offset. Nobody in these trials was eating it instead of a meal.
What the review supports is the narrower claim: cocoa flavanols do something measurable to executive function, the effect survives a formal grading of the evidence, and the literature is no longer just a run of small studies pointing hopefully in one direction.
Whether that is a reason to eat chocolate is a question the review does not attempt, and neither should anyone quoting it.
People also ask
What did the review find?
Chronic dark chocolate or cocoa intake significantly improved results of the Trail Making Test part A (weighted mean difference -4.51; 95% CI -8.41 to -0.60; P=.024) and part B (-13.1; -17.73 to -8.47). Eight of the 16 included studies found a significant effect on at least one of the cognitive function domains.
What is the Trail Making Test?
A timed pencil task. Part A asks you to connect numbered circles in order. Part B alternates numbers and letters, which requires holding two sequences at once. Lower times are better, and part B is the one that taxes executive function.
Is 13 seconds a lot?
On part B, where typical completion times run from under a minute to well over two, it is a real difference and not a dramatic one. It sits in the range of what a good night's sleep or a cup of coffee moves, which is worth knowing before anyone buys chocolate as medicine.
What is doing the work?
Probably flavanols, a class of polyphenol abundant in cocoa. They increase blood flow to the brain in short-term studies, which is the mechanism usually proposed. This review measured outcomes rather than mechanism.
Does milk chocolate count?
The trials used dark chocolate or cocoa preparations, which is where the flavanols are. Milk chocolate has far less cocoa solids and far more sugar, and nothing in this review supports it.
Why were the results controversial before?
Because the individual trials are small, use different cognitive tests, and vary in dose and duration. Eight of 16 found an effect on at least one domain, which is the pattern that produces both enthusiastic headlines and unconvinced reviewers.
What are the limits?
Cognitive test scores are not the same as cognitive health, the trials were mostly short, and the specific domains that improved varied between studies. This is a signal worth following, not a reason to change what you eat.