Berry powder 'lowered anxiety compared with cocoa' in a 12-week trial of 59 adults. No group's score actually changed, and nobody got a placebo

Berries and cocoa are two of the best-known sources of polyphenols, the plant compounds that keep turning up in research on mood and the ageing brain. Berries are rich in anthocyanins, the pigments that make them red and blue; cocoa is rich in flavanols. A team at the Spanish National Research Council's food science institute in Madrid (ICTAN-CSIC) set out to see whether a small daily amount of either, or both together, would shift anxiety or depression scores in older adults. Their paper was published in Molecular Nutrition & Food Research on 7 October 2026.

Of 144 volunteers, 60 signed up and 59 finished. Men had to be at least 50 and women past the menopause; the group ended up aged 47 to 81. All were mentally healthy by design: anyone with symptoms or a history of depression or anxiety was excluded. They were randomly assigned, without knowing which product they had, to 5 g a day of a red-berry powder (20 people), 2.5 g a day of a polyphenol-rich cocoa powder (20), or 7.5 g of the two combined (19), stirred into water, juice, yogurt or cereal at breakfast for 12 weeks. Before and after, a neuropsychologist sat with each person while they filled in two standard questionnaires: the Beck Depression Inventory and the State-Trait Anxiety Inventory, which separates how anxious you feel right now (state) from how anxious you tend to be (trait).

The headline finding is that berries 'significantly reduced state anxiety compared with cocoa'. Here is what sits behind it. The berry group's average state-anxiety score went from 16.05 to 13.91. The cocoa group's went from 12.30 to 15.35. Neither change was statistically significant on its own (P = 0.185 and P = 0.241), and nor was any other before-and-after change in any group on any of the three scores. Taking all 59 people together, depression went from 6.02 to 5.36 and state anxiety from 14.82 to 13.8, both well within chance. What passed the test was the gap between the two groups' changes: one going down a little, the other going up a little. The authors say as much in the body of the paper: 'In the present study, no significant effects were observed.'

From here the reading is ours. First, there was no placebo group, so the trial can compare berry powder with cocoa powder but cannot say whether either did anything compared with doing nothing. Second, the two groups did not start level. The berry group began with the highest anxiety score of the three and the cocoa group with the lowest, and scores at the extremes tend to drift back toward the middle on a second measurement whatever people eat. Third, the significant result is reported as z = -1.88 with P = 0.031. A z of 1.88 gives a P of about 0.03 only if the test is one-sided; the usual two-sided version would be about 0.06, over the paper's own 0.05 line. The paper does not say which was used. And it was one of nine possible pairings (three scores, three pairs of groups), with no adjustment mentioned.

The second half of the paper is about urine. The team looked for chemical traces of what people had eaten, and the clearest signal was a simple one: two breakdown products of theobromine, cocoa's main stimulant, rose to roughly two and a half to three and a half times their starting level in the people taking cocoa alone and did not move in the berry group. That is useful as proof that people took their powder. The team then tested whether changes in a list of urine compounds tracked changes in mood scores. Most links faded once the number of comparisons was accounted for. One held: people whose p-cresol glucuronide, a compound that starts with gut bacteria breaking down protein, rose more tended to see their trait anxiety fall. Another ran against expectations, with higher epicatechin sulfate (a cocoa flavanol product) going with more anxiety, not less, which the authors flag as needing caution.

The authors list their own limits plainly: a small sample, a dose kept deliberately low so that it stays within what a person could eat, and volunteers who had little anxiety or depression to lose in the first place. They call for trials in people who actually have those conditions. That seems the fair summary. Berries are a good food for many reasons, and nothing here counts against them or against cocoa. But this study does not show that a spoonful of berry powder calms anyone down. The work was funded by Spain's state research agency and European regional funds, and the authors declare no conflicts of interest.