Berries, tea, coffee, cocoa and olive oil are often promoted as foods that can protect the aging brain. A new review finds reasons these foods may support brain health, but it does not prove that their plant compounds reduce the risk of dementia.
Published in Nutrients, the paper examined previous laboratory, animal, population-based and clinical research on polyphenols, a large family of compounds produced by plants. The authors explored how these compounds might affect inflammation, oxidative stress, cellular energy, the gut microbiome and other biological processes associated with brain aging.
This was a narrative review, not a new clinical trial, systematic review or meta-analysis. The authors did not test a diet in participants or calculate an overall reduction in dementia diagnoses. Instead, they brought together findings from many different types of studies, which vary considerably in their ability to show whether a food or compound improves human health.
“Polyphenols are not miracle cures, but research suggests they may be promising tools for supporting healthy brain aging,” said Mónika Fekete, an assistant professor at the Institute of Preventive Medicine and Public Health at Semmelweis University and the review’s senior author. “The focus, however, should not be on dietary supplements but on a varied diet rich in plant-based foods.”
Polyphenols include thousands of compounds found throughout the plant kingdom. Among those discussed in the review were anthocyanins, which give many berries their red, blue and purple colors; flavanols found in cocoa; EGCG from green tea; curcumin from turmeric and compounds found in extra virgin olive oil.
These substances can behave differently in the body. Evidence involving one polyphenol cannot automatically be applied to the entire category, and findings from concentrated extracts or supplements may not reflect what happens after eating an ordinary serving of food.
Many of the studies included in the review focused on possible biological mechanisms. In laboratory and animal experiments, certain polyphenols have shown antioxidant or anti-inflammatory activity and have affected pathways involved in nerve cell function, protein accumulation and cellular energy production.
Those findings can help researchers identify promising questions, but influencing a biological pathway is not the same as preventing dementia. A compound may change inflammation in cells or improve memory in an animal model without producing a meaningful benefit in people.
Human evidence is also difficult to interpret as one unified body of research. Some studies track what people eat and compare those habits with later cognitive outcomes. Others test supplements or measure changes in memory scores, blood markers or brain imaging.
These outcomes are related, but they are not interchangeable. A change in an inflammatory marker does not necessarily translate into preserved memory, and performing better on a short-term cognitive test does not establish that someone is less likely to develop Alzheimer’s disease years later.
The review paid particular attention to Mediterranean-style and MIND eating patterns. Both emphasize combinations of plant foods rather than a single ingredient.
Mediterranean-style diets commonly include vegetables, fruits, legumes, whole grains, nuts, fish and olive oil. The MIND diet draws from Mediterranean and heart-supportive eating patterns while placing additional emphasis on leafy green vegetables and berries.
Some observational studies have linked closer adherence to these patterns with slower cognitive decline or a lower risk of dementia. However, people who follow these diets may also differ in physical activity, education, health care access, smoking, income and other factors that affect brain health.
Diet-pattern research also cannot isolate polyphenols as the source of any apparent benefit. These eating patterns differ in fiber, types of fat, vitamins, minerals, fish intake and overall food quality, among other characteristics.
The paper also explored how gut bacteria metabolize polyphenols. A substantial proportion of these compounds is not absorbed in its original form. Instead, microbes in the digestive system transform them into other substances that may enter circulation and affect inflammation or cellular metabolism.
“This may help explain why the same diet does not affect everyone in the same way,” said Noémi Mózes, an assistant professor at the Institute of Preventive Medicine and Public Health at Semmelweis University and the review’s first author. “In the future, personalized nutrition could help us better understand who is most likely to benefit from a polyphenol-rich diet.”
That is a plausible area for future research, but it is not yet a practical personalized nutrition tool. Scientists cannot currently examine a person’s microbiome and reliably prescribe the right berry, tea or polyphenol for dementia prevention.
The review’s discussion of supplements also deserves caution. Concentrated curcumin, green tea extracts and other polyphenol products may deliver doses far beyond what people obtain from food. They may also differ in absorption, purity and safety.
Evidence involving supplements should not be used to promise that a food will have the same effect. The reverse is also true: Findings involving a whole dietary pattern cannot establish that an isolated supplement captures its benefits.
The strongest message from the review is therefore broader and less dramatic than the idea of eating particular foods to ward off dementia.
A varied eating pattern that includes vegetables, fruits, legumes, whole grains, nuts and other plant foods may support cardiovascular and metabolic health, both of which are connected with brain health. But no berry, beverage, spice, nutrient or supplement has been shown to prevent Alzheimer’s disease on its own.
The authors acknowledge that there is not enough evidence to recommend one food or compound as a stand-alone dementia prevention strategy. Future randomized trials will need to determine whether specific dietary changes produce lasting improvements in cognitive function or reduce diagnosed dementia, rather than simply changing short-term biomarkers.
The review received public and academic support through Hungary’s National Research, Development and Innovation Fund, the Ministry of Innovation and Technology, the National Cardiovascular Laboratory Program, the European University for Well-Being program and Semmelweis University’s Cooperative Translational Research Program.