Photo by Caleb Wright on Unsplash
In A Nutshell
- A large UK study found that each extra 25 grams of processed meat eaten daily was linked to a 44 to 52 percent higher risk of dementia and Alzheimer’s disease.
- A separate modeling analysis estimated that swapping 10 percent of daily calories from ultraprocessed foods for unprocessed food could lower dementia risk by 19 percent.
- Higher ultraprocessed food intake has also been tied to depression, anxiety, and signs of brain shrinkage on scans.
- All of the human evidence is observational, so it cannot prove cause and effect, and most clues about how these foods might harm the brain come from animal studies.
Eating just one more ounce of processed meat every day was tied to as much as a 52 percent higher risk of developing dementia, according to a massive UK study cited in a new scientific review. Researchers drew on health data from nearly 500,000 people and found that each additional 25 grams of processed meat consumed daily, roughly the weight of two slices of deli turkey, corresponded with a 44 to 52 percent increase in the risk of all-cause dementia and Alzheimer’s disease specifically.
Published in Annual Review of Nutrition, the review pulls together decades of research on ultraprocessed foods and brain aging. Processed meat is only one piece of a much bigger picture that includes sodas, packaged snacks, and other heavily manufactured products now making up more than half the calories many Americans eat in a day.
Researchers behind the review aren’t claiming a hot dog directly causes memory loss. Instead, they point to a growing, if still incomplete, body of evidence that heavily processed food may work against the brain over the course of a lifetime, through inflammation, damage to blood vessels, and disruption of the chemical signals brain cells use to communicate. Its authors describe the biology behind these links as poorly understood and call for better-designed human studies before anyone draws firm conclusions.
Ultraprocessed Food and Dementia Risk: Defining the Problem
Ultraprocessed foods are not simply foods that have been touched by a factory. Nearly every packaged item on a grocery shelf has gone through some kind of processing. According to the review, ultraprocessed foods stand apart because two things happen at once: natural ingredients like fiber and plant compounds get stripped out, while sugar, fat, salt, and lab-made additives like sweeteners, colorings, and thickeners get added in. Sodas, packaged snack cakes, frozen ready meals, and many processed meats all fit the bill.
Consumption has been climbing worldwide, but the amount people actually eat varies sharply by country. In the United States, ultraprocessed foods make up somewhere between 55 and 58 percent of daily calories, among the highest rates in the world, and the United Kingdom similarly exceeds 50 percent. Brazil sits at around 21.5 percent of daily calories and Mexico at roughly 29.8 percent, according to the review. Italy had the lowest rate among the countries studied, at about 10 percent.

Processed Meat, Soda, and Dementia Risk in Human Studies
Rather than running a single new experiment, a team of 26 researchers from institutions around the world combed through existing studies, both from labs working with animals and from long-term studies following human populations, to build a picture of what’s known and what remains uncertain about ultraprocessed food and brain aging.
Beyond the processed meat numbers, a separate analysis of UK Biobank data used mathematical modeling to estimate that replacing 10 percent of calories from ultraprocessed foods with unprocessed food could lower dementia risk by 19 percent. In Brazil, a study following more than 10,000 adults for about eight years found that people who got more than roughly 20 percent of their daily calories from ultraprocessed foods experienced faster declines in memory and reasoning skills, though this effect showed up only in people who didn’t closely follow a brain-healthy eating pattern known as the MIND diet.
Mental health showed up as another area of concern. Pooled data across multiple large reviews found that people with higher ultraprocessed food intake had roughly 22 to 53 percent higher odds of depression or anxiety symptoms, depending on which outcome and dataset researchers examined. Brain scan studies linked higher intake of fat, sugary sodas, meat, and processed meat products to smaller brain volume, while sugary sodas and certain fats were tied to more damage in the brain’s white matter, the tissue that helps different brain regions talk to each other.
Animal research offered a window into possible mechanisms that human studies can’t easily test. Mice bred to develop Alzheimer’s-like disease and fed high-fat diets with added sugar or salt built up more of the toxic proteins linked to Alzheimer’s in the brain’s memory center. Other lab studies pointed to specific additives: mice given aspartame, an artificial sweetener, over long periods showed brain changes associated with Parkinson’s disease, while titanium dioxide, a whitening agent used in some processed foods, impaired memory and learning in animal studies and killed brain cells grown in the lab.
Reviewers were careful to note that confidence in this evidence varies quite a bit. Across mental health, memory and thinking, and dementia, the best available evidence earned a “moderate” rating in the review’s own tally. Brain scan research fared somewhat better, while studies of blood and spinal fluid markers numbered just four, and no studies at all have yet examined brain electrical activity.
Mixed Results on Dementia Risk and Brain Health
Science rarely delivers a clean, unanimous verdict, and this review doesn’t pretend otherwise. One study found a strange pattern in which people who ate either very little or very much ultraprocessed food scored worse on tests of memory and thinking, suggesting the relationship might not be a simple straight line. Other research on general “Western-style” diets found no clear link to memory decline at all, a contradiction the reviewers attribute to differences in how studies measured diet, how long they followed people, and which mental tests they used.
Researchers also flagged a basic scientific headache: it’s tough to separate the effects of processing itself from the effects of the added fat, sugar, and salt these foods typically contain. Untangling the two will likely require future trials designed specifically to isolate that difference, something the field has largely not attempted yet.
A Case for Caution, Not Panic
Trials proving cause and effect are still missing, and the authors say so plainly. Even so, diets heavy in ultraprocessed foods keep showing up alongside worse outcomes for memory, mood, and brain structure, across multiple countries and multiple types of studies. With no cure for dementia and lifestyle changes among the few tools available for lowering risk, trading some of the deli meat and packaged snacks for whole foods looks like a low-cost bet, backed by a body of science that is growing but still has real gaps to fill.
Disclaimer: This article summarizes a scientific review of existing research, not a new clinical trial. The human studies it describes are observational, meaning they show associations between diet and brain health outcomes but cannot prove that ultraprocessed foods or processed meat cause dementia, depression, or other conditions. Findings from animal and cell studies have not been confirmed in humans. This content is for informational purposes only and is not medical advice; readers should talk with a doctor or registered dietitian before making significant changes to their diet.
Paper Notes
Limitations
This review is a synthesis of existing research rather than a single new experiment, and its authors are candid about the gaps in that underlying evidence. Much of the strongest mechanistic evidence comes from animal studies, which only partially mirror human biology, behavior, and genetics. Human studies on the topic remain limited in number, often rely on self-reported food questionnaires not specifically designed to measure ultraprocessed food intake, and there is no single, universally agreed-upon system for classifying which foods count as ultraprocessed. Randomized controlled trials, the gold standard for proving cause and effect, are largely missing from this field because tracking the long-term brain effects of diet changes is difficult and expensive. According to the authors, understanding of the biological mechanisms remains incomplete, and they call for more rigorous, longer studies before firm conclusions can be drawn.
Funding and Disclosures
According to the paper’s disclosure statement, lead author Owen T. Carmichael chairs the Alzheimer’s Association ISTAART Nutrition, Metabolism and Dementia professional interest area and has a consulting agreement with Clario, Inc. Debora Melo van Lent is the immediate past chair of that same professional interest area, and Hom Lal Shrestha is a board member of Dementia Alliance International. ISTAART, the Alzheimer’s Association’s International Society to Advance Alzheimer’s Research and Treatment, facilitated the manuscript through its Nutrition, Metabolism and Dementia and Neurophysiology professional interest areas, and the authors note that the views expressed are their own and do not necessarily reflect those of that organization. In the acknowledgments, Carmichael reports research support from the National Institutes of Health, Eli Lilly Corporation, and the Pennington Biomedical Research Foundation; other authors list support from sources including the Alzheimer’s Association and national research agencies in several countries.
Publication Details
Titled “Ultraprocessed Foods and the Aging Brain: State of the Science,” the paper was authored by Owen T. Carmichael, Puja Agarwal, Benedict C. Albensi, Vyshnavy Balendra, Nathan M. D’Cunha, Diandra N. Denier-Fields, Naomi Vidal Ferreira, Natalia Gomes Gonçalves, Rashi Kumar, Corinne Labyak, Susanna Lopez, Mychael V. Lourenco, Ilianna Lourida, Rishav Nag, Zuzana Nedelska, Juliet Nnenda Olayinka, Jung Hyun Park, Kayalvizhi Rajendran, Ashley Shaw, Hom Lal Shrestha, Aaron N. Smith, Natalia Soldevila-Domenech, Débora G. Souza, Claudia Kimie Suemoto, Jodi R. Titiner, Rafael de la Torre, and Debora Melo van Lent. Carmichael, the corresponding author, is based at the Pennington Biomedical Research Center in Baton Rouge, Louisiana. It was published in Annual Review of Nutrition in 2026 (volume 46, pages 533-561), first appearing online as a Review in Advance on May 15, 2026, with the DOI 10.1146/annurev-nutr-061824-053453.







