Two studies link ultra-processed foods to cognitive decline, with the strongest signal in adults under 60
STUDY 1: Lee H et al, Am J Public Health 2026 (cohort study, funded NIH)
STUDY 2: Zhang S et al, Journal of Neurology 2026 (meta-analysis of prospective cohort studies, independent funding).
Background
Ultraprocessed foods — packaged snacks, sugar-sweetened beverages, processed meats, ready-to-eat meals — now make up more than half the calories in the average American diet. They’re linked to obesity, diabetes, and cardiovascular disease.
The Cohort Study
- 5,370 US adults aged 50 and older, drawn from the Health and Retirement Study
- Dietary intake assessed by food frequency questionnaire in 2013; cognitive function tracked biennially through 2020
- Participants divided into five groups by ultraprocessed food intake; top fifth (heaviest consumers) compared to the bottom fifth
Compared to the lowest consumers, those in the highest fifth had a 58% higher risk of dementia (hazard ratio 1.58) and a 46% higher risk of cognitive impairment without dementia. The combined risk of either outcome was 47% higher. Eating more unprocessed or minimally processed foods was associated with lower risks across all three outcomes. Among ultraprocessed food subtypes, processed meat drove most of the signal — the top consumers had more than double the dementia risk of the lowest consumers (hazard ratio 2.25).
The Meta-analysis
- Eight prospective cohort studies (not including the new one above), totaling more than 250,000 adults in the US and UK, followed for 3 to 35 years.
- Each study measured overall ultra-processed food intake using the NOVA classification system and tracked cognitive outcomes, including cognitive dysfunction, dementia, and Alzheimer’s disease.
High combined ultra-processed food intake was associated with a 14% higher risk of cognitive impairment (hazard ratio 1.14, 95% CI 1.04–1.25). The effect was strongest in adults under 60 (hazard ratio 1.19), and not statistically significant in those over 60.
In secondary analyses restricted to younger adults, hazard ratios for cognitive dysfunction, dementia, and Alzheimer’s disease specifically were 1.13, 1.13, and 1.17, respectively. Larger studies and those with longer follow-up showed more stable estimates and less variability across studies.
Limitations
Both studies employed observational design, which cannot prove causation. People who eat more processed food also tend to smoke more, exercise less, and have lower incomes. That said, the findings are consistent with prior work from the UK Biobank and the Framingham Heart Study. One controlled trial finds improved cognition through a diet that reduced ultraprocessed foods (in people with metabolic syndrome).
Practice Implications
- The Mediterranean diet avoids ultraprocessed foods. While the support for its cognitive benefits is preliminary, it has robust trials in depression and is broadly recommended for physical health. Here’s a simple version to share with patients.
—Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report
