Ultra-Processed Foods May Raise Dementia Risk by 58%

Hot dogs, packaged snacks, sugary cereals. Most of us know they're not great. But a study published this month in the American Journal of Public Health puts a sharper number on the brain cost: people who eat the most ultra-processed foods had a 58% higher risk of developing dementia — and a 46% higher risk of cognitive impairment — compared to those who ate the least.

That's not a marginal effect. That's a large, prospective study of 5,370 adults over 50, tracked for close to nine years.

Key Takeaway: A Harvard-led study found that the highest ultra-processed food consumers had a 58% greater chance of developing dementia and 46% greater cognitive impairment risk compared to the lowest consumers, over a median follow-up of nearly nine years.

What counts as ultra-processed?

The NOVA classification system (developed by Carlos Monteiro at USP) defines ultra-processed foods as industrial formulations made mostly from substances extracted from foods — sugars, fats, starches, protein isolates — plus additives like colorings, flavorings, emulsifiers, and preservatives you'd never keep in a home kitchen.

Practically speaking: packaged chips, deli meats, soft drinks, flavored yogurts, instant noodles, mass-produced bread, most fast food. In the US, these foods account for more than 50% of daily calories for the average adult.

Stat: Ultra-processed foods make up more than half of daily calorie intake for the average American adult, according to NIH data.

What the Harvard study found

Researchers used data from the University of Michigan's Health and Retirement Study — a nationally representative longitudinal survey of roughly 20,000 Americans. For this analysis, they tracked 5,370 adults aged 50 and older from 2013 to 2020 (median follow-up: 8.7 years).

Diet was assessed at baseline via a detailed food frequency questionnaire. Cognitive status was tracked at each wave. The team controlled for education, income, smoking, alcohol use, and physical activity — the usual suspects that confound dietary studies.

The top quintile of ultra-processed food intake, compared to the bottom quintile: 58% higher dementia risk, 46% higher cognitive impairment risk.

What's equally striking is where the risk starts. Senior author Cindy Leung (associate professor of public health nutrition at Harvard) told the Wall Street Journal: "There may not be a safe level." Even moderate consumption — roughly a pound of UPF per day, which is unremarkable for many people — showed measurable associations.

Key Takeaway: Even moderate ultra-processed food consumption (around 450g per day) was associated with higher cognitive risk, not just extreme intake. This suggests there is likely no "safe floor" below which the risk disappears entirely.

Why would processed food affect the brain?

The study is observational, so it can't prove causation. But several plausible mechanisms have been proposed.

First, gut-brain axis disruption. Additives and emulsifiers in ultra-processed foods may alter the composition of gut bacteria, triggering low-grade inflammation that reaches the brain over time. Research on the gut-brain connection is still developing, but a recent gut microbiome study found that even a single bacterial species can have measurable downstream effects on metabolism and weight.

Second, nutrient displacement. People who eat a lot of ultra-processed foods tend to eat fewer minimally processed foods — whole grains, vegetables, legumes — which are rich in fiber, antioxidants, and polyphenols that have been linked to better brain aging. The Harvard study found that high consumption of minimally processed foods was associated with lower dementia risk, which supports this explanation.

Third, metabolic damage. Ultra-processed diets are associated with higher rates of obesity, type 2 diabetes, and cardiovascular disease — all of which are established risk factors for dementia. The harm to the brain may partly flow through these intermediate conditions.

Stat: Processed meats like bacon showed the strongest individual link to dementia and cognitive impairment in the Harvard analysis, more than any other UPF subcategory.

The foods that appeared protective

The same study identified foods associated with lower dementia risk: whole grains, fruits, and vegetables — "minimally processed foods," as Harvard describes them. These patterns broadly map to what research says about diet and mental health, where similar anti-inflammatory, whole-food diets showed consistent benefits.

This isn't surprising. The brain is metabolically expensive — it consumes about 20% of the body's energy — and it's also one of the most sensitive organs to what circulates in the bloodstream over decades.

What this means practically

The study doesn't suggest you need to eliminate every processed food. What it does suggest is that the overall composition of your diet matters, accumulated over years. A few things that seem to follow from the data:

• Whole foods most of the time. Fruits, vegetables, legumes, whole grains — the boring stuff that consistently shows up in the protective category.
• Minimizing processed meats in particular. Bacon, deli meats, hot dogs showed the strongest dementia signal in this study.
• Reading labels for additives. If the ingredient list reads like a chemistry class, it's almost certainly ultra-processed.
• Being aware of displacement. If UPF is filling your plate, something else isn't. That "something else" is often what the brain needs most.

Aware of how your food environment shapes what you reach for? The research is consistent: changing what's around you tends to work better than willpower alone.

Key Takeaway: No single food causes dementia. The evidence points to overall dietary patterns built up over years — with ultra-processed foods appearing to accumulate risk and whole, minimally processed foods appearing to reduce it.

Limits to keep in mind

This is an observational study. Diet was self-reported at a single time point. The linear trend across quintiles didn't reach statistical significance, which means the dose-response relationship is suggestive but not ironclad.

What it adds to is a growing body of evidence — previous meta-analyses, cohort studies in Europe and the US — all pointing in the same direction. The consistency across different populations, different methods, and different outcomes is the more compelling signal here.

Sources

• Harvard T.H. Chan School of Public Health, June 10, 2026
• Samuthirapandian R et al., "Ultraprocessed Foods and the Risk of Cognitive Impairment and Dementia in Older US Adults: 2013-2020 Health and Retirement Study," American Journal of Public Health, 2026
• The Independent, June 3, 2026
• PsyPost, 2026
• NOVA Food Classification, Carlos Monteiro, University of São Paulo

FAQ

What are ultra-processed foods? Ultra-processed foods are industrial formulations made primarily from food-derived substances (refined sugars, fats, starches, protein isolates) combined with additives like emulsifiers, colorings, and preservatives. Common examples include packaged snacks, soft drinks, deli meats, instant noodles, and most fast food.

Does eating ultra-processed food definitely cause dementia? No. The Harvard study is observational, meaning it shows an association, not a causal link. Researchers can't rule out confounding factors entirely. That said, the biological mechanisms are plausible, and the finding is consistent with prior research.

Which ultra-processed foods were most strongly linked to dementia risk? In this study, processed meats — bacon, deli meats, hot dogs — showed the strongest individual association with dementia and cognitive impairment. Sugary packaged foods were also in the high-risk category.

What foods may help protect the brain? The study found that high consumption of minimally processed foods — fruits, vegetables, and whole grains — was associated with lower dementia risk. Diets rich in fiber, polyphenols, and antioxidants (such as the Mediterranean diet) have also shown consistent associations with better cognitive aging.

Does this apply to all age groups? The study focused on adults aged 50 and older. What seems likely is that dietary patterns built over decades shape cumulative risk — so what you eat in your 40s and 50s probably matters too.

-- Selena