Original studyJAMA Network Open, 2026Read the paper →
Ultra-processed foods make up a large share of what many people eat, from packaged snacks and sugary drinks to ready meals. Most of the worry about them has centred on weight, heart disease, and diabetes. A large Canadian study now suggests that eating more of these foods is also linked to silent damage in the brain in middle age, before any change in thinking skills can be seen.
Researchers at McMaster University and partner institutions across Canada analysed data from 8,783 adults aged 35 to 69 with no heart disease. The study appears in JAMA Network Open. Participants came from two large Canadian research programmes.
All measurements were taken at a single visit. Each person reported their eating habits over the previous year, had their body fat measured, and underwent MRI scans of the brain and abdomen. They also took two short thinking tests, and the average age was about 58.
Ultra-processed foods are industrially made products, often high in added sugar, salt, saturated fat, and additives. Vascular brain injury here meant either small silent strokes, known as covert brain infarcts, or a heavy build-up of white matter changes, which signal disease in the brain’s small blood vessels. Neither usually causes symptoms, but both are warning signs for future stroke and are studied as early markers of cognitive ageing.
Participants were split into four equal groups by how much ultra-processed food they ate for every 2,000 calories. Those in the top group had about 40% higher odds of vascular brain injury than those in the bottom group, after accounting for age, sex, income, education, smoking, exercise, diabetes, high blood pressure, and total calories.
Brain injury appeared on the scans of 11.3% of the highest consumers, compared with 8.1% of the lowest. The top group also carried more body fat, including more of the deep visceral fat around the organs.
Thinking skills told a different story. Scores on the Montreal Cognitive Assessment and the Digit Symbol Substitution Test, which assess memory, attention, and mental speed, did not differ between the highest and lowest consumers. The researchers suggest that damage to blood vessels may appear on scans years before any decline shows up on brief tests.
The study has clear limits. Because diet and brain scans were measured at the same time, it cannot show that ultra-processed food caused the damage. The brain finding was also sensitive to how intake was counted, and when it was measured as a share of calories or of total food servings, the link was no longer statistically significant.
Diet was self-reported, and the group was mostly well educated. People eating the most ultra-processed food also tended to have lower incomes, smoke more, and exercise less, so other factors may play a part.
The findings add to growing concern about diet and brain health, following earlier research that linked ultra-processed foods to a higher risk of depression. Studies that follow people over many years will be needed to show whether eating less of these foods protects the brain with age.
Eating more ultra-processed food was linked to silent brain injury in middle-aged adults, but not to poorer thinking skills.
