A massive new study tracking more than 200,000 adults in the United Kingdom suggests that a diet heavy in ultra-processed foods may be linked to a higher risk of certain blood cancers, according to research published in The American Journal of Clinical Nutrition.
Researchers followed 204,494 participants enrolled in the UK Biobank for a median of 10.5 years, analyzing dietary questionnaires to estimate the share of daily calories coming from ultra-processed sources — packaged snacks, sugary drinks, processed meats and many ready-made meals. During that period, 848 participants developed myeloid neoplasms, a group of cancers affecting blood-forming cells in the bone marrow.
For every one-standard-deviation increase in the percentage of calories from ultra-processed foods, participants faced a 12% higher relative hazard of developing a myeloid neoplasm. The association was stronger for myeloproliferative neoplasms (MPNs), which cause the bone marrow to overproduce blood cells, where the relative hazard rose 15% per standard-deviation increase.
The biological clues researchers are chasing
The team also examined clonal hematopoiesis of indeterminate potential (CHIP), an age-related condition involving mutations in blood-forming cells that can raise blood cancer risk. The link between ultra-processed food intake and myeloid neoplasms persisted even among participants without detectable CHIP at baseline, suggesting the association is not solely explained by that pathway.
Elevated markers of inflammation, immune dysregulation and altered fat metabolism also correlated with higher ultra-processed food consumption, offering potential biological mechanisms — though the study cannot prove causation.
What the numbers actually mean for your pantry
These percentages describe differences between groups with different long-term eating patterns, not an individual’s absolute risk from a single snack bar or frozen dinner. The researchers adjusted for age, smoking, physical activity, body measurements and overall diet quality, but observational studies cannot fully rule out residual confounding.
Nutritional quality varies widely within the ultra-processed category; a fiber-rich whole-grain cereal and a sugary soda share the label but not the same metabolic impact. The findings point to cumulative dietary patterns over years, not occasional convenience foods.