People who more closely followed the plant-forward EAT–Lancet eating pattern had a lower rate of developing type 2 diabetes in a large UK study. The association was meaningful, but this was not a randomized diet trial and does not prove the diet prevented diabetes.
What changed
Researchers followed 155,413 UK Biobank participants who did not have diabetes at the start. Diet was estimated from at least one 24-hour food recall. Compared with the lowest-adherence group, people in the highest quarter had a 15% to 17% lower rate of developing type 2 diabetes, depending on which EAT–Lancet scoring method was used. Blood-based “biological age” scores statistically accounted for about one-third to 44% of the association.
How strong is the evidence?
Large prospective observational cohort. The sample size and diabetes follow-up are strengths. The article was published online November 11, 2026, after acceptance October 23; the journal notes that final formatting is still pending.
What remains uncertain
Food recalls can misclassify usual eating, and people choosing healthier diets may also exercise more, smoke less, or receive better care. KDM biological age and PhenoAge are calculated surrogate markers—not clocks that prove slower aging. Statistical mediation cannot establish that slower aging caused the lower diabetes rate.
Why it matters
The practical pattern emphasizes vegetables, fruit, whole grains, beans, nuts, and other minimally processed foods while limiting red and processed meat and added sugar. It supports familiar food choices, not a supplement. See our Pantry & Kitchen guide and plant-protein research brief.
Illustrative stock photo by Ella Olsson on Unsplash. It does not show study participants, their meals, blood samples, or results.
Sources
Original study in Frontiers in Nutrition, published November 11, 2026.
