UK Biobank data link low ultra-processed food intake plus 150 minutes weekly activity to 49-70% lower type 2 diabetes incidence across genetic risk strata

The analysis used baseline Nova-classified food-frequency data and one-week wrist accelerometry to quantify ultra-processed food weight percentage and moderate-to-vigorous physical activity. Incident type 2 diabetes was ascertained via hospital and primary-care records; polygenic risk scores stratified participants into high, intermediate, and low genetic susceptibility. Each 10% higher ultra-processed food proportion raised diabetes hazard by 7%, while meeting 150 minutes moderate-to-vigorous activity cut risk 39%. In the high-genetic-risk group, the combined low-ultra-processed-food and activity pattern halved incidence.

These observational associations align with prior UK Biobank and EPIC-InterAct reports showing independent protective gradients for physical activity and detrimental gradients for ultra-processed food. The present work adds an interaction test indicating activity attenuates ultra-processed-food risk even within high-genetic-risk strata, yet residual confounding by socioeconomic position, overall diet quality, and unmeasured behaviors remains possible. Reverse causation is limited by the prospective design but cannot be excluded.

Modifiable lifestyle levers therefore retain leverage irrespective of inherited risk. Public-health messaging should emphasize simultaneous reduction of ultra-processed food and attainment of activity targets rather than isolated changes. Next required evidence is a randomized trial testing joint dietary-plus-activity interventions with objective adherence measures and diabetes incidence as primary endpoint.