Summary
Background & aim
The association between ultra-processed foods (UPF) consumption and the risk of type 2 diabetes (T2D) has not been much explored. We aimed to evaluate the association between consumption of UPF and the incidence of T2D.

Methods
We assessed 20,060 participants (61.5% women) from the SUN project (Seguimiento Universidad de Navarra) followed-up every two years (median follow-up 12 years). Food and drink consumption were evaluated through a validated 136-item food frequency questionnaire and grouped according to their degree of processing by the NOVA classification. Participants were categorized into tertiles of UPF consumption adjusted for total energy intake. We fitted Cox proportional hazard models with repeated dietary measurements at baseline and updating information on food consumption after 10 years of follow-up to minimise the potential effect of diet variation.

Results
During 215,149 person-years of follow-up, 175 new-onset T2D cases were confirmed. Participants in the highest baseline tertile (high consumption) of UPF consumption had a higher risk of T2D as compared to those in the lowest tertile (multivariable adjusted hazard ratio [HR] 1.53, 95% confidence interval [CI]: 1.06 to 2.22) with a significant dose–response relationship (p for linear trend = 0.024). The multivariable adjusted HR using repeated measurements of UPF intake was 1.65 (95% CI 1.14–2.38) when comparing extreme tertiles.

Conclusions
In a highly-educated Mediterranean cohort with a low absolute risk, a higher intake of UPF was independently associated with a higher risk for T2D. These results provide more evidence to encourage the limitation of UPF consumption to reduce the population burden of T2D.