Plant-Based Diets, Ultra-Processed Foods, and Risks of Mortality and Major Chronic Diseases: A Prospective Cohort Study

Published August 2026

Background

Higher-quality plant-based diets (PBDs) are associated with lower risks of mortality and chronic disease, but whether ultra-processed food (UPF) content affects these associations remains unclear. We examined whether UPF content influences the relationship between plant-based dietary patterns and risks of mortality and major chronic diseases, accounting for nutrient quality.

Methods

This prospective cohort study included 124,836 UK Biobank participants aged 40–70 years (recruited 2006–2010). Dietary intake was assessed using the Oxford WebQ 24-h recall. Four modified Plant-Based Diet Indices (PDIs) were derived to distinguish healthy (hPDI) and unhealthy (uPDI) patterns with high- and low-UPF content, using the Nova classification and a Modified Nutrient Quality Index (mNQI). Participants were followed for 8.3–10.5 years for all-cause mortality and incident T2DM, CVD, and cancer. Multivariable Cox models estimated hazard ratios (HRs) and 95% confidence intervals (CIs).

Findings

Among 124,836 participants (mean [SD] age 56.2 [7.8] years; 55.8% women), there were 5780 deaths, 3420 T2DM cases, 6078 CVD cases, and 9437 cancer cases. Higher adherence to healthy plant-based diets—whether high- or low-UPF—was associated with 8–28% lower risk of all-cause mortality [HRQ4vsQ1 (95% CI): high-UPF hPDI, 0.92 (0.85–1.00); low-UPF hPDI, 0.91 (0.84–0.98)] and type 2 diabetes [high-UPF hPDI, 0.89 (0.79–0.99); low-UPF hPDI, 0.72 (0.65–0.79)]. Higher adherence to the high-UPF hPDI was also associated with 11% lower cardiovascular disease risk [0.89 (0.82–0.96)], while no clear association was observed for the low-UPF hPDI. Nutrient quality was similar across high- and low-UPF hPDI patterns.

Interpretation

Adherence to healthful PBDs is associated with more favourable health outcomes irrespective of UPF content, suggesting that overall PBD quality may be more important than processing level for chronic disease prevention.