Background
Comprehensive research evidence is lacking on the role of ultra-processed foods (UPF) in the relationship between the consumption of plant-sourced foods and their impact on cardiovascular disease (CVD) outcomes. This study aims to assess CVD risk associated with the dietary contribution of food groups that consider both plant or animal origin and food processing categories, within a large cohort of British adults.

Methods
Data from the UK Biobank participants (40–69 y) who completed at least two 24-h dietary recalls between 2009 and 2012 (n = 126,842; median follow-up: 9 y), with subsequent data linkage to hospital and mortality records, were used. Food groups were classified as either plant-sourced or non-plant/animal-sourced foods. These groups were further divided into non-UPF and UPF, and expressed as a percentage of total energy intake.

Findings
Every 10 percentage points increase in plant-sourced non-UPF consumption was associated with a 7% lower risk of CVD (95% CI 0.91–0.95) and a 13% lower risk of CVD mortality (0.80–0.94). Conversely, plant-sourced UPF consumption was associated with a 5% increased risk (1.03–1.07) and a 12% higher mortality (1.05–1.20). The contribution of all UPF was linked to higher CVD risk and mortality, and no evidence for an association between contribution of all plant-sourced foods and CVD incidence and mortality was observed.

Interpretation
The dietary contribution of plant-sourced non-UPF inversely linked to CVD risk, while plant-sourced UPF contribution showed a positive association. Recognizing the role of food processing is crucial for favourable CVD outcomes, even in plant-sourced diets.

Funding
World Cancer Research Fund.