Meta-Analysis of 51 Studies Associates Highest Ultra-Processed Food Intake With 12-31% Elevated Risks Across Multiple Chronic Conditions

The Family Medicine and Community Health meta-analysis pooled dose-response data from cohorts with 2- to 32-year follow-up that applied the NOVA classification to dietary records. Highest consumers showed graded increases in risk for cardiometabolic, neoplastic, gastrointestinal, and mental-health outcomes after adjustment for measured confounders. Absolute risks were not reported, leaving base-rate context limited.

Prior large observational syntheses in The Lancet and BMJ have reported comparable effect sizes for cardiovascular and cancer endpoints when ultra-processed food intake is expressed as a percentage of total energy. Residual confounding by socioeconomic status, physical activity, and overall diet quality remains plausible, and reverse causation cannot be excluded in studies with shorter follow-up.

The current evidence base consists entirely of observational designs; no large randomized trials have tested substitution of ultra-processed foods with minimally processed alternatives on hard clinical endpoints. Regulatory bodies have therefore not yet altered labeling or procurement policies on the basis of these associations.

Next required studies are long-term pragmatic trials measuring substitution effects on incident disease and mortality, plus repeated dietary assessment with biomarker validation to reduce measurement error.