Across 51 prospective cohorts encompassing nearly 9 million adults, higher ultra-processed food (UPF) consumption was associated with elevated hazard ratios for cardiovascular events (HR 1.24), depression/anxiety (HR 1.27), digestive disease (HR 1.31), metabolic syndrome/diabetes (HR 1.24), obesity (HR 1.23), cancer (HR 1.12), and all-cause mortality (HR 1.18). Crucially, dose-response analyses quantified the per-100g/day increment: each additional 100g of daily UPF intake raised cardiovascular risk by 14% and all-cause mortality by 3%.
The sheer scale here—8.8 million participants—makes this among the most statistically powered dietary meta-analyses ever published, and the consistency across seven distinct disease categories is striking. Yet size doesn't resolve the field's central limitation: all constituent studies are observational cohorts, meaning residual confounding by socioeconomic status, overall dietary quality, and health behaviours remains a genuine concern. UPF consumption correlates with poverty, shift work, and poor sleep—each independently harmful.
The NOVA classification used to define UPFs lumps together meaningfully different products (diet sodas, whole-grain breads, yogurts), and FFQ-based dietary assessment introduces measurement noise. Still, the dose-response linearity across multiple endpoints strengthens causal inference beyond a simple high-vs-low comparison. For practical guidance, even modest reductions—swapping 100g of processed snacks for minimally processed alternatives—appear to carry measurable mortality benefit. This meta-analysis is confirmatory but substantially upgrades the evidence base, moving UPF reduction from a plausible recommendation to one grounded in robust epidemiological signal.