Ultra-processed food intake and risk of digestive system cancers: a systematic review and dose-response meta-analysis

Ultra-processed food (UPF) intake has been associated with digestive system cancer risk, but evidence remains scattered across cancer sites, study designs, and exposure metrics. This systematic review and meta-analysis evaluated associations between UPF intake and digestive system cancer risk overall and by cancer site. We further examined UPF subcategories and dose-response relationships.

We systematically searched PubMed, Embase, Web of Science, and the Cochrane Library for English-language cohort and case-control studies published from database inception to May 1, 2026. Hazard ratios (HRs) from cohort studies and odds ratios (ORs) from case-control studies were pooled separately with their corresponding 95% confidence intervals (CIs) using random-effects models. Additional analyses were conducted to investigate site-specific digestive system cancers, UPF subcategories, and dose-response relationships, alongside subgroup, sensitivity, meta-regression, and publication bias assessments. The study protocol was registered in PROSPERO (CRD420261404299).

A total of nine prospective cohort studies and five case-control studies were included. The main pooled estimates showed a 12% higher risk in cohort studies (HR = 1.12, 95% CI: 1.05-1.20; I ² = 37.6%) and 42% higher odds in case-control studies (OR = 1.42, 95% CI: 1.25-1.60; I ² = 26.7%). Site-specific analyses showed relatively consistent positive associations for colorectal and colon cancer outcomes. Among UPF subcategories, ultra-processed meat/protein products were associated with higher digestive system cancer risk (HR = 1.33, 95% CI: 1.15-1.53). In dose-response analyses using UPF weight share as the primary exposure metric, each 10-percentage-point increase in UPF intake was associated with a 6% higher risk (HR = 1.06, 95% CI: 1.01-1.10), with evidence of non-linearity, although no reliable threshold could be identified. Supplementary analyses based on g/day showed a consistent direction of association. Sensitivity analyses supported the stability of the main findings, and no clear evidence of publication bias was observed.

Higher UPF intake was associated with increased digestive system cancer risk, with relatively consistent evidence for colorectal and colon cancers. Ultra-processed meat/protein products may be a relevant contributing subcategory, and dose-response analyses suggested an exposure-response gradient. Because evidence came mainly from observational studies and dose-response data were limited, findings require confirmation in high-quality prospective studies.

https://www.crd.york.ac.uk/PROSPERO/view/CRD420261404299, identifier PROSPERO (CRD420261404299).