Ultra-processed food consumption linked to several cancer types

The evidence suggested associations between UPF consumption and increased risks of overall cancer, colorectal cancer, breast cancer, and gastric cancer. However, the certainty of this evidence was weak: the GRADE framework rated 2 associations as low certainty and 8 as very low certainty. Methodological quality was also limited, with the AMSTAR 2 assessment finding that 3 reviews were of low quality and 4 of very low quality. Subgroup analyses indicated that age, geographic region, dietary assessment methods, and study design may explain some of the variation in observed associations across studies. The authors emphasize that while the pattern suggests a connection between UPF and cancer risk, the fragmented nature of existing research and inconsistent methods across studies limit confidence in the findings.

The heterogeneity across studies is substantial. Different researchers used different methods to define and measure ultra-processed food intake—some relied on food frequency questionnaires, others on 24-hour recalls—which may explain why some associations appeared strong while others were weak or absent. The association was strongest for colorectal and gastric cancers but weaker for breast cancer, suggesting cancer-specific pathways or that confounding factors (like overall diet quality or physical activity) may play a larger role for some malignancies than others. Three reviews failed AMSTAR 2 criteria due to incomplete searches, lack of protocol registration, or unaddressed publication bias. This umbrella review does not establish causation: most underlying studies were observational, meaning people who eat more ultra-processed foods may differ in unmeasured ways (exercise, smoking, socioeconomic status) from those who eat less. The authors call for prospective cohort studies using standardized UPF classification and adjusted analyses before public health directives change.