Iron intake and cancer risk: an umbrella review with causal methods audit

Background: The potential carcinogenic role of iron—particularly heme iron—has been proposed based on its pro-oxidant and nitrosating capacity, but systematic reviews show heterogeneous results. The objective of this study is to synthesize the evidence from systematic reviews on iron consumption/exposure and cancer risk, assessing methodological quality, certainty of evidence, and use of causal inference frameworks. Methods: Systematic reviews of observational studies on dietary iron (total/heme/non-heme) or biomarkers vs . cancer incidence were included. Searches were conducted in PubMed, Scopus, Embase, and Web of Science, without language restriction. A MeaSurement Tool to Assess systematic Reviews (AMSTAR-2), Risk Of Bias In Systematic reviews (ROBIS), and Grading of Recommendations Assessment, Development and Evaluation (GRADE) were applied. Results: Nine reviews met the inclusion criteria. For colorectal cancer, three cohort meta-analyses found a positive association between heme iron and risk (per 1 mg/day: relative risk [RR]: 1.08–1.18), with low to moderate heterogeneity and extensive adjustment; GRADE: low. In esophageal cancer, total iron was associated with lower risk and heme iron with higher risk in dose–response analyses; GRADE: very low due to inconsistency and imprecision. In lung and breast cancer, syntheses showed null associations for total and heme iron; a subgroup of women suggested a modest increase with heme iron in lung cancer; GRADE: low. Biomarkers yielded mixed results. Methodological quality was low to moderate; no use of directed acyclic graphs (DAGs), G-methods, or causally justified confounder sets was identified. Conclusions: The evidence shows a consistent and modest association between heme iron and colorectal cancer (RR: 1.08–1.18 per additional mg/d), with a dose–response gradient and low heterogeneity, although certainty is low due to residual confounding and measurement error. No consistent evidence of association was found between total dietary iron or non-heme iron and cancer risk. For esophageal cancer, findings are paradoxical and of very low certainty; for lung and breast cancer, null effects predominated. The universal absence of formal causal frameworks in the included reviews limits causal inference.