Iron feeds tumors (and also kills them)

Mechanism (strong, but cuts both ways)

The classic rationale is that iron is a pro-oxidant: free iron catalyzes Fenton chemistry, generating reactive oxygen species that cause oxidative DNA damage. Iron has been suggested as a risk factor for several cancers mainly due to its prooxidant activity, which can lead to oxidative DNA damage, and people with hemochromatosis or iron overload have a higher risk of developing liver cancer. Iron is also required for proliferation (ribonucleotide reductase, DNA synthesis), and many tumors display an iron-overload phenotype to fuel rapid growth.

The complication is ferroptosis. Ferroptosis is an iron-dependent form of cell death characterized by iron overload and lipid peroxidation, and it acts as a double-edged sword in cancer. Iron is critical for proliferation and many tumor cells exhibit iron overload to support rapid growth, yet iron simultaneously acts as a double-edged sword for those same cells — high intracellular iron makes some tumors (especially metastasis-prone, drug-resistant, mesenchymal-state cells) vulnerable to ferroptotic death. So at the cellular level iron both promotes carcinogenesis and provides an exploitable kill-switch, which is why iron-loading nanoparticles and GPX4/xCT inhibitors are being explored therapeutically.

Observational epidemiology (moderate, heterogeneous, site-specific)

The largest systematic review (Fonseca-Nunes 2014, 59 studies) found that higher intake of heme iron showed a tendency toward a positive association with cancer risk, with per-1-mg/day heme iron relative risks of 1.08 for colorectal, 1.12 for colon, 1.03 for breast, and 1.12 for lung cancer. Notably, the picture flips depending on the exposure: three of four studies using serum ferritin as an indicator of high body iron stores showed an inverse relation, so dietary heme iron and measured body stores don't point the same way.

By Cancer site:

Colorectal is the strongest signal — roughly three-quarters of 33 studies of colorectal neoplasia supported an association between increased iron intake or stores and increased risk.
Liver (HCC): a meta-analysis found high serum ferritin and high serum iron both associated with primary liver cancer (HR 1.49 for ferritin; HR 2.47 for serum iron). This is the cleanest link, consistent with the hemochromatosis story.
Breast: modest. Heme iron intake was significantly associated with increased risk (pooled RR 1.12), and serum/plasma iron showed RR 1.22, but ferritin, transferrin saturation, and total/dietary/supplemental iron showed no significant association.
Lung: genuinely mixed — a meta-analysis of 13 studies found no association between serum iron and lung cancer risk, despite some positive heme-iron findings in men.

Causal inference / Mendelian randomization (mixed — undercuts a uniform effect)

MR is worth weighting heavily here because it sidesteps reverse causation and confounding. The results argue against iron being a uniform causal driver. For lung cancer specifically, MR analysis demonstrated that genetically-predicted iron status did not causally increase lung cancer risk across serum iron, ferritin, transferrin saturation, or transferrin, and one study even found higher genetically-predicted serum iron status inversely associated with lung squamous cell carcinoma. Other MR work does find site-specific positive signals — for thyroid cancer, increases in iron, serum ferritin, and transferrin saturation were associated with an elevated risk, and elevated ferritin was positively correlated with small cell lung cancer in one 2025 analysis. The takeaway: causal effects appear cancer-specific rather than global.

Randomized trial evidence (limited — essentially one trial)

The only RCT-level interventional data come from the VA FeAST trial (Zacharski). VA Cooperative Study #410 randomized 1,277 peripheral arterial disease patients without recent malignancy to control (n=641) or iron reduction by phlebotomy (n=636). In the reduction group, 38 of 636 patients developed malignancies and 14 died of cancer, a statistically significant difference versus control. A 2025 reanalysis reaffirmed and quantified this: iron reduction via calibrated phlebotomy targeting a ferritin of 25 ng/mL significantly decreased cancer risk overall and reduced cancer-specific and all-cause mortality among those diagnosed with cancer, with effects first becoming significant within 6 months of the initial phlebotomies.

Note: cancer was a secondary/substudy outcome (the primary vascular endpoint was null), the population was a Veterans Affairs PAD cohort, mean age 67, 98.5% self-reported male, and it's a single trial — so it's suggestive of causality and direction but not definitive, and not generalizable to women or younger populations.

Summary: lausible and bidirectional

The honest synthesis: mechanistically plausible and bidirectional (ferroptosis), epidemiologically supported mainly for heme iron intake and for liver/colorectal cancer, weaker and inconsistent elsewhere, with MR suggesting the causal effect is site-specific rather than universal, and a single RCT providing the only interventional support. The cleanest gap to flag is the dissociation between dietary heme iron (positive) and measured body-iron biomarkers (mixed/inverse in several datasets), which suggests heme iron's effect may operate partly through gut-luminal mechanisms rather than systemic iron load alone.

Forbidden Supplements in Cancer #2 Iron — The Nutrient Cancer Craves

mandakingnd.substack - June 2026 Iron is also involved in:

Oxygen transport
Mitochondrial energy production
DNA synthesis
Immune function
Neurotransmitter production
Cellular respiration

Some cancers may be particularly responsive to iron availability. * Breast cancer * Colorectal cancer * Pancreatic cancer * Lung cancer * Liver cancer * Leukemias * Lymphomas

Oncology becomes a constant exercise in balancing competing priorities. * Sometimes iron supplementation is necessary. * Sometimes intravenous iron is used. * Sometimes erythropoiesis-stimulating agents are considered. * Sometimes blood transfusions become necessary. * None of these decisions are simple, and none should be made in isolation.