It is unclear whether prediagnostic iron-related biomarkers in circulation are associated with cancer risk. We constructed a case-cohort of participants who had plasma samples available from the Japan Public Health Center—based Prospective Study and determined the incidence of cancer in these participants. We measured plasma concentrations of iron, ferritin, and hepcidin, and assessed the association between each biomarker and cancer incidence using a weighted Cox regression model. There were 4,253 participants in the sub-cohort (the randomly selected participants from an eligible, at-risk population) and 3,596 incident cancer cases (499 cases occurred in the sub-cohort). Median follow-up was for 16.5 years. In the multivariable adjusted analysis, iron deficiency (plasma ferritin <30 ng/mL) was associated with a higher risk of total cancer [adjusted HR, 1.23; 95% confidence interval (CI), 1.07–1.42] and the association was weaker after excluding those followed-up for <3 years. Iron overload was not significantly associated with total cancer (HR, 1.04; 95% CI, 0.82–1.33), but was associated with liver cancer (HR, 4.49; 95% CI, 2.71–7.43). Lower plasma levels of hepcidin and ferritin are associated with an increased gastrointestinal cancer risk. Meanwhile, lower plasma hepcidin and higher plasma ferritin levels were associated with an increased liver cancer risk. In conclusion, there was no association between iron overload and cancer risk, besides liver cancer.
High ferritin and low hepcidin levels in the plasma were associated with increased liver cancer risk. Evaluating iron metabolism including hepcidin levels may help identify people with high liver cancer risk.