BackgroundExcess iron has been shown to induce diabetes in animal models. However, the results from human epidemiologic studies linking body iron stores and iron intake to the risk of type 2 diabetes mellitus (T2DM) are conflicting. In this study, we aimed to systematically evaluate the available evidence for associations between iron intake, body iron stores, and the risk of T2DM.MethodsA systematic search of the PubMed/MEDLINE and EMBASE databases to the end of 22 April 2012 was performed, and reference lists of retrieved articles were screened. Two reviewers independently evaluated the eligibility of inclusion and extracted the data. Pooled relative risks (RRs) and 95% confidence intervals (CIs) were calculated using random-effects models.ResultsWe reviewed 449 potentially relevant articles, and 11 prospective studies were included in the analysis. A meta-analysis of five studies gave a pooled RR for T2DM of 1.33 (95% CI 1.19 to 1.48; P<0.001) in individuals with the highest level of heme iron intake, compared with those with the lowest level. The pooled RR for T2DM for a daily increment of 1 mg of heme iron intake was 1.16 (1.09 to 1.23, P<0.001). Body iron stores, as measured by ferritin, soluble transferrin receptor (sTfR) and the sTfR:ferritin ratio, were significantly associated with the risk of T2DM. The pooled RRs for T2DM in individuals with the highest versus the lowest intake of ferritin levels was 1.70 (1.27-2.27, P<0.001) before adjustment for inflammatory markers and 1.63 (1.03-2.56, P = 0.036) after adjustment. We did not find any significant association of dietary intakes of total iron, non-heme, or supplemental iron intake with T2DM risk.ConclusionHigher heme iron intake and increased body iron stores were significantly associated with a greater risk of T2DM. Dietary total iron, non-heme iron, or supplemental iron intakes were not significantly associated with T2DM risk.