Hyper-insulinaemia and cancer, meta-analyses of epidemiological studies.

摘要

BACKGROUND: A substantial body of evidence links sex hormones, diet, excess body weight and physical activity to the risk of developing cancer at several sites common in affluent countries. The hypothesis that high circulating levels of insulin could be the underlying factor increasing cancer risk has been proposed. Epidemiological studies on markers of hyper-insulinaemia and cancer are reviewed and summarized. METHODS: Studies of cancers of the colon and rectum, pancreas, breast, and endometrium examining the association with blood levels of C-peptide, insulin, glucose, glycated haemoglobin (HbA1c) were searched in PubMed. Multivariate, adjusted relative risks (RR) and their 95% confidence intervals were abstracted and summarized by meta-analyses. RESULTS: Most of the studies identified were cohorts that relied on measurements obtained at baseline or assessed in blood stored at low temperature several years before the onset of cancer. The meta-analyses showed excess risks of colorectal and pancreatic cancers associated with higher levels of circulating C-peptide/insulin and with markers of glycaemia. Significant heterogeneity was found among four epidemiological studies of endometrial cancer and C-peptide giving a summary RR compatible with no association. Overall breast cancer risk was significantly higher in the upper categories of C-peptide/insulin, however, the excess derived entirely from retrospective studies. CONCLUSION: Current evidence suggests that subjects who develop colorectal and pancreatic cancers have increased pre-diagnostic blood levels of insulin and glucose.