Effect of metformin on the risk of prostate cancer in patients with type 2 diabetes by considering different confounding factors: a meta-analysis of observational studies.

Preclinical studies have suggested the antitumorigenic properties of metformin on prostate cancer; results from epidemiological studies remain contradictory. We aim to investigate the evidence of metformin and the risk of prostate cancer. PubMed, Embase, Cochrane Library, and Web of Science databases were searched for eligible studies. Meta-analyses were carried out using the most fully adjusted hazard ratios and the corresponding 95% confidence intervals. Eighteen cohort studies and six case-control studies representing 2 009 504 male patients with type 2 diabetes mellitus were identified. The pooled HR of prostate cancer for metformin therapy was 0.97 (0.84-1.12) in case-control studies and 0.94 (0.79-1.12) in cohort studies, respectively. In cohort studies, we found that there was a modest association in studies with samples from Europe, but not in studies with samples from North America, Asia, and Oceania. In addition, metformin showed a slightly protective effect compared with sulfonylurea, but not insulin and other comparators. Meta-regression analyses found that obesity and prostate-specific antigen adjustment in statistical models may be the sources of heterogeneity. However, there were no significant differences in subgroups stratified by time-related biases, analytical approaches, types of risk estimates, study quality, publication year, and whether adjusted for smoking, alcohol abuse, hemoglobin A1c, diabetes duration, and other confounding factors. Our study showed that metformin therapy was not associated with the risk of prostate cancer in patients with type 2 diabetes mellitus. However, exploratory analyses suggest that metformin use may be protective in a certain subgroup of patients.