Reduced risk of lung cancer with metformin therapy in diabetic patients: a systematic review and meta-analysis.

Zhi-Jiang Zhang,Yongyi Bi,Shiyue Li,Qingjun Zhang,Genming Zhao,Yi Guo,Qibin Song

DOI: https://doi.org/10.1093/aje/kwu124

2014-01-01

American Journal of Epidemiology

Abstract:Emerging evidence from epidemiologic studies and basic science suggests an inverse association between metformin use and cancer risk in diabetic patients. However, the association with lung cancer is not consistent. We summarized the evidence currently available (2009-2013) and explored sources of heterogeneity. Metformin therapy was associated with significantly lower risks of cancers of the lung (4 studies; pooled relative risk = 0.71, 95% confidence interval (CI): 0.55, 0.95; P = 0.02) and respiratory system (6 studies; pooled relative risk = 0.85, 95% CI: 0.75, 0.96; P = 0.01). There was evidence of moderate heterogeneity (I-2 > 50%). The major sources of heterogeneity were smoking adjustment status and cancer site. The relative risk from studies that adjusted for smoking was 1.16-fold (95% CI: 1.00, 1.35) closer to the null than that from studies not adjusting for smoking. The relative risk of respiratory cancer was 1.23-fold (95% CI: 1.02, 1.49) closer to the null than that for lung cancer. In conclusion, metformin use appears to be associated with lower risks of lung and respiratory cancer in diabetic patients. However, caution regarding overestimation is needed, since adjustment for smoking attenuates the association.