RT Journal A1 Roussel R, Travert F, Pasquet B, et al T1 MEtformin use and mortality among patients with diabetes and atherothrombosis JF Archives of Internal Medicine JO Archives of Internal Medicine YR 2010 FD November 22 VO 170 IS 21 SP 1892 OP 1899 DO 10.1001/archinternmed.2010.409 UL http://dx.doi.org/10.1001/archinternmed.2010.409 AB Background  Metformin is recommended in type 2 diabetes mellitus because it reduced mortality among overweight participants in the United Kingdom Prospective Diabetes Study when used mainly as a means of primary prevention. However, metformin is often not considered in patients with cardiovascular conditions because of concerns about its safety.Methods  We assessed whether metformin use was associated with a difference in mortality among patients with atherothrombosis. The study sample comprised 19 691 patients having diabetes with established atherothrombosis participating in the Reduction of Atherothrombosis for Continued Health (REACH) Registry between December 1, 2003, and December 31, 2004, treated with or without metformin. Multivariable adjustment and propensity score were used to account for baseline differences. The main outcome measure was 2-year mortality.Results  The mortality rates were 6.3% (95% confidence interval [CI], 5.2%-7.4%) with metformin and 9.8% 8.4%-11.2%) without metformin; the adjusted hazard ratio (HR) was 0.76 (0.65-0.89; P < .001). Association with lower mortality was consistent among subgroups, noticeably in patients with a history of congestive heart failure (HR, 0.69; 95% CI, 0.54-0.90; P = .006), patients older than 65 years (0.77; 0.62-0.95; P = .02), and patients with an estimated creatinine clearance of 30 to 60 mL/min/1.73 m2 (0.64; 95% CI, 0.48-0.86; P = .003) (to convert creatinine clearance to mL/s/m2, multiply by 0.0167).Conclusions  Metformin use may decrease mortality among patients with diabetes when used as a means of secondary prevention, including subsets of patients in whom metformin use is not now recommended. Metformin use should be tested prospectively in this population to confirm its effect on survival.