Effect of metformin on all-cause and cardiovascular mortality in patients with coronary artery diseases: a systematic review and an updated meta-analysis

Abstract

Background: Metformin is the most widely prescribed drug to lower glucose and has a definitive effect on the cardiovascular system. The goal of this systematic review and meta-analysis is to assess the effects of metformin on mortality and cardiac function among patients with coronary artery disease (CAD).

Methods: Relevant studies reported before October 2018 was retrieved from databases including PubMed, EMBASE, Cochrane Library and Web of Science. Hazard ratio (HR) was calculated to evaluate the all-cause mortality, cardiovascular mortality and incidence of cardiovascular events (CV events), to figure out the level of left ventricular ejection fraction (LVEF), creatine kinase MB (CK-MB), type B natriuretic peptide (BNP) and to compare the average level of low density lipoprotein (LDL).

Results: In this meta-analysis were included 40 studies comprising 1,066,408 patients. The cardiovascular mortality, all-cause mortality and incidence of CV events were lowered to adjusted HR (aHR) = 0.81, aHR = 0.67 and aHR = 0. 83 respectively after the patients with CAD were given metformin. Subgroup analysis showed that metformin reduced all-cause mortality in myocardial infarction (MI) (aHR = 0.79) and heart failure (HF) patients (aHR = 0.84), the incidence of CV events in HF (aHR = 0.83) and type II diabetes mellitus (T2DM) patients (aHR = 0.83), but had no significant effect on MI (aHR = 0.87) and non-T2DM patients (aHR = 0.92). Metformin is superior to sulphonylurea (aHR = 0.81) in effects on lowering the incidence of CV events and in effects on patients who don't use medication. The CK-MB level in the metformin group was lower than that in the control group standard mean difference (SMD) = - 0.11). There was no significant evidence that metformin altered LVEF (MD = 2.91), BNP (MD = - 0.02) and LDL (MD = - 0.08).

Conclusion: Metformin reduces cardiovascular mortality, all-cause mortality and CV events in CAD patients. For MI patients and CAD patients without T2DM, metformin has no significant effect of reducing the incidence of CV events. Metformin has a better effect of reducing the incidence of CV events than sulfonylureas.

Keywords: Coronary artery disease; Diabetes; Meta-analysis; Systemic review.

Fig. 1

Flowchart of study selection

Fig. 2

Forest plot of hazard ratio of cardiovascular mortality among patients with metformin therapy vs no-metformin therapy

Fig. 3

a Forest plot of hazard ratio of all-cause mortality among patients with metformin therapy vs no-metformin therapy. b Forest plot of hazard ratio of all-cause mortality among patients with MI at baseline, metformin therapy vs no-metformin therapy. c Forest plot of hazard ratio of all-cause mortality among patients with HF at baseline, metformin therapy vs no-metformin therapy

Fig. 4

a Forest plot of hazard ratio of CV events among patients with metformin therapy vs no-metformin therapy. b Forest plot of hazard ratio of CV events among patients with MI at baseline, metformin therapy vs no-metformin therapy. c Forest plot of hazard ratio of CV events among patients with HF at baseline, metformin therapy vs no-metformin therapy

Fig. 5

a Forest plot of hazard ratio of CV events VENTS among patients with T2DM at baseline, metformin therapy vs no-metformin therapy. b Forest plot of hazard ratio of CV events among patients without T2DM at baseline, metformin therapy vs no-metformin therapy. c Forest plot of hazard ratio of CV events among patients with metformin monotherapy vs sulphonylurea monotherapy. d Forest plot of hazard ratio of CV events among patients with metformin monotherapy vs no-drug therapy

Fig. 6

a Forest plot of mean difference of LVEF% among patients with metformin therapy vs no-metformin therapy. b Forest plot of mean difference of CK-MB among patients with metformin therapy vs no-metformin therapy. c Forest plot of mean difference of BNP among patients with metformin therapy vs no-metformin therapy. d Forest plot of mean difference of LDL among patients with metformin therapy vs no-metformin therapy

Fig. 7

a Funnel plot of hazard ratio of cardiovascular mortality among patients with metformin therapy vs no-metformin therapy. b Funnel plot of hazard ratio of all-cause mortality among patients metformin therapy vs no-metformin therapy. c Funnel plot of hazard ratio of CV events among patients with metformin therapy vs no-metformin therapy. d Funnel plot of hazard ratio of CV events among patients with T2DM at baseline, metformin therapy vs no-metformin therapy