Background
Coronary artery disease (CAD) is the most common cardiovascular disease. It is a major cause of death and permanent disability and carries heavy economic and social costs due to its impaired functioning. It is estimated that up to 23.3 million people will die of cardiovascular disease by 2030 [
1]. CAD posed challenges to developed countries as well as to developing countries. With the aging of the global population, CAD has become a major public health problem that seriously threatens human life and health [
2,
3].
Metformin, as a biguanide derivative (dimethylbiguanide), has always been the primary drug in hypoglycemic treatment of patients with type 2 diabetes mellitus (T2DM) since its introduction in 1957 [
4]. Studies have found that metformin not only has a hypoglycemic effect, but also has a protective effect on various diseases such as kidney cancer [
5‐
7], pancreatic cancer, periodontal disease. Recent studies have also found that metformin has a positive effect on cardiovascular protection [
8‐
12]. Metformin also lowers risk factors for cardiovascular disease such as blood fats [
13‐
15], body weight and blood pressure. Compared with insulin and/or oral hypoglycemic agents (except metformin), metformin reduces the risk of all-cause mortality and the incidence of cardiovascular disease [
16,
17], infection, or acidosis. Metformin can reduce the incidence of myocardial infarction (MI) in newly diagnosed obese diabetic patients [
18]. Similarly, in animal models of myocardial infarction, metformin can effectively limit ischemia–reperfusion injury and reduce the infarct area, which is also shown in non-diabetic animals. Some placebo-controlled trials of metformin even support findings that incorporate cardiovascular endpoints [
19]. However, the effect of metformin on cardiovascular disease, especially coronary heart disease, remains controversial. Therefore, it is necessary to provide these data to equip patients with treatment guidelines and prescribing decisions.
Therefore, in this changing context, it seems timely to review the evidence of effects of metformin in preventing and improving cardiovascular disease. So, the goal of this systematic review and meta-analysis was to assess the effects of metformin on cardiovascular mortality, all-cause mortality, cardiac function and low density lipoprotein (LDL) levels in patients with CAD.
Methods
This study was reported in accordance with the PRISMA statement for reporting systematic reviews and meta-analysis [
20].
Search strategy
A systematic literature search of the PubMed, EMBASE, Cochrane Library and Web of Science databases was conducted by two study investigators independently. The cutoff date of the search was October 31, 2019. The following free text or MeSH terms were used in searching: “metformin” combined with “coronary heart disease” or “CHD” or “myocardial infarction” or “myocardial ischemia” or “cardiovascular disease” or “cardiovascular mortality” or “coronary artery disease” or “CAD” or “heart failure” or “HF” or “CHF”. The search was restricted to human studies. The titles and abstracts of studies identified in the search were independently reviewed by the two authors to exclude studies that were not meaningful to our research. References to identified studies were also retrieved to identify studies that may be eligible. The scope of inclusion was not limited by the language of publication.
Selection criteria
Eligible patients: patients with CAD, and patients with other age-related comorbidities were not excluded.
Eligible interventions: in our study, patients in the intervention group had been treated by metformin for a period of time.
Eligible controls: no medication or drugs other than metformin were used in the control groups.
Primary outcome: cardiovascular events (CV events), defined as recurrent MI, heart failure (HF), recurrent angina, malignant arrhythmia, cardiogenic death), cardiovascular mortality and cardiac function.
Two authors independently extracted data from selected studies, one author use a predefined data extraction sheet to extract data from each of the included studies, the second author independently reviewed the data to ensure accuracy.
Assessment of risk of bias
We used the Newcastle–Ottawa Scale (NOS) for assessing the quality of cohort studies and case–control studies based on three categories and eight items. The NOS uses a star rating system (semi-quantitative) where a maximum of nine stars can be awarded in assessing quality of studies.
Data extraction and analysis
The analysis was carried out from three perspectives and six indexes. Firstly, HR was calculated between effects of metformin and non-metformin on cardiovascular mortality, all-cause mortality and incidence of CV events. Secondly, three cardiac function indexes including LVEF, CK-MB and BNP were tested, each of which got the mean value or mean standard differences pooled in metformin and non-metformin groups. Thirdly, mean difference of LDL level was calculated between metformin and non-metformin groups.
Statistical methods
Review Manager 5.3 software from the Cochrane Collaboration (London, United Kingdom) was used to estimate the pooled effect size, the inverse variance approach was used to pool HR and SMD and draw the forest plot. P values for all comparisons were two-tailed and P < 0.5 of all tests was considered statistically significant except for heterogeneity. I2 values and Q statistic were used to evaluate heterogeneity across studies. Statistically significant heterogeneity was present at the P < 0.1. The random-effects model was used to calculate the pooled effect values where significant heterogeneity was present. Otherwise the fixed effects model was used.
Discussion
40 clinical trials were included in this study involving 1,066,408 subjects. We found: (1) metformin could remarkably reduce cardiovascular mortality; (2) metformin could significantly reduce all-cause mortality, including in patients with MI and HF; (3) metformin could reduce the incidence of cardiovascular events. Metformin could significantly reduce the incidence of cardiovascular events in HF patients, but wasn’t effective in MI patients. Metformin could significantly reduce the incidence of cardiovascular events in T2MD patients, but wasn’t effective in patients without T2MD. Furthermore, metformin was effective in reducing the incidence of cardiovascular events compared to those who take sulfonylureas or those who didn’t take anything; (4) metformin could reduce CK-MB level, but couldn’t improve LEVF and BNP; (5) metformin couldn’t lower LDL levels.
Metformin is the most widely used oral antihyperglycemic agent for the treatment of type 2 diabetes. Twenty years ago, experimental evidence showed that lowering blood glucose reduced the risk of microvascular complications among patients with type 2 diabetes [
33,
60,
61]. Several studies have subsequently indicated that metformin decreased mortality and adverse cardiovascular events [
62‐
66]. A large nationwide study revealed that patients treated with metformin had significantly lower all-cause mortality than those treated with sulfonylureas (including glimepiride, glyburide, glipizide, and tolbutamide) [
54]. A retrospective cohort study of 23,915 patients with type 2 diabetes mellitus reported that patients treated with glipizide [
67], glibenclamide or glimepiride had higher mortality than those treated with metformin after a median follow-up of 2.2 years. However, Bannister et al. [
68] found that the survival rate of diabetic patients using metformin was almost the same as that of normal healthy people. Inzucchi et al. [
69] also found that the mortality risk of patients using metformin at discharge was similar to that of patients not using metformin. Therefore, the cardiovascular protective effect of metformin was not clear. This meta-analysis indicates that metformin can reduce the cardiovascular mortality, all-cause mortality and incidence of cardiovascular events. In order to make the discussion more accurate, we conducted subgroup analyses. We find that metformin can reduce the incidence of cardiovascular events not only in HF patients, but also in T2MD patients, though, it isn’t effective in MI patients and patients without T2DM. In addition, metformin reduces the incidence of cardiovascular events better than sulfonylureas and better than non-medication. Metformin can significantly reduce mortality and reduce the incidence of cardiovascular events, that means metformin can be used as a cardiovascular protective agent to prolong life and reduce mortality. However, the cardiovascular protective mechanism of metformin has not been fully elucidated. T2DM is often associated with cardiovascular complications, and the two often coexist as comorbidity. Patients with T2DM often have elevated levels of inflammatory cytokines, while hyperglycemia and high concentration of glycation end products can also cause vascular endothelial cell damage and calcification [
70], which have important effects on the onset and progression of atherosclerosis [
61,
71], leading to related cardiovascular diseases. Metformin activates AMPK phosphorylation, which reduces oxidative stress, reduces the production of inflammatory cytokines, and increases eNOS (endothelial nitric oxide synthase) activity, which may be an important mechanism for metformin cardiovascular protection. Recent studies have also shown that metformin has protective effects on vascular endothelial function and angiogenesis [
72], which may be the pathway of metformin’s cardiovascular protective effects. However, this protective effect also has some limitations. For patients with severe diseases, such as MI, metformin has no effect, nor does it have cardiovascular protective effect for CAD patients without diabetes. Therefore, metformin cannot be used as first-line cardiovascular therapy. However, metformin can effectively reduce all-cause mortality, cardiovascular mortality and the incidence of cardiovascular events in patients with T2DM. Based on this cardiovascular protective effect, it is recommended that patients with T2DM should be given priority in the use of metformin, which can reduce the incidence of cardiovascular events while treating diabetes. Even if patients with T2DM have already had cardiovascular complications, it is recommended to use metformin in combination with cardiovascular therapy.
In diabetic patients, metformin and improved prognosis are independent of glycemic control, and there are indications that it may have direct cardioprotective effects [
37,
73]. Metformin has also been reported to improve cardiac cell function by altering cardiac metabolism and remodeling. Studies in mice found metformin reduced infarct size by 22 to 65%. Long-term preoperative use of metformin can reduce the risk of non-infarction in patients with acute myocardial infarction [
74]. A 2014 retrospective cohort study found that metformin also reduced the size of myocardial infarction [
37]. However, metformin’s role in improving cardiac function remains controversial. In one study, while patients with acute st-segment elevation patients (STEMI) were treated with metformin for 4 months, metformin was not found to increase ejection fraction (53.1% vs 54.8%,
P = 0.1) or decrease n-terminal pro-brain natriuretic peptide (nt-probnp) levels (167 vs. 167 ng/L,
P = 0.66) [
39]. Another STEMI study also found no improvement in ejection fraction in diabetic patients treated with metformin and a 0.7% difference between those taking metformin and those not taking metformin [
23]. After it was shown it could reduce mortality, metformin was further analyzed in this meta-analysis for its effect on cardiac function: LVEF, CK-MB and BNP were specifically analyzed. Our analysis revealed that the use of metformin is not associated with increased LVEF and decreased BNP, but with decreased mean value of CK-MB. In short, we found that metformin may have some effects on cardiac function. However, due to the small number of studies on CK-MB, the bias could be very large and there might be errors. More studies are needed to confirm the effect of metformin on CK-MB.
Hyperlipidemia is an important risk factor for coronary heart disease [
75]. For every 1.0 mmol/L increase in LDL, the risk of acute cardiovascular events increases by about 40%. Lowering blood lipids can effectively reduce the incidence of coronary heart disease [
1,
2]. Previous studies have shown that metformin can alter hepatic lipid homeostasis by inhibiting acetyl CoA carboxylase by AMPK, thereby enhancing insulin action. The level of total blood cholesterol and LDL can be reduced by taking metformin in the elderly [
76]. The plasma total cholesterol (TG) is remarkably reduced by using high dose metformin (> 1700 mg/day). Lexis et al. [
38] found that LDL levels in non-diabetic acute STEMI patients significantly decreased after 4 months of administration of metformin. These studies show that metformin also lowers blood lipid levels. For further assessment of the effect of metformin on LDL levels nine studies were included in this meta-analysis, though six studies showed that metformin reduced LDL levels, our meta-analysis found that metformin was not associated with reduced LDL levels. Luo et al. [
62] also concluded that metformin does not reduce LDL concentration, but it plays a cardiovascular protective role by increasing cholesterol outflow [
17]. So, more research is needed on whether metformin may be protective by affecting LDL.
As a hypoglycemic drug, metformin has been used clinically for up to 60 years thanks to its low cost, safety and effectiveness. In our meta-analysis, it was found that besides the antihyperglycaemic effect, metformin can also reduce all-cause mortality, mortality of cardiovascular disease and incidence of CV events. Besides the antihypoglycemic effect, clinicians should consider metformin’s cardiovascular protection when selecting drugs. Metformin is more secure because it has a very low risk of hypoglycemia compared to other drugs such as sulfonylureas, and very little risk of causing lactic acidosis. We have more sufficient reasons to recommend metformin.
Some limitations should be considered when interpreting our findings. First of all, most of the patients included in this study were diabetic, so the results do not stand for the majority of non-diabetic patients with coronary heart disease treated with metformin. Secondly, there is large heterogeneity between some studies related to population, research design and other aspects. Thirdly, this meta-analysis doesn’t reflect other language studies as literature included is written in English.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.