Association of Metformin Use With Risk of Venous Thromboembolism in Adults With Type 2 Diabetes

A General-Population-Based Cohort Study

Tingting Sha; Yuqing Zhang; Changjun Li; Guanghua Lei; Jing Wu; Xiaoxiao Li; Zidan Yang; Chao Zeng; Jie Wei


Am J Epidemiol. 2022;191(5):856-866. 

In This Article

Abstract and Introduction


Metformin is hypothesized to protect against the risk of venous thromboembolism (VTE); however, there is a paucity of data supporting this hypothesis. Among individuals aged 40–90 years with a diagnosis of type 2 diabetes in the Health Improvement Network database (2000–2019), we compared the risks of incident VTE, pulmonary embolism, and deep vein thrombosis among metformin initiators with those among sulfonylurea initiators. Individuals were followed from their first prescription refill to an incident VTE, drug discontinuation, switching or augmenting, plan disenrollment, or the end of the study, whichever occurred first. Hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated using the Cox model, adjusting for confounders using inverse probability of treatment weighting. Among 117,472 initiators of metformin and 13,835 initiators of sulfonylureas, 555 (1.3/1,000 person-years) and 75 (2.1/1,000 person-years) VTE cases occurred in each group, respectively. The multivariable-adjusted HR was 0.65 (95% CI: 0.51, 0.84). The corresponding risks for pulmonary embolism (adjusted HR = 0.71, 95% CI: 0.50, 1.01) and deep vein thrombosis (adjusted HR = 0.64, 95% CI: 0.48, 0.87) were also lower in metformin initiators than in sulfonylurea initiators. Our study provided empirical evidence to support a lower risk of VTE after initiation of metformin as compared with sulfonylureas among patients with type 2 diabetes.


Venous thromboembolism (VTE), including pulmonary embolism (PE) and deep vein thrombosis (DVT), is associated with substantial morbidity and mortality.[1] Studies have shown that 8.3% of the population is expected to develop VTE after 45 years of age[2] and approximately 30% of patients with PE die within 1 year after PE diagnosis,[3,4] making VTE the third most common cause of vascular mortality after acute myocardial infarction and stroke.[1] Furthermore, although the overall incidence of VTE has remained stable over time in the United States and Europe (e.g., 1–2 cases per 1,000 inhabitants), the proportion of PE, a more severe subtype, among VTE appears to be increasing.[5,6]

Type 2 diabetes mellitus is a common metabolic disorder, affecting more than 450 million adults worldwide,[7] and its incidence and prevalence are still increasing.[8] Patients with type 2 diabetes often present early vascular endothelial dysfunction and imbalance between coagulation and fibrinolysis—for example, increased activity of platelets and resistance to fibrinolysis[9,10]—and thus are at increased risk of venous thrombotic complications (odds ratio = 1.42, 95% confidence interval (CI): 1.12, 1.77).[11–13]

Metformin is one of the most commonly prescribed medications for the control of diabetes. Both in vitro and in vivo studies have shown that, in addition to its role in glycemic control, metformin could decrease the activation, adhesiveness, and aggregation of platelets,[14,15] modulate the release of procoagulant factors,[16,17] and improve the integrity and senescence of endothelial cells.[18,19] Therefore, one may speculate that metformin plays a beneficial role in preventing VTE through its anticoagulation and endothelial protection properties. However, to date there is a paucity of data supporting this hypothesis. In a cohort study, Lu et al.[20] reported that metformin was associated with a lower occurrence of DVT compared with nonuse. However, that study did not adjust for obesity, a strong risk factor for VTE,[1] and used nonusers as a comparison group, which may lead to intractable confounding by indication bias or channeling bias.[21]

Thus, our objective in this population-based cohort study was to assess whether initiation of metformin use is associated with lower risks of incident VTE, PE, and DVT than initiation of sulfonylureas (an active comparator) in individuals with type 2 diabetes.