Metformin and Varicose Vein Risk

First study to examine metformin and varicose vein risk looks at whether diabetes drug may have anti-inflammatory properties that affect varicose vein risk.

A study that was done in Finland revealed that varicose veins were associated with certain risk factors such as age, body height, body mass index, standing at work, high socioeconomic status, and a number of births. A positive association was established between obesity and varicose veins. However, an association between diabetes and varicose veins was not established. The fact that before 1995 only sulfonylureas and metformin were available to treat diabetes can suggest that metformin might be a protective factor against varicose veins in patients with diabetes. Metformin is believed to have anti-inflammatory properties that help with endothelial dysfunction and inflammation. There are no studies done on metformin and varicose veins. This study wanted to explore the relationship between the two. 

This study, which was done in Taiwan, included both an unmatched cohort and a well-matched cohort (derived from the unmatched cohort) of patients with type 2 diabetes. International classification of disease (ICD) codes were used to obtain the diagnoses of diabetes and varicose veins. They excluded patients who were treated with other antidiabetic medications before metformin, patients with type 1 diabetes, missing data, varicose veins at entry or within 6 months of diabetes diagnosis, cancer within entry or 6 months of diabetes diagnosis, central nervous system disorders, and patients who were followed for less than 180 days. The unmatched cohort had 123,710 metformin ever users and 15,095 metformin never users. Final randomization in the matched cohort included a 1:1 ratio (15,088 in each group) of metformin ever users and never users. The occupation was also classified in the study. Class 1 included civil servants, teachers, employees of governmental or private business, professionals, and technicians; class 2 included people without a specific employer, self-employed people; class 3 included farmers or fishermen; class 4 included low-income families supported by social welfare, or veterans. Rubin’s B and Rubin’s R were used in the matched cohort to assess for the balance between covariates between metformin ever users and never users after matching. Cumulative duration of metformin therapy and incidence density of varicose veins were calculated.  Kaplan Meier curves were used to compare never users vs ever users and for never users vs tertiles of cumulative duration of metformin therapy in patients who were varicose vein-free. Log-rank test was used to test the significance in different groups of metformin exposure. Hazard ratios were used forever, the user. Cox proportional hazard model with the inverse probability of treatment weighting (IPYW) was used using the propensity score to reduce potential confounding between the difference in characteristics. Sensitivity analyses were also done. 

The two groups were balanced with Rubin’s B of less than 25 and Rubin’s R of between 0.5 and 2 in the matched cohort. Regarding varicose vein-free population, Kaplan Meier curves revealed statistical significance (P<0.01) in the comparison of never users vs ever users. Regarding comparing the duration of metformin treatment vs never users, the log-rank test revealed statistical significance in favor of metformin therapy (P<0.01). All analyses revealed a lower risk of varicose veins with metformin. The hazard ratio that compared ever users vs. never users was 0.62 (95% CI, 0.5-0.75). There was a dose-response benefit when metformin was used for more than 2 years in the unmatched cohort (HR, 0.55; 95% CI, 0.44 -0.69) and 5 years in the matched cohort (HR 0.24; 95% CI 0.13, 0.42).

This was the first study to assess the relationship of metformin use and the reduction in risk of getting varicose veins. Some limitations of the study were multiple confounders such as lifestyle, physical activity, daily standing time, number of births, cigarette smoking, alcohol drinking, dietary pattern, family history, and genetic parameters. Another limitation was the sole use of ICD codes which can lead to misclassification and underestimation and a lack of data concerning medication adherence. A measure of how controlled diabetes (HbA1c) was, was missing, which can lead to potential confounding. A strength of this study was that they used medical records and not self-reporting. More research will need to be done to establish a causation relationship rather than associative.  

Practice Pearls:

• Metformin was associated with less risk of getting varicose veins. Duration of treatment showed significance starting with 2 years in 1 cohort and 5 years in the other.
• A proposed mechanism for metformin would be the anti-inflammatory properties that can modulate endothelial dysfunction and inflammation.
• This was the first study to ever associate metformin with reduced risk of varicose veins.

Reference for “Metformin and Varicose Vein Risk”:
Tseng, Chin‐Hsiao. “Metformin Reduces Risk of Varicose Veins in Patients with Type 2 Diabetes.” Diabetes/Metabolism Research and Reviews, 2019

Nour Salhab, Pharm.D. Candidate, USF College of Pharmacy