Metformin Prescription is Associated with Reduced AAA Growth Rate in a European Cohort

Introduction: Diabetes has been associated with reduced AAA incidence and growth rate in several large cohorts. Recent studies from USA, Australia and Taiwan have reported that the common antidiabetic drug metformin may be responsible for this protective effect. We sought to investigate this association in a European cohort. Methods: Patients under surveillance for small AAA were identified through local registers at three sites. Inclusion criteria were initial abdominal aortic diameter of ≥30mm and repeated ultrasonography scans of the abdominal aorta separated by at least six months. Data on medical history and prescribed medication at the time of the initial scan of the aorta were retrieved from local registers and individual patient records. One way-ANOVA and linear regression models were used to assess the association between metformin prescription and AAA growth rate. In the regression models known factors associated with AAA growth were included (gender, age, initial AAA diameter, active smoking and hypertension) along with kidney failure as it was more common in patients with diabetes mellitus type two (DM2) without metformin prescription. In a second model, cardiovascular and antidiabetic medications were added. Results: A total of 526 AAA patients were included in the analysis; 428 without DM2, 33 DM2 patients without metformin prescription, and 65 DM2 patients with metformin. Mean follow-up time was 3.2 (1.7) years. The mean (SD) initial AAA diameter was 38.0 (6.1) mm and annual growth rate was 2.11 (2.13) mm. Mean annual AAA growth rates were 2.31 (2.24) mm in patients without DM2 vs 1.14 (1.14) mm in DM2 patients with metformin prescription (p< 0.001) and 1.56 (1.40) mm in DM2 patients without metformin prescription (ns), figure 1. As growth rates were right skewed Kruskal-Wallis with post hoc tests were used for the ANOVA and growth rate data log transformed in the regression analysis. Cases with incomplete data were excluded in the regression models. With non-DM2 patients as reference group, the whole group of DM2 patients and those with metformin prescription, but not those without metformin prescription had a significantly reduced AAA growth rate in both the unadjusted and adjusted models. Including only DM2 patients in the regression analysis, metformin prescription was not significantly associated with AAA growth rate in neither model, table 1.Tabled 1ModelDiabetes vs non diabetes patientsDiabetes with metformin vs non diabetes patientsDiabetes without metformin vs non diabetes patientsMetformin treatment in diabetes patients onlypnpnpnpnUnadjusted<.001526<.001493.137461.12798Model 1.005501.005470.331438.25494Model 2.014500.009469.479437.34094[Association between AAA growth rate and diabetes with or without metfromin prescription] Open table in a new tab [Association between AAA growth rate and diabetes with or without metfromin prescription] Conclusion: These results confirms previous reports of a drastically reduced AAA growth rate in DM2 patients prescribed metformin compared to non-DM2 patients. The contrasting result of a lack of a statistically significant association of metformin prescription when analyzing only DM2 patients is likely due to insufficient power, only 33 DM2 patients without metformin prescription were included, and/or to confounding variables not captured in the models. As metformin is almost exclusively used in DM2 patients, a randomized controlled trial is needed to assess its effect on AAA growth in non-DM2 AAA patients. Disclosure: Nothing to disclose