Original Research

Journal of General Internal Medicine

, Volume 25, Issue 8, pp 853-858

First online:

Diabetes Mellitus, Glycemic Control, and Risk of Atrial Fibrillation

  • Sascha DublinAffiliated withGroup Health Research InstituteDepartment of Epidemiology, University of Washington Email author 
  • , Nicole L. GlazerAffiliated withDepartment of Medicine, University of Washington
  • , Nicholas L. SmithAffiliated withDepartment of Epidemiology, University of WashingtonSeattle Epidemiology Research and Information Center, VA Puget Sound Health Care System
  • , Bruce M. PsatyAffiliated withGroup Health Research InstituteDepartment of Epidemiology, University of WashingtonDepartment of Medicine, University of WashingtonDepartment of Health Services, University of Washington
  • , Thomas LumleyAffiliated withDepartment of Biostatistics, University of Washington
  • , Kerri L. WigginsAffiliated withDepartment of Medicine, University of Washington
  • , Richard L. PageAffiliated withDepartment of Medicine, University of Wisconsin School of Medicine and Public Health
  • , Susan R. HeckbertAffiliated withGroup Health Research InstituteDepartment of Epidemiology, University of Washington

Rent the article at a discount

Rent now

* Final gross prices may vary according to local VAT.

Get Access



Diabetes may be an independent risk factor for atrial fibrillation. However, results from prior studies are in conflict, and no study has examined diabetes duration or glycemic control.


To examine the association of diabetes with risk of atrial fibrillation and to describe risk according to diabetes duration and glycemic control.


A population-based case-control study.


Within a large, integrated healthcare delivery system, we identified 1,410 people with newly-recognized atrial fibrillation from ICD-9 codes and validated cases by review of medical records. 2,203 controls without atrial fibrillation were selected from enrollment lists, stratified on age, sex, hypertension, and calendar year.


Information on atrial fibrillation, diabetes and other characteristics came from medical records. Diabetes was defined based on physician diagnoses recorded in the medical record, and pharmacologically treated diabetes was defined as receiving antihyperglycemic medications. Information about hemoglobin A1c levels came from computerized laboratory data.


Among people with atrial fibrillation, 252/1410 (17.9%) had pharmacologically treated diabetes compared to 311/2203 (14.1%) of controls. The adjusted OR for atrial fibrillation was 1.40 (95% CI 1.15-1.71) for people with treated diabetes compared to those without diabetes. Among those with treated diabetes, the risk of developing atrial fibrillation was 3% higher for each additional year of diabetes duration (95% CI 1-6%). Compared to people without diabetes, the adjusted OR for people with treated diabetes with average hemoglobin A1c ≤7 was 1.06 (95% CI 0.74-1.51); for A1c >7 but ≤8, 1.48 (1.09-2.01); for A1c >8 but ≤9, 1.46 (1.02-2.08); and for A1c >9, 1.96 (1.22–3.14).


Diabetes was associated with higher risk of developing atrial fibrillation, and risk was higher with longer duration of treated diabetes and worse glycemic control. Future research should identify and test approaches to reduce the risk of atrial fibrillation in people with diabetes.


arrhythmia atrial fibrillation diabetes mellitus glycemic control diabetes complications