Use of Coumarin Anticoagulants and Cerebral Microbleeds in the General Population

Study Questions:

Does the use of warfarin relate to a higher prevalence and incidence of cerebral microbleeds? Are cerebral microbleeds more frequent in patients with a higher maximal international normalized ratio (INR) value? Does variability in INR for warfarin users correlate with cerebral microbleed presence?


From the population-based Rotterdam Study, 4,945 participants ages ≥45 years with brain magnetic resonance images (MRIs), but no prior infarct, were analyzed, including 3,069 who had follow-up MRIs. Warfarin use and INR values were analyzed using logistic regression in comparison to the presence or absence of cerebral microbleeds on brain MRIs. INR variability was calculated based on the first 10 values following warfarin initiation.


Overall, 8.6% of the study population used warfarin before the initial MRI. The prevalence of microbleeds in the initial MRI was 19.4%. Including the initial and follow-up MRI results, the incidence rate of microbleed development was 6.9% per year (mean follow-up 3.9 years). Warfarin use was associated with a higher prevalence of cerebral microbleeds (odds ratio, 1.70; 95% confidence interval, 1.24-2.34). There was a trend toward higher microbleed prevalence in patients with a higher maximal INR value compared to warfarin never users (p = 0.073). Patients with higher INR variability had a higher prevalence of cerebral microbleeds compared to those with the least INR variability (p = 0.048).


The authors concluded that use of warfarin is associated with cerebral microbleeds, and that this association is strongest in patients with greater INR variability.


This study describes the association between warfarin use and the presence of cerebral microbleeds. In particular, they associate high INR variability with cerebral microbleeds. Microbleeds have been associated with cognitive impairment (i.e., amyloid angiopathy) and risk of ischemic stroke. It is not clear if use of the newer direct oral anticoagulants (e.g., dabigatran, rivaroxaban, and apixaban) would have different associations with cerebral microbleeds. However, phase III trials have consistently shown a marked reduction in the risk of intracranial hemorrhage with use of the direct oral anticoagulants compared to warfarin for stroke prevention in atrial fibrillation.

Clinical Topics: Anticoagulation Management, Arrhythmias and Clinical EP, Heart Failure and Cardiomyopathies, Anticoagulation Management and Atrial Fibrillation, Atrial Fibrillation/Supraventricular Arrhythmias

Keywords: Stroke, Intracranial Hemorrhages, Follow-Up Studies, Cerebral Amyloid Angiopathy, Anticoagulants, Warfarin, Atrial Fibrillation, Confidence Intervals, Magnetic Resonance Spectroscopy, Brain

< Back to Listings