Public Release: 

Study clarifies impact of age on safety of warfarin treatment for atrial fibrillation

Appropriate anticoagulation target seen as most important for avoiding hemorrhage

Massachusetts General Hospital

A study conducted at Massachusetts General Hospital (MGH) has clarified the risk of intracranial hemorrhage in older patients with atrial fibrillation who take the drug warfarin to prevent ischemic stroke. The report in the November 16 Annals of Internal Medicine showed that treatment at moderate levels of intensity has no greater risk of hemorrhage than does lower intensity treatment. Although the risk of hemorrhage did significantly increase in patients over 85, that risk often can be minimized by tight control of warfarin intensity.

"Studying how warfarin therapy is associated with intracranial hemorrhages is challenging because these events are rare, albeit devastating," says Margaret Fang, MD, MPH, who led the study as an MGH research fellow and is now at the University of California at San Francisco (UCSF). "Because lower level warfarin therapy can lead to a much higher risk of ischemic stroke, we also wanted to examine whether less intensive treatment, which recent guidelines have suggested for older patients, actually results in a lower risk of hemorrhage."

Atrial fibrillation, a type of irregular heartbeat, is a common and strong risk factor for stroke. By leading to the formation of blood clots that travel to the brain, the condition is believed to account for about 80,000 ischemic strokes a year and can increase a patient's overall stroke risk fivefold. Many patients with atrial fibrillation are treated with blood-thinning medications like warfarin, and previous studies by members of this research team and others have confirmed that achieving appropriate levels of anticoagulation - reflected by levels of 2.0 or more on a blood test called the INR - can significantly reduce the risk that a stroke will occur.

However, elevated anticoagulation levels can increase the risk of brain hemorrhage, a rare but dangerous complication of warfarin therapy. Concern about the risk of hemorrhage, especially among older people, has led some patients to avoid anticoagulation therapy. The current study was designed to determine at what age and INR level the risk for intracranial hemorrhage becomes significant.

The research team compiled information on 170 adult atrial fibrillation patients who had developed intracranial hemorrhage while being treated with warfarin and compared that data to information from 744 randomly selected patients from the MGH anticoagulation clinic who had not developed hemorrhage during the same time period. The study groups included only patients taking warfarin for atrial fibrillation.

Although the risk of hemorrhage did increase with patients' age, the most significant increase in risk occurred after age 85. The risk of bleeding increased in patients with INR levels over 3.5 and sharply rose after 4.0. But even among older patients, achieving an of INR of less than 2.0 did not reduce risk below that seen at an anticoagulation level of 2.0 to 3.0.

"Our study emphasizes that physicians should aim for an INR range of 2.0 to 3.0 when prescribing warfarin for atrial fibrillation, even for older patients," says Fang. "Although we found that people aged 85 and older had a higher risk for intracranial hemorrhage, these are also the patients that gain the greatest benefit from warfarin for stroke protection. Therefore, the net benefits of warfarin generally outweigh the potential risks in older patients with atrial fibrillation." Fang is an assistant adjunct professor of Medicine in the UCSF Division of General Internal Medicine Hospitalist Group.

The study authors are also participating in a continuing study of more than 13,000 atrial fibrillation patients to assess the long-term risk for hemorrhage both with and without warfarin therapy. That study is being led by Daniel Singer, MD, of MGH, senior author of the current study, and Alan Go, MD, of UCSF and Kaiser Permanente of Northern California, also a co-author of the current study. Additional co-authors are Yuchiao Chang, PhD, Elaine Hylek, MD, MPH, Jonathan Rosand, MD, and Steven Greenberg, MD, PhD, all of MGH. This study was supported by a National Research Service Award, the National Institute on Aging, and the Eliot B. Shoolman Fund of MGH.


Contact for UCSF: Carol Hyman, 415 502-9553,

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.