Adults age 80 and older with atrial fibrillation who continued taking blood-thinning medication had lower one-year mortality than those who never received the treatment, according to new research from WashU Medicine. The study is led by Gizem Kaya, MD, a hospitalist in the Division of General Internal Medicine.
The findings, published Oct. 27 in the Journal of General Internal Medicine, suggest that age alone should not be a reason to withhold anticoagulation, even as clinicians carefully consider the risk of bleeding.
Atrial fibrillation, an irregular heart rhythm, increases the risk of blood clots and ischemic stroke. Anticoagulants, commonly called blood thinners, help prevent clots from forming. However, concerns about bleeding and other health conditions can make physicians and patients hesitant to use these medications in older adults, who have been underrepresented in clinical trials.
“This study looks at a question we face all the time in practice: whether to continue, start, or withhold anticoagulation in very elderly patients with atrial fibrillation,” says Kaya.
To better understand the benefits and risks in this population, researchers analyzed health record data from 45,305 adults age 80 and older with atrial fibrillation. The study compared outcomes among patients who continued anticoagulation, started treatment, stopped treatment or never received it.
Among patients who continued taking anticoagulants, 8.7% died within one year, compared with 17.7% of those who never received the medication. After accounting for differences between patients and other factors associated with mortality, anticoagulation was associated with a 25% lower risk of death during the study period.
The researchers also found important differences in bleeding outcomes. Gastrointestinal bleeding occurred more frequently among patients who continued anticoagulation than among those who never received it — 13% compared with 6.6%. Rates of bleeding within the brain, a particularly serious complication, were similar across the treatment groups.
The survival differences narrowed over time. By five years, mortality rates were similar across the groups, suggesting that other health conditions and age-related risks may increasingly influence long-term outcomes.
The findings underscore the complexity of treatment decisions for older adults. Although anticoagulation was associated with lower mortality in this study, the retrospective design cannot establish that the medication directly caused the difference. Treatment decisions should account for each patient’s overall health, frailty, bleeding risk and goals of care rather than relying on age alone.
The research team used a synthetic electronic health record database derived from BJC HealthCare records to study outcomes in a large, real-world population.