Common blood thinners may slow cognitive decline in Alzheimer’s patients

People with both atrial fibrillation and Alzheimer’s disease who took a newer class of blood thinner declined more slowly on standard memory tests than those taking an older drug or no anticoagulant at all, according to a Swedish national registry study led by researchers at Karolinska Institutet. The drugs, known as non-vitamin K oral anticoagulants, or NOACs, are already prescribed widely to prevent stroke in people with irregular heartbeats. Now the same medications appear tied to a modest but measurable slowing of cognitive decline in patients who also have dementia.

Comparing NOACs, warfarin and no treatment at all

Maria Eriksdotter, a professor at Karolinska Institutet’s Department of Neurobiology, Care Sciences and Society and a senior consultant in geriatric medicine at Karolinska University Hospital, led the analysis using Sweden’s national dementia registry, SveDem. The study drew on 7,308 people diagnosed with both atrial fibrillation and Alzheimer’s disease, sorted into three groups: those taking NOACs, those taking the older blood thinner warfarin, sold in Sweden as Waran, and those receiving no anticoagulant medication at all. Cognitive function was tracked using the Mini-Mental State Examination, the standard bedside test doctors use to score memory, attention and orientation over time.

People taking NOACs showed a significantly slower annual decline than either comparison group, by slightly more than 0.2 points per year on that scale. “There are reasons to believe that the treatment could have a positive effect on cognition, for example by improving blood flow and reducing small-scale damage in the brain,” Eriksdotter said, describing a plausible mechanism separate from the drugs’ primary job of preventing stroke-causing clots.

A small yearly gap that Karolinska says compounds

Nanbo Zhu, a researcher on Eriksdotter’s team at Karolinska Institutet, cautioned against reading too much into a single year’s difference on its own. “The difference is modest for an individual patient from one year to the next, but over a longer period even such an effect could influence how cognitive function develops,” Zhu said. A 0.2-point annual gap sounds small against an MMSE scale that runs to 30 points, but Alzheimer’s disease typically erodes scores gradually over many years, meaning a consistent slowdown could add up to a meaningfully longer stretch of preserved function for some patients.

The NOAC group also showed lower risks of death, stroke, blood clots and fractures compared with people on no anticoagulant, benefits consistent with what cardiologists already expect from the drug class in atrial fibrillation generally. Warfarin users saw lower risks of death, stroke and blood clots too, but carried a higher risk of major bleeding than either of the other two groups, a tradeoff that has shaped anticoagulant prescribing for years independent of any cognitive question.

That bleeding risk is central to why researchers were curious about cognition in the first place. Warfarin requires regular blood monitoring and carries a well-documented risk of small, repeated bleeds in the brain over time, a possible mechanism for accelerating rather than slowing decline in people who already have dementia. NOACs, a group that includes drugs such as apixaban, rivaroxaban and dabigatran, need no routine monitoring and are generally considered to carry a lower bleeding risk, which is part of why they have gradually replaced warfarin as the default anticoagulant prescribed for atrial fibrillation over the past decade. Whether that safety profile, rather than any direct effect on brain blood flow, explains the slower decline is a question the registry data cannot fully separate out.

Why an observational study cannot prove cause and effect

The Karolinska team was careful to note the study’s limits. Because it draws on registry data rather than a randomized trial, it cannot establish that NOACs directly caused the slower decline rather than simply being prescribed more often to patients who were healthier, more mobile or under closer medical supervision to begin with. Patients in the registry could also have switched between drug types during the study period, and factors that influenced which anticoagulant a doctor chose for a given patient were not fully captured in the analysis.

Full details of the analysis appear in the European Heart Journal, where the Karolinska group lays out the registry methodology and statistical adjustments used to compare the three patient groups. A separate summary distributed to the scientific press through EurekAlert and picked up by outlets including Healthline echoed the same caution: the finding is an association worth testing further, not a reason for doctors to switch a patient’s anticoagulant on cognitive grounds alone.

Eriksdotter’s group has not announced a follow-up randomized trial, and until one exists, the registry link between NOACs and slower decline will remain what it is now: a large, carefully adjusted association rather than proof that the drugs protect the aging brain. For the doctors managing atrial fibrillation in dementia patients today, the practical takeaway is narrower than the topline result: the choice between NOACs and warfarin was already leaning toward the newer drugs on stroke and bleeding grounds alone, and this study gives that existing preference one more plausible, if unproven, reason behind it.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview