Preventing Dementia May Be Another Reason to Take Oral Anticoagulants for A-fib

On top of staving off stroke, oral anticoagulation may serve another critical purpose in patients with atrial fibrillation: preventing dementia.

In a large, Swedish registry study, patients taking oral anticoagulation had a relative 29% lower risk of dementia compared with their untreated peers (HR 0.71; 95% CI 0.68-0.74), according to Leif Friberg, MD, PhD, and Mårten Rosenqvist, MD, PhD (Danderyd University Hospital, Stockholm, Sweden).

The magnitude of the apparent benefit was greater when treatment was initiated earlier, they report in a study published online October 24, 2017, ahead of print in the European Heart Journal.

“If there is a high probability that this relationship we have shown is actually true, then it constitutes a very strong argument for both doctors and patients to make sure that you stay on treatment,” Friberg told TCTMD, noting that prior studies have demonstrated that 10% to 15% of patients with A-fib discontinue oral anticoagulation annually. “A knowledgeable patient who knows why he or she is taking a drug is much better at using it and uses it in a much safer way. And that is, I think, the most important part of this study: that it should improve compliance and make people stay on treatment.”

Similar Outcomes With NOACs, Warfarin

It’s well-known that patients with A-fib carry a heightened risk of dementia, but the potential for oral anticoagulation to modify that risk remains unclear. A recent systematic review encompassing 19 studies and 15,876 patients found no definitive evidence supporting either benefit or harm of anticoagulation in terms of cognitive function.

A randomized, placebo-controlled trial to evaluate the impact of oral anticoagulation on dementia would not be ethically possible at this point—given the established benefit for stroke prevention—so researchers are left with observational data to explore the issue, Friberg said.

In this study, he and Rosenqvist examined Swedish national registry data on 444,106 patients who were diagnosed with A-fib between 2006 and 2014 and who had no prior diagnosis of dementia. At baseline, 54.3% were not taking an oral anticoagulant, 42.9% were taking warfarin, and 2.9% were taking a non-vitamin K antagonist oral anticoagulant (NOAC).

Overall, 5.9% of patients developed dementia during follow-up. The rate of dementia was lower in anticoagulated patients (1.14 vs 1.78 per 100 patient-years at risk; P < 0.001).

After propensity-score matching (leaving 80,948 matched pairs), patients taking oral anticoagulation carried a lower risk of dementia compared with those not taking the drugs, with an even stronger association in an on-treatment analysis (HR 0.52; 95% CI 0.50-0.55).

Friberg and Rosenqvist used four falsification endpoints to test whether confounding by indication could be influencing the strong association between treatment and dementia risk. They found that use of oral anticoagulation was either not related or only weakly related to risks of influenza, fall accidents, diabetes, or chronic obstructive pulmonary disease. “This, we think, is an important indicator that [the dementia] findings are not just incidental. They are true,” Friberg said.

The investigators further assessed whether NOACs and warfarin were differentially associated with dementia risk, and found no difference (HR 0.97; 95% CI 0.67-1.40).

Theoretically, Friberg explained, NOACs might be expected to provide greater protection against dementia because they’ve been shown to carry a lower risk of intracerebral bleeding, which could hasten cognitive decline, compared with warfarin. It could be that the excellent warfarin management achieved in Sweden hampered the possibility of finding a difference between the two types of agents and that an advantage for NOACs could emerge in parts of the world with poorer warfarin control, he said.

He noted that there is an ongoing trial comparing dabigatran (Pradaxa; Boehringer Ingelheim) and warfarin in patients with nonvalvular A-fib that has incident dementia as the primary endpoint. The projected completion date is April 2021.