A Swedish study of people who take anticoagulants suggests that these drugs are associated with lower rates of dementia. While some limitations of the study were noted, this HealthDay story neglected to mention important details such as the costs of treatment, the magnitude of the treatment effect, and the potential for serious harms such as uncontrolled bleeding. Nor did the story acknowledge the researchers’ ties to numerous drug companies that make anticoagulants.
Many of the shortcomings of this story can be traced back to the news release that apparently sparked it. When reporters don’t dig beyond the information they are provided in a news release, there’s a good chance that they’ll present an unbalanced and incomplete portrait of the research to their readers.
There is an apparent association between atrial fibrillation and the risk for dementia. If an anticoagulant prescribed for atrial fibrillation could also prove to be protective against dementia, that would indeed be a medical advance. Yet despite the study in which more than 26,000 of 440,000 participants with atrial fibrillation were diagnosed with dementia, it remains unclear if the oral anticoagulants are solely responsible for the reduction in dementia rates. It’s possible, for example, that patients on oral anticoagulation therapy are different in important ways from those who are not — and that these differences (not the drug treatment) help explain the reduction in dementia rates. The story would have been wise to acknowledge this likelihood.
No mention of costs, which is unfortunate because there can be substantial costs associated with anticoagulation, especially with the newer anticoagulants.
The story reported only the relative benefits, that “people taking anticoagulants were 29 percent less likely to develop dementia than were those who were not taking the blood thinners.” We ask: “29% of what?”
It would have been much more informative to provide the absolute rates of dementia in both groups. We explain why in this primer: Reporting the findings: Absolute vs relative risk.
People stop taking these drugs for a reason, and some of those reasons might have to do with the side effects or adverse effects which can also include unstoppable bleeding. The story should have mentioned this. As the article indicates, “patients start on oral anticoagulation for stroke prevention but they stop after a few years at an alarmingly high rate.” It would seem an important question to explore: why do approximately 15 percent of patients stop taking the drugs within the first year, and then 10 percent each year afterwards?
This was a mixed bag. The researchers noted the drawbacks of such a cohort study, pointing to the fact that it cannot provide definitive proof linking anticoagulant use and reduced risk of dementia. They noted that a randomized placebo-controlled trial would need to be done, yet doing those studies might be considered unethical.
While this is all helpful context, the story uses inappropriate language throughout the story, including in the headline and lead sentence, to suggest a cause-and-effect relationship between anticoagulants and dementia. For example: “New research suggests they help prevent dementia as well as stroke.” This statement goes beyond what an observational study like this one can tell us.
We’ll give the benefit of the doubt on the rating, but note that these findings may simply be a manifestation of the “healthy user effect” rather than a true benefit of these drugs. A stronger story would have alerted readers to this possibility.
There is no obvious disease mongering here.
We learn that the researchers came from the Karolinska Institute in Stockholm yet are not given information about any potential financial conflicts between those researchers or their institution to the makers of anticoagulants.
As the study itself makes clear, the researchers have financial ties to numerous drug companies that make anticoagulants including Bayer, Pfizer, and Bristol-Myers Squibb.
We learn that “along with not taking blood thinners, the strongest predictors for dementia were age, Parkinson’s disease and alcohol abuse.” While this hints at some of the other factors that might influence dementia risk, it doesn’t really get at strategies one might consider to actively stave off risk. The story could have mentioned some of these strategies including exercise and a healthful diet.
It may seem fairly obvious that these anticoagulants are widely prescribed. The fact that the study was among community-dwelling people who were already taking these treatments reinforces their availability.
The story makes it sound like the anticoagulant-dementia risk connection is a brand new discovery. However, the study that’s the basis of the story notes that numerous researchers have already examined this hypothesis and come to conflicting conclusions. The story should have indicated what’s new and different about this study compared to the previous ones, if anything.
Several quotes are taken directly from this European Heart Journal news release. While the story acknowledges the source of the quotes, it doesn’t seek out any outside voices or perspectives to complement the coverage. The predictable result is a relatively weak and unbalanced story.