More Evidence Links Atrial Fibrillation and Dementia Risk

Pauline Anderson

March 07, 2013

Atrial fibrillation (AF) is significantly associated with cognitive impairment and dementia, independent of history of clinical stroke, a new meta-analysis has found.

Previous research had highlighted the association between AF and cognitive decline, but it was assumed that this association was mediated through clinical stroke because patients with AF are 4 to 5 times more likely to develop clinical stroke, said lead author Shadi Kalantarian, MD, MPH, research fellow, Massachusetts General Hospital Cardiac Arrhythmia Service, Boston.

"We wanted to know if this association was independent of that stroke history, and we observed that it was," he told Medscape Medical News.

The study, the first to collect and present separate data for dementia and cognitive impairment, provides the most comprehensive evidence to date of the potential effects that AF has on cognition, but also highlights "critical gaps" in understanding the mechanism underlying the association, said Dr. Kalantarian.

The study is published in the March 5 issue of the Annals of Internal Medicine.

21 Studies

Researchers searched 5 large databases — MEDLINE, PsycINFO, Cochrane Library, CINAHL, and EMBASE — as well as reference lists, with no language restrictions, for studies that reported the association between AF and cognitive impairment or total dementia.

Dr. Shadi Kalantarian

The analysis included 21 studies representing geographically diverse regions, such as Asia, North and South America, Europe, and Australia. Fourteen studies reported the association between AF and cognitive impairment or dementia in patients with or without a history of stroke, and 7 studies examined the association of AF with cognitive impairment or dementia after stroke.

In a combined analysis of the 14 studies that investigated the association between AF and dementia or cognitive impairment, 9 of which were prospective, AF was significantly associated with the risk for cognitive impairment (relative risk [RR], 1.40; 95% confidence interval [CI], 1.19 - 1.64).

There was significant heterogeneity, however, possibly due to variance in participant characteristics. With use of a random-effects model, pooled estimates were almost the same for prospective and cross-sectional studies.

Heterogeneity was still seen in prospective studies, possibly because of different outcome measures. When the analysis was restricted to dementia outcomes, which are more accurately diagnosed than cognitive impairment, the heterogeneity was eliminated but the association didn't change substantially (RR, 1.38; 95% CI, 1.22 - 1.56).

Limiting the analysis to studies using the Mini-Mental State Examination, the most widely used screening tool for cognitive decline, also did not appreciably change the results.

In the 7 studies reporting cognitive impairment or dementia after a stroke, AF was associated with a more than 2-fold increased risk (RR, 2.70; 95% CI, 1.82 - 4.00).

To evaluate the association independent of stroke history, the investigators did a meta-analysis of studies that either excluded patients with a history of stroke or adjusted for stroke in the multivariate adjusted model. For this analysis, the RR was 1.34 (95% CI, 1.13 - 1.58).

Dr. Kalantarian emphasized that the association was independent of "clinically overt stroke," that is, strokes with which patients had been diagnosed, and that they had reported.

Possible Mechanisms

Shared risk factors, such as hypertension, congestive heart failure, and diabetes, which tend to accumulate as the population ages, might explain the association between AF and cognitive decline. Another explanation could be the hypercoagulable state in patients with AF or stasis of blood in the left atrium that may lead to formation of thrombi in the left atrial appendage and then to stroke.

Other unproven mechanisms, said Dr. Kalantarian, could involve cerebral hypoperfusion — that is, less blood going to the brain might affect cognition — or the proinflammatory state in both AF and dementia.

Perhaps the most plausible explanation is that AF increases the risk for silent infarcts, which would mediate the association between AF and cognitive impairment, said Dr. Kalantarian. However, 1 study in the analysis that excluded patents with a history of stroke with detailed imaging also showed an association between AF and cognitive impairment.

"That's only 1 study in the literature, and it says that there's a gap in our knowledge," said Dr. Kalantarian. "We think that silent infarct is a major contributor, but we can't say that it's 100% the underlying mechanism and that other things are not contributing. It's probably multifactorial; that other factors are contributing as well."

Priority should be given to studies that will elucidate the underlying mechanism, she added.

The authors used several objective criteria to assess the quality of individual studies. Of the 9 prospective studies in the broader population, 6 had "reasonably good quality" in that they matched 5 or 6 of the 7 quality criteria, said Dr. Kalantarian.

However, 6 of the 21 studies met 3 or fewer quality criteria, mainly because of a higher potential for misclassification of AF or outcome, inadequate adjustments for potential confounders, and the presence of attrition bias. Sensitivity analyses that excluded these studies did report similar results, though.

Another possible limitation was that stroke was mostly self-reported or derived from medical records and rarely confirmed by imaging.

The study results warrant further well-designed longitudinal studies with better adjustment for potential confounders and with detailed information on subtype of dementia, according to the authors. As well, clinical trials to evaluate interventions that may lower the risk for cognitive impairment in patients with AF are also needed, they said.

In future trials, "investigators should consider cognitive function as a new outcome to be assessed in interventional studies for the treatment of AF," the authors write.

Critical Question

Asked to comment on these findings for Medscape Medical News, Stefan Knecht, MD, professor, neurology, St. Mauritius Therapieklinik, Meerbusch, Germany, said he's not convinced that the meta-analysis adequately answers the critical question of whether cognitive impairment is linked to something other than stroke in patients with AF.

For one thing, he said, asking patients for their history of stroke is a weak tool because if they have had "silent" strokes, they wouldn't be aware of it, and so would not report it.

. Given the propensity of AF to cause strokes, it's not clear by what means other than some sort of stroke AF would affect brain function, he pointed out.

"There are actually several questions still to be answered," he said. "What kinds of strokes occur in AF to cause cognitive impairment? How much of a problem is this? How much of this is related to the type of treatment patients receive? And what can we do about it?"

The paper simply confirms that a relationship does exist between AF and cognitive impairment, but the authors do little to disentangle possible mechanisms, added Dr. Knecht.

"Could they have done so? I think so. For example, they could have tried to sort out differences in cognitive impairments in patient with versus patients without a history of stroke."

Dr. Kalantarian noted that she and her colleagues didn't have enough data to separate out different domains of cognition.

"The methodology of cognitive assessment is so variable and not many studies look at specific domains of cognitive function," she said. "That's why we were not able to say, for example, that patients with AF are more prone to have executive dysfunction."

Dr. Kalantarian and Dr. Knecht have disclosed no relevant financial relationships.

Ann Intern Med. 2013;158:338-346. Abstract