Rizwan Kalani, MD

Chen LY, Lopez FL, Gottesman RF, Huxley RR, Agarwal SK, Loehr L, et al. Atrial Fibrillation and Cognitive Decline−The Role of Subclinical Cerebral Infarcts: The Atherosclerosis Risk in Communities Study. Stroke. 2014

Atrial Fibrillation (AF) is the most common cardiac arrhythmia, with increasing prevalence, incidence, and associated mortality worldwide. The increased risk of stroke is well known to all of us, but more recent studies have demonstrated an association of AF with cognitive impairment or dementia. Large epidemiologic series have shown that patients with AF have both an increased risk of dementia and faster cognitive decline, even without a clinical history of symptomatic stroke. We all probably have seen patients with a history of AF with clinically silent infarcts (SI) on neuroimaging; the authors of this study hypothesized that these lesions are associated with greater cognitive impairment in AF patients.

Chen et al evaluated data from the Atherosclerosis Risk in Communities (ARIC) biracial, multicenter, population-based study. The 935 patient cohort analyzed completed study visits that included three serial cognitive assessments (between 1993-1995, 1996-1998, and 2004-2006) and brain MRI at two points in time (between 1993-1995 and 2004-2006). AF diagnosis was obtained from ECG’s at study visits and hospital records, with cardiologist confirmation; patients were also clinically monitored for incident clinical stroke. Attention, executive function, and recent memory were assessed using the digit symbol substitution (DSS), word fluency (WF), and delayed word recall (DWR) tests, respectively. SI were defined as asymptomatic T2/PD hyperintense, T1 hypodense, focal, >3mm non-mass lesions on MRI.
The main novel findings were that patients who developed AF had significantly greater annual average rate of decline in DSS (-0.77) and WF (-0.80) in a linear model compared to those who did not develop AF, after adjusting for demographics/education/vascular risk factors. In subgroup analysis, patients who had more prevalent SI’s on baseline imaging who developed AF had greater decline in WF (-2.65) than those who didn’t develop AF during follow-up. In those that were found to have new SI’s on serial imaging during the study period, the ones that developed AF had greater decline in DSS (-1.51) than those who did not develop AF. Furthermore, in patients with incident AF, the proportion that developed SI’s was nearly twice that of those without new AF. In individuals without SI’s, incident AF was not associated with cognitive testing scores.
This study suggests that the association between incident AF and cognitive decline is mediated by SI’s. Most of these lesions turned out to be in the deep grey nuclei or deep supratentorial white matter; though not classic AF-related ischemia, up to 15% of patients with lacunar infarction have demonstrated an embolic source in prior reports. It is important to note that the role of shared vascular risk factors (HTN, DM, APO-e4 genotype, etc) contributing to cognitive decline cannot be excluded in this study.

The implications and questions raised by the results are important given the global epidemics of stroke, dementia, and AF. Future studies will need to evaluate risk of SI in those with low CHA2DS2-VASc scores and if anticoagulation reduces risk of cognitive decline in AF patients. Other questions raised are if we should be looking for occult AF in patients with SI’s or should we treat them with anticoagulants without documented AF?