Risk factor management and cognitive function in patients with atrial fibrillation

Abstract Background/Introduction Atrial fibrillation (AF) is associated with an increased risk of dementia, even in the absence of stroke. Risk factor management (RFM), a newly defined pillar of AF management, reduces AF burden and symptoms. However, data on brain health endpoints are lacking. Purpose This study investigated the impact of RFM and other pillars of AF management on cognitive impairment in AF. Methods 215 patients with a clinical diagnosis of AF and without a diagnosis of dementia were recruited (RECORD-AF, ANZCTR ACTRN12621001506886). Retrospective medical records audit, prospective assessments, and participant interviews were conducted. Cognitive assessment was conducted by blinded trained personnel using the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). Successful RFM was defined as weight loss to BMI<27 or ≥10% weight loss, treatment of OSA, alcohol intake <3SD/week and heart failure, hypertension and diabetes as per standard guidelines, and smoking cessation. RBANS index scores were regressed on demographic, comorbidities, RFM and AF-related variables in hierarchical linear regressions. Missing data was dealt with using multiple imputations with chained equations and a pooled analysis from 10 imputations. Results 196 participants completed RBANS for the cognitive function assessment. The mean age was 65.4 ± 10 years, 51% paroxysmal AF (PAF), 60% males and the median CHA2DS2VA score = 2 [1-3]. All received oral anticoagulation appropriate for CHA2DS2VA score. The mean RBANS index score was 94.4 ± 12.7. The visuospatial domain was most affected (RBANS 83.7 ± 13.5). Cognitive scores 1.5 SD below normative means were observed in 9.7% of the participants for the indexed RBANS and in 20.9% for immediate memory, 37.8% for visuospatial, 3.6% for language, 4.1% for attention and 15.3% for delayed memory domains. There was no association between risk factors at the time of AF diagnosis and cognitive function (β=-3, p=0.45). The number of unmanaged modifiable risk factors at RBANS was associated with poorer performance on immediate memory (β=-3, p=0.01), attention (β=-2.9, p=0.01), and total index score (β=-1.8, p=0.02) in analyses adjusted for age, sex, education, stroke, depression, CHA2DS2VA and AF severity. Compared to rate control, rhythm control was protective for the attention domain (β=7, p=0.01), and there was a trend toward improved total index RBANS scores (β=3.9, p=0.07). However, the benefit of rhythm control on cognition was no longer observed in the presence of un-modified risk factors (β=-3.08, p = 0.004 for rhythm*RF interaction term). Conclusion(s) Early impairment in cognitive function is common in individuals with AF. Successful risk factor management is associated with a lower risk of cognitive impairment. [LJP1]Correct?