P4625Association between pattern of atrial fibrillation and neurocognitive function in patients at low risk of stroke

Abstract

Growing evidence suggests that the rate of cognitive impairment and dementia is magnified in patients with AF, independently of clinical stroke. The strongest association is observed in patients <75 years. However, it remains unclear whether type of AF influences cognitive function. Method: Patients with AF and a low risk of stroke randomized in the BRAIN‐AF trial (NCT02387229) underwent global cognitive assessment at baseline using the MoCA score (range 0‐30). A cut‐off value <26 has a sensitivity of 80‐100% and specificity of 50‐76% for detecting mild cognitive impairment. A cut‐off value <24 is less sensitive for this purpose (84%) but has greater specificity (81%). Inclusion criteria were: documented AF; CHADS2 score of 0, and age <62 years. Main exclusion criteria were: known dementia; major depression; need for anticoagulant or antiplatelet therapy and increased risk of bleeding. Results: A total of 219 patients, mean age 52.9±7.3 years, 20% female were enrolled and had permanent (N=30, 13.7%), persistent (N=28, 12.8%), or paroxysmal (N=161, 73.2%) AF (Table). All subjects had depression and global cognitive function assessments. Multivariate analyses adjusted for age and education level. A higher proportion of patients with permanent and persistent AF had a MoCA score <26 (23.3% and 21.4%, respectively) compared to patients with paroxysmal AF (8.7%; p=0.0461). Similar results were obtained with a MoCA cutoff value<24 (13.3% versus 7.1% versus 1.9%, respectively, p=0.0115). There was a trend towards a lower mean MoCA score in patients with permanent versus paroxysmal AF (p=0.0515). Conclusions: In conclusion, in a young AF population at a very low risk of stroke risk (CHADS2 score of 0), cognitive function is modulated by pattern of AF, with a gradient of risk for cognitive impairment observed from paroxysmal to persistent to permanent AF. (Table Presented).