Cerebrovascular Reactivity in Atrial Fibrillation

Atrial fibrillation is associated with substantial risk of severe stroke, cognitive decline and dementia. Elevated plasma biomarkers of vascular damage/dysfunction and reduced conduit (brachial) artery function have been reported in atrial fibrillation. We sought to test the hypothesis that patients with atrial fibrillation have a reduced cerebrovascular reactivity to carbon dioxide (CO 2 ) in comparison with age‐matched, healthy control and hypertensive participants. Middle cerebral artery flow velocity was measured at rest and during step‐wise changes in partial pressure of end‐tidal CO 2 (P ET CO 2 ) in 14 patients with atrial fibrillation (median [IQR]; 66.5 [59.5–69.3] yr, 3 women), 8 patients with hypertension (65.5 [63.5–71.5] yr, 4 women) and 10 healthy control participants (67.5 [66–70] yr, 4 women). Incremental hypercapnia was delivered via the open circuit steady‐state method, with stages of 4% and 7% CO 2 . Incremental hypocapnia was comprised of two 2‐min stages of voluntary hyperventilation to reduce P ET CO 2 in an equal but opposite direction to that observed during hypercapnia. Cerebrovascular reactivity was calculated as the change in middle cerebral artery mean flow velocity, per mmHg change in P ET CO 2 . Resting middle cerebral artery mean flow velocity tended to be lower in patients with atrial fibrillation (mean [SD]; 54.4 [14.5] cm/s) and hypertension (54.6 [13.0] cm/s), when compared to healthy controls (65.9 [14.4] cm/s) (P=0.127). Hypercapnic cerebrovascular reactivity was ≈50% lower in patients with atrial fibrillation (median [IQR]; 1.7 [1.5–2.4] cm/s/mmHg) and ≈22% lower in patients with hypertension (2.7 [2.2–3.1] cm/s/mmHg), compared to healthy controls (3.4 [2.8–3.8] cm/s/mmHg) (P<0.05). Hypocapnic cerebrovascular reactivity was not significantly different between groups (P>0.05). In summary, atrial fibrillation patients exhibit blunted hypercapnic cerebrovascular reactivity when compared to healthy control participants. Such impairments in cerebrovascular reactivity may contribute to the well‐established increased risk of cerebrovascular events in this patient group. Support or Funding Information Funded by BHF project grant PG/15/45/31579.