Background Greater awareness of the importance of stroke prevention in AF, greater screening, and the availability of warfarin for primary prevention over the past 2 decades should have reduced the incidence of AF-related ischaemic stroke despite the ageing population.

Methods We studied all AF-related incident ischaemic strokes (IIS) from 2002 to 2008 in the Oxford Vascular Study (OXVASC) and compared our findings with results from a previous population-based study in the same population in 1981–1986 (Oxford Community Stroke Project; OCSP).

Results AF-related stroke accounted for 29.4% (210/715) of IIS in OXVASC (mean/SD age=79.8/9.9) vs 17.9% (85/474) in OCSP (77.3/9.0): relative rate =1.64 (1.32–2.04, p<0.0001). However, after adjustment for ageing of the underlying population between the studies the rates were similar (1.19, 0.96–1.48, p=0.109). Of the 154 OXVASC patients with incident ischaemic stroke and prior AF, 116 had a CHADS2 score of ≥2, of whom only 19 (16.4%) had been on warfarin for primary prevention.

Conclusions There has been no reduction in age-adjusted incidence of AF-related ischaemic stroke in Oxfordshire over the last 25 years and absolute numbers have increased substantially. Underuse of anticoagulation in primary prevention in eligible patients is a major barrier to effective stroke prevention in the face of an ageing population with high rates of AF.