Author's reply

EDITOR,—Nigel Wheeldon discusses the association between atrial fibrillation and coronary disease, but much depends on the criteria used for diagnosis. Many epidemiological studies have depended on electrocardiographic criteria or the clinical history. Studies depending solely on ST-T wave abnormalities may encounter difficulties in interpretation, especially if patients are taking digoxin or have left ventricular hypertrophy.

The frequency of coronary disease or other aetiological causes of atrial fibrillation depends on the population studied. In surveys of patients admitted to hospital acutely with atrial fibrillation, ischaemic heart disease …