In order to study the usefulness of ambulatory ECG recording in detecting asymptomatic coronary
artery disease associated with diabetes mellitus, 136 diabetic subjects (type 1, n = 72; type 2, n = 64) were studied by
24-hour ECG monitoring. Positive test results on exercise electrocardiography and/or thallium scintigraphy with
angiographically documented coronary artery disease served as a standard for significant coronary artery disease.
Transient ST segment depressions were detected in 6 patients on 24-hour ECG recording. Seventeen diabetic subjects
had ischemia on exercise electrocardiography and 31 on dynamic thallium scintigraphy. Twelve patients with
evidence of myocardial ischemia in any of the noninvasive tests had significant (≥50% coronary artery narrowing)
coronary artery disease, and insignificant (<50% coronary artery narrowing) coronary artery wall abnormalities
were detected in 7 diabetic subjects; 6 refused coronary angiography. In the 6 subjects with a positive 24-hour ECG
monitoring result, coronary angiography showed significant coronary artery lesions in 3 patients, insignificant coronary
artery disease in 1 subject, and 2 had patent coronary arteries. Thus the sensitivity of the 24-hour monitoring
was only 25%, the specificity 98% and the predictive accuracy 50%. These results show that ambulatory ECG
recording is an insensitive method for detecting asymptomatic coronary artery disease in otherwise apparently
healthy diabetic subjects.