Predictive value of clinical and exercise variables for detection of coronary artery disease in men with diabetes mellitus.

Sixty-eight men with diabetes mellitus (mean age 53 +/- 10 years) and no symptoms of cardiac dysfunction enrolled in a long-range study for detection of latent coronary artery disease. The testing included maximal treadmill stress with thallium-201 scintigraphy and echocardiography. Radionuclide angiography was available in 35 men (52%), and 24 (35%) had gated scanning with exercise. Of the 68 patients, 14 (21%) had a mild (9 patients) or moderate (5 patients) decrease in ejection fraction on radionuclide angiography, echocardiography or both. Fifty-two men agreed to remain in the study and have been followed for 12 to 18 months (mean 41 +/- 19). Ten coronary events have occurred. Four of the men died (2 suddenly) and 6 have angina pectoris. Three patients have had vascular complications. Of the clinical and exercise variables studied, exercise duration effectively predicted an adverse outcome, while the odds ratio in favor of a coronary event increased by 36 times in those with thallium-201 defects and 7 times in those with ST-segment changes on exercise. Radionuclide angiographic responses during exercise were abnormal in 5 of 6 patients with events, but were also abnormal in 12 of 29 men (41%) who did not have coronary artery disease. Clinical variables such as blood pressure, cholesterol level and family history were not predictive of outcome, nor was maximal heart rate during exercise. Thus, diabetic mean who can exercise for 440 seconds on a treadmill using a bruce protocol are at low risk of a coronary event.(ABSTRACT TRUNCATED AT 250 WORDS)