Random exercise stress test in diagnosing effort angina

To improve the performance of exercise stress testing in the diagnosis of effort angina while minimizing risks of serious complications, we evaluated an impulse response of ST changes, which is a transient ST response resulting from a hypothetical, strenous-impulselike exercise, without actually imposing the strenuous load. To obtain the impulse response, subjects walked intermittently according to a computer-generated random binary sequence on a treadmill for 20 minutes (with a constant speed of 1.7 mph and a slope of 10%). We used Fourier transform for beat-to-beat changes in ST level and the binary sequence of exercise. We then determined the transfer function by taking the ratio of Fourier transformed ST level to exercise over the frequency range of 0.5 through 5.0 cycles/min. Converting the transfer function to the time domain yielded the impulse response of ST change. The subjects consisted of 49 patients (60 ± 9 years) with effort angina, 13 patients with atypical chest pain (56 ± 9 years), and 30 healthy, male volunteers (23 ± 7 years). In 82 subjects (89%), the ST impulse response showed an initial depression followed by a smooth, gradual restoration toward the preexercise ST level (type I response). The average duration of the initial depression was 8 ± 3 seconds in the healthy volunteers, whereas it was significantly prolonged to 23 ± 14 seconds in effort angina (p < 0.05). The depression in patients with atypical chest pain was not significantly different from that in the healthy volunteers. Although the level of exercise was milder in the proposed exercise test than in the conventional treadmill exercise test, the sensitivity and the specificity were significantly better in the proposed exercise test than in the conventional one in the same population. We conclude that this random exercise test is a sensitive, safe tool and is very accurate for the diagnosis of effort angina.