To investigate the mechanism of sympathovagal imbalance due to myocardial ischemia, we studied 42 consecutive patients undergoing successful percutaneous transluminal coronary angioplasty by correlating frequency domain and time domain measures of heart rate variability with parameters such as echocardiography, stress thallium scanning and radionuclide angiography before, immediately after and 2 months after the procedure. Of these, 20 patients (Group N) had normal and 22 patients (Group A) had abnormal regional wall motion. A control group of 20 healthy subjects (Group C) underwent echocardiography and examination of heart rate variability twice at 2-month intervals to check for spontaneous variations. At baseline, frequency domain measures such as low and high frequency power and time domain measures such as SDANN index (the mean of the standard deviations of the average of RR intervals) were lower in Group A than in Groups N and C, whereas no differences were detectable in ultra low and very low frequency, total power, SDNN index (the mean of the standard deviations of the mean of normal RR intervals), and r-MSSD (the root mean square of successive RR differences). There was high association between the diastolic wall stress index and both high frequency (r = -0.82) and low frequency power (r = -0.77). There were similar findings for the systolic wall stress index (r = -0.72 for high frequency and r = -0.64 for low frequency power). After successful coronary angioplasty, regional wall motion, left ventricular wall stress indices and all measures of heart rate variability were unchanged in Group N. In Group A the mean summed segment score improved from 15.9 +/- 2.6 to 12.2 +/- 1.7 (p < 0.0001), and mean low frequency, mean high frequency power (logarithmic units), and SDANN index (msec) increased from 6.10 +/- 0.23 to 6.36 +/- 0.28 (p < 0.005), from 5.36 +/- to 0.40 to 5.70 +/- 0.39 (p < 0.01) and from 70 +/- 18 to 83 +/- 18 (p < 0.01) respectively. In addition, low and high frequency power and SDANN index, lower at baseline in Group A than in the other two groups, were comparable in the three groups after coronary angioplasty. The evolution of diastolic and systolic wall stress indices paralleled that of the above three parameters. In conclusion, diastolic and systolic wall stress indices, in addition to segmental left ventricular dysfunction, were synergistically involved in determining sympathovagal imbalance in patients with significant coronary artery disease; the reversal of left ventricular dysfunction and wall stress indices improves the profile of heart rate variability. Alterations in cardiac geometry and wall stress influence mainly the discharge of afferent sympathetic and efferent parasympathetic innervations and also principally the long-term heart rate variations instead of short-term modulation.