Decreased heart rate variability (HRV) correlates with increased sympathetic or decreased vagal tone. This could contribute to increase local coronary hyperreactivity caused by atherosclerotic plaque disruption, thus facilitating progression from unstable angina to acute myocardial infarction (AMI). To test this hypothesis we studied 92 patients admitted to the coronary care unit for episodes of chest pain at rest associated with transient ST shifts (> 0.15 mV). Patients who developed AMI in the first 24 hours, as well as those with previous AMI, concomitant valvular or myocardial diseases or diabetes mellitus were not enrolled in the study. Thirty age-matched subjects without any evidence of coronary artery disease were chosen as controls. All patients underwent a 2 to 5 day continuous Holter monitoring during full medical treatment (including beta-blockers, heparin and aspirin). Angiography was performed within 1 week in 88 of the 92 patients. During follow-up (mean duration of 16 +/- 5 days), 26 patients (Group I) had a major coronary event (6 deaths, 7 non fatal AMI, 13 urgent revascularizations). The remaining 66 patients (Group II) had a good clinical outcome. ECG recordings during ST shifts were excluded from Holter monitoring analysis. Time domain measurements of HRV predicted mortality and total events. The most powerful predictors was the standard deviation of the means of the 5 min R-R intervals (SDANN index) which was significantly (p < 0.001) lower in Group I than Group II (55 +/- 18 versus 87 +/- 29).(ABSTRACT TRUNCATED AT 250 WORDS)