Amiodarone, a potent class III antiarrhythmic agent with adrenergic antagonism properties, is administered increasingly to diabetic patients with cardiac arrhythmias refractory to all other available forms of therapy. Because a large percentage of diabetic patients show a perturbed autonomic regulation of the cardiovascular system, including a pertubed regulation of heart rate, we studied the antiarrhythmic response as well as the early effects (within 5 days) on heart rate of an intravenous amiodarone loading dose in diabetic patients. Seven type II (noninsulin-dependent) diabetic patients (age 64.7 +/- 9.7 years), affected by uncontrolled atrial fibrilation or atrial flutter, were enrolled for the study and a group of 12 well-matched (for age, sex and arrhythmia) nondiabetic patients served as a control group. It was found that before amiodarone administration, nondiabetic patients showed significantly wider variations in the circadian rhythm of heart rate values than diabetic patients (p = 0.0062, unpaired t-test). In all patients but one (who was nondiabetic), amiodarone treatment resulted in a cardioversion to sinus rhythm. After amiodarone administration, nondiabetic patients showed a significantly greater decrease (p = 0.0011) in heart rate values in comparison with the diabetic group (-35% vs. -20% on average, at the end of the study). Furthermore, in nondiabetic patients there was also an earlier significant fall (within the first 4 h after the start of treatment with amiodarone, p < 0.001) in the heart rate values in comparison with diabetic patients, in whom a significant decrease (p < 0.001) was found only at the 4th day. A significant (p = 0.0004), more rapid onset of the antiarrhythmic response to the drug was found in nondiabetic patients (6.8 +/- 6.0 h) in comparison with diabetic patients (98.0 +/- 14.8 h). Our findings suggest that the antiarrhythmic effects of amiodarone in diabetic patients with uncontrolled atrial fibrilation or atrial flutter may be delayed in comparison with nondiabetic patients. This altered response may be (at least in part) due to the diabetic autonomic neuropathy. Our study indicates that the presence of diabetes mellitus always must be taken into account when patients are enrolled for large, prospective, randomized trials, planned to evaluate the antiarrhythmic effects of amiodarone given intravenously.