BACKGROUND The combination therapy of low-dose diltiazem or bexatolol with digoxin can be a useful adjunct for achieving heart rate control with minimal side effects. But there has not been a study including patients with impaired left ventricular function and evaluating whether the beneficial effects of medication will be maintained during a follow-up period. OBJECTIVE The purpose of this study was three-fold: (1) to compare the efficacy of digoxin with low-dose diltiazem and digoxin with low-dose betaxolol on randomized crossover study; (2) to evaluate whether the beneficial effects of medication will be maintained after 7 months; (3) to evaluate the safety of the combination therapy in patients with impaired left ventricular function. METHODS We did a prospective randomized crossover study in 35 patients with chronic atrial fibrillation (AF) including 15 patients with left ventricular dysfunction. After enrollment, each patient was evaluated for heart rate, blood pressure, rate-pressure products, maximal exercise tolerance at rest and during symptom-limited treadmill test before medication, at 4 weeks after medication of digoxin (0.125-0.5 mg daily) with diltiazem (90 mg twice daily), and at 4 weeks after digoxin with betaxolol (20 mg once daily). We performed 24-h ambulatory electrocardiogram (ECG) in 15 patients at the end of each phase of treatment. We repeated symptom-limited treadmill test like above method in 15 patients at 7 months of medication. RESULTS (1) Ventricular rates were significantly reduced in digoxin with low-dose betaxolol therapy at rest and during exercise (67 +/- 3, 135 +/- 5 (mean +/- S.E.M.) beats/min, respectively) in comparison to digoxin with low-dose diltiazem therapy (80 +/- 7, 154 +/- 5) (P < 0.05). (2) Rate-pressure products were significantly less in digoxin with low-dose betaxolol at rest and during exercise (85 +/- 4, 213 +/- 12 x 10(2) mmHg/min) than in digoxin with low-dose diltiazem therapy (105 +/- 6, 269 +/- 12) (P < 0.05). (3) Exercise capacity was significantly improved in digoxin with low-dose betaxolol (9.3 +/- 0.5 METS) or digoxin with low-dose diltiazem (9.7 +/- 0.5) in comparison to control state (8.3 +/- 0.5) (P < 0.05). (4) At 7 months evaluation, there was no significant difference between at 4 weeks and at 7 months. (5) Results on 24-h ambulatory ECG showed the same findings as on treadmill test. (6) Although side effects occurred more frequently in digoxin with low-dose betaxolol therapy, they were minimal and no patient had to withdraw medication. Worsening of left ventricular dysfunction was not observed. CONCLUSIONS Our study suggested that (1) combination therapy of low-dose betaxolol with digoxin was more superior to low-dose diltiazem with digoxin in controlling ventricular rate and reducing rate-pressure products; (2) the effects controlling ventricular rate, reducing rate-pressure products and improving exercise capacity have been well maintained even after 7 months of medication with each combination therapy.