Assessment of Nonlinear Heart Rate Dynamics after Beating-Heart Revascularization

Abstract

Advanced nonlinear methods of measuring heart rate variability (HRV) derived from the mathematics of complex dynamics and fractal geometry have provided new insights into the abnormalities of heart rate behavior in various pathologic conditions. These methods have provided additional prognostic information compared with traditional HRV measures and clearly have complemented the conventional linear methods. Knowledge about the behavior of complex cardiac dynamics indices after different cardiac procedures is very limited, however. We aimed to clarify how nonlinear heart rate dynamics are affected by beating-heart revascularization (off-pump coronary artery bypass graft [CABG] surgery) within the first week after the procedure. Included in the study were 66 patients who had isolated stable multivessel coronary artery disease and were in normal sinus rhythm. The patients were on chronic beta-blocker therapy and were scheduled for off-pump CABG. We performed 15-minute high-resolution electrocardiographic recordings preoperatively and on the third and seventh postoperative days to assess linear and nonlinear heart rate dynamics. Frequency-domain measurements, detrended fluctuation analysis (DFA) with short-term (<or=11 beats, alpha1) and long-term (>11 beats, alpha2) correlation properties of RR-intervals, and fractal dimension (FD) measurements (average, high, and low) were made. Arrhythmia was monitored preoperatively with 24-hour Holter recordings, postoperatively by continuous monitoring for the first 4 days after the procedure, and subsequently by clinical monitoring; 24-hour Holter recordings were obtained again on the seventh postoperative day. We used the paired-samples Student t test, the Mann-Whitney U test, and the Fisher exact test for statistical analyses. Differences in arrhythmia occurrence before and after the procedure were tested with the Wilcoxon signed rank test and the McNemar test. A P level < .05 was considered statistically significant. Values for all frequency-domain parameters decreased significantly after off-pump CABG (P< .001). Values for the alpha1 and high FD parameters decreased significantly after the procedure (P= .028 and .001, respectively), whereas alpha2 increased significantly (P= .023). DFA alpha1 was significantly lower in patients with postoperative atrial fibrillation than in patients remaining in sinus rhythm (mean +/- SD, 0.79+/-0.32 versus 1.13+/-0.45 [P= .003] on the third postoperative day; 0.89+/-0.31 versus 1.22+/-0.34 [P< .001] on the seventh postoperative day), whereas low and average FDs were significantly higher (1.84+/-0.16 versus 1.68+/-0.19 [P= .003] on the third postoperative day and 1.77+/-0.18 versus 1.66+/-0.17 [P= .01] on the seventh postoperative day for the low FD; 1.83+/-0.09 versus 1.76+/-0.10 [P= .011] on the third postoperative day and 1.80+/-0.11 versus 1.73+/-0.10 [P= .014] on the seventh postoperative day for the average FD). The low FD was significantly higher on the third postoperative day in patients with postoperative deterioration of ventricular ectopy than in patients with improved ventricular ectopy (1.74+/-0.17 versus 1.48+/-0.08, [P= .03]). The decreases in alpha1, average FD, and high FD indicate that a profound decay of cardiac complexity and fractal correlation can be observed after off-pump CABG. Furthermore, a more extensive impairment of nonlinear indices was observed in patients who developed postoperative arrhythmias than in those who remained in stable sinus rhythm. Our findings suggest that the postoperative hyperadrenergic setting acts as a preliminary condition in which both reduced and enhanced vagal activity may predispose patients to arrhythmia, indicating that postoperative rhythm disturbances are an end point associated with divergent autonomic substrates.