Relation between heart rate recovery index and syntax score

Background: Heart rate recovery is a function of vagal reactivation, and its impairment is an independent prognostic indicator for cardiovascular and all-cause mortality. We aimed to search relationship between heart rate recovery after exercise stress test and SYNTAX score. Methods: This was a single-center retrospective cohort study. 346 individuals were included. Heart rate recovery index (HRRI) is defined as the reduction in heart rate from the rate at peak exercise to the rate at the 1st, 2nd and 3rd minutes after the cessation of the exercise stress test, and these results were indicated as HRRI1, HRRI2 and HRRI3 respectively. The SYNTAX scores were calculated for all patients. The SYNTAX score was divided into tertiles as low (≤22), intermediate (2332), and high (≥33). Results: Compared to control group, patients with impaired HRRI1 had significantly higher SYNTAX scores while HRRI2 and HRRI3 demonstrated a marginally significant relation with high SYNTAX score. In bi-serial and partial correlational analyses, HRRI1, HRRI2 and HRRI3 were positively correlated with the high SYNTAX score (r = -0.366, p < 0.001; r = 0.313, p < 0.001; r = -0.322, p < 0.001). After adjustments were made for age, gender, hypertension, diabetes mellitus, hyperlipidemia and uric acid levels, HRRI1 was still significantly associated with high SYNTAX score (r = -0.236, 95% CI, p = 0.002), while the relationships between the high SYNTAX score and HRRI2 and HRRI3 lost significance (r = 0.131, p =0.091; r = -0.159, p =0.040). ROC analyses showed HRRI 1,2 and 3 had moderate predictive ability for high SYNTAX score (AUC=0.713, 0.636-0.790, cut-off:22.5 bpm), (AUC=0.673, 0.593-0.752, cut-off:45.5 bpm) and (AUC=0.674, 0.5940.764, cut-off:60.5 bpm) respectively with CI 95%. Conclusions: A delayed decline of heart rate has been associated with increased risk of cardiovascular mortality. In our study lower heart rate recovery at first minute was associated with higher SYNTAX score. *Correspondence to: Sezen Bağlan Uzunget, MD, Ankara Ufuk University Hospital, Faculty of Medicine, Cardiology Department, Turkey, E-mail: sezenbaglan@hotmail.com


Introduction
Exercise stress testing, which is now widely available at a relatively low cost, is currently used most frequently to estimate cardiovascular performance. Basically, it is performed to identify angina or electrocardiographic changes induced by exercise. Determining exercise capacity, chronotropic response, heart rate reserve, heart rate recovery index and ventricular ectopic beats, increases the value of the test.
Heart rate recovery (HRR) is a function of vagal reactivation, and its deterioration is an independent prognostic indicator for cardiovascular events and all-cause mortality. HRRI is calculated by extracting the heart rates at 1, 2 and 3 minutes after cessation the test, from the patient's heart rate at finalizing the exercise. Many studies have shown that the HRRI in the 1st and 2nd minutes highly predicts prognosis in coronary artery patients and reported that those with impaired HRRI have a significantly higher risk of mortality [1][2][3].
SYNTAX score is an angiographic tool for grading the complexity of coronary artery disease according to the lesion number, lesion locations and functional importance. It was shown that SYNTAX scoring system could predict in-hospital, early and six months adverse cardiac outcomes [4,5]. Early prediction of mortality and morbidity enables taking precautions for complications in high risk patients.
In this study we aimed to search relationship between heart rate recovery index and SYNTAX score.

Methods
This was a single-center retrospective cohort study approved by the relevant Ethics Committee. All consecutive candidates that had had a coronary angiography within 30 days of exhibiting an abnormal exercise test result at our clinic, between January 2014 and December 2016, were enrolled. Of the subjects which we met the exercise test results on file scans, the ones reached the target heart rate were included in this study. The exclusion criteria were as follows: previous diagnosis of CAD, myocardial infarction, slow coronary flow, previous revascularization, over 75 years old, inability to reach 85% of their age-predicted maximum heart rate in exercise test, presence of atrial fibrillation or flutter, history of chronic obstructive pulmonary disease, congestive heart failure, moderate to severe valvular heart disease, and the use of beta blocker (BB) medication, digoxin or calcium channel blockers. According to our inclusion and exclusion criteria we collected data on 346 subjects who exhibited ST depression and reached at least 85% of the age-predicted maximal heart rate during treadmill exercise test. Baseline demographic, clinical, and angiographic characteristics data and stress test results were retrieved from the patient files and the electronic centralized clinical database of our center. All angiographic parameters were calculated by an experienced cardiologist who was blinded to the procedural data. The SYNTAX score was derived from the summation of the individual scores for each separate lesion producing ≥50% luminal obstruction in vessels ≥ 1.5 mm diameter.
All patients in the study had underwent exercise treadmill testing using the Bruce protocol. In our clinical applications, patients do not perform a cool-down walk and are placed supine as soon as possible after exercise. The predicted peak heart rate was calculated as (220 − age), and the aim was to reach at least 85% of the age-predicted heart rates. The end of exercise was tagged, and at least 3 minutes of postexercise heart rate recorded with the subject at rest. Heart rate recovery indices were defined as the reduction in heart rate from the rate at finalizing exercise to the rate at the 1 st , 2 nd and 3 rd minutes after the cessation of the exercise stress test, and these results were indicated as HRRI1, HRRI2 and HRRI3 respectively.

Statistical analysis
Data analyses were performed by using SPSS for Windows, version 22.0 (SPSS Inc., Chicago, IL, United States). Whether the distribution of continuous variables was normal or not was determined by Kolmogorov Smirnov test. Levene test was used for the evaluation of homogeneity of variances. Unless specified otherwise, continuous data were described as mean +/-SD for normal distributions, and median (minimum -maximum value) for skewed distributions. Categorical data were described as number of cases (%).
While the differences in normally distributed variables among more than two independent groups were analyzed by One-Way ANOVA, otherwise, Kruskal Wallis test was applied for comparisons of the normally data. When the p-value from One-Way ANOVA or Kruskal Wallis test statistics were statistically significant post hoc LSD or Conover's non-parametric multiple comparison test were used to know which group differ from which others.
Univariate and multivariate logistic regression were used to determine predictors for SYNTAX score. Any variable having a significant univariate test at some arbitrary level is selected as a candidate for the multivariate analysis. We base this on the Wald test from logistic regression and p-value cut-off point of 0.25. Receiver Operating Characteristic (ROC) curve analysis was performed to detect the best sensitivity and specificity value of HRRI in predicting SYNTAX score.

Results
All patients and healthy participants had sinus rhythm and normal 12-lead ECG results at rest. The baseline characteristics of the study groups are shown in Table 1 Severity of coronary artery diseases was determined by SYNTAX Score. Data are expressed as mean±standard deviation or median (minimum-maximum) for continuous variables and number (percentage) for categorical variables; Φ: One way Anova test, β: Kruskal wallis test and ŧ: Chi-square; least significant difference (LSD) or conover-Inman test were performed for the binary comparisons among the groups and the p value was set at 0.05. Significant differences were found between; a: C vs GI, b: C vs GII, c: C vs GIII, d: GI vs GII, e: GI vs GIII, g: GII vs GIII; BMI -body mass index; LDL -low-density lipoprotein; HDL -high density lipoprotein; SBP -systolic blood pressure; DBP -diastolic blood pressure; METs -metabolic equivalent; bpm -beat per minute; HRRI1 -heart rate recovery index at first minute; HRRI2 -heart rate recovery index at second minute; HRRI3 -heart rate recovery index at third minute; ACE -accreditation for cardiovascular excellence; mm-millimeter.
All 346 subjects were categorized into 4 groups based on SYNTAX scores: control group, low (≤22), intermediate (23)(24)(25)(26)(27)(28)(29)(30)(31)(32), and high (≥33) SYNTAX score. Significant differences for gender, DM and HT status were found between four groups. The binary comparisons of these four groups for HRRI1, HRRI2 and HRRI3, also demonstrated statistically significant differences (p<0.001). Fasting glucose and serum creatinine levels, duration of the exercise period and the workload in METS were statistically significant in some of the binary comparisons. Univariate analyses of each risk factor for predicting SYNTAX scores were performed. Because of high correlation between HRRI1, 2 and 3, influence of each of them was assessed separately. Any variable having a significant univariate test (p<0.25) was selected as a candidate for the multivariate analysis (Tables 2-4). Mean HRRI1, 2 and 3 durations of control group and SYNTAX groups are shown in Figure 1. Mean reduction in heart rate declines as the severity of CAD increases.  Table 5). Figures 2,3 & 4: ROC curve analyses for predictive values of HRRI in detecting CAD severity evaluated by SYNTAX score.In the low SYNTAX (≤22) group; no cut-offs were given for HRRI1, HRRI2 and HRRI3 due to all of the asymptotic significances were greater than 0.5 ( Figure 2).
In the high SYNTAX group(≥33); HRRI 1,2 and 3 showed a moderate predictive ability for high SYNTAX score according to the area under the ROC curve (AUC=0.    (Figure 4).

Discussion
Coronary artery disease (CAD) is the leading cause of morbidity and mortality in many developed countries [6]. Scientists and researchers have concentrated to reduce the burden by enhancing new identification and treatment strategies. Besides it is inexpensive and noninvasive, due to its reliability, exercise stress test is the initial test employed in the diagnosis of CAD. Exercise testing has been suggested to be valuable and less dependent on the patient's pre-test possibility for the purpose of risk stratification [7,8]. While evaluating a stress test, ST segment deviation, exercise hypotension or hypertension, maximum exercise capacity, chronotropic response to exercise, heart rate variability and heart rate recovery are the parameters that have been shown to carry prognostic values [9][10][11][12][13][14][15]. Several studies have attempted to determine exercise test scores that incorporate clinical and demographic risk factors not based on the exercise test itself. An example of such is the Duke Treadmill Score. In a latest study by Dzenkeviciute et al., a negative correlation between Duke treadmill score and significant coronary artery stenosis (r = -0.181, p = 0.005), SYNTAX score (r=-0.173; p=0.007), cardiac revascularizations (r=-0.213; p=0.001) were found, however, its significance to predict cardiac revascularization was dependent on confounding clinical findings [16].
There is considerable recent evidence that heart rate recovery is an important parameter thought to reflect the cardiac autonomic nervous system function [17]. Changes in heart rate during and following exercise are determined by the balance between the excitatory sympathetic influence and the inhibitory parasympathetic influence. Sympathetic activity that increased during exercise reduces after finalizing exercise, however, recovery of the heart rate immediately after exercise is primarily a function of vagal reactivation, which is thought to be most important during the first minute after exercise [18]. In healthy and asymptomatic individuals, the heart rate falls rapidly within the first 30 seconds after exercise followed by a slower reduction [19]. In the case of presence of ischemia, ischemia-induced locally liberated adenosine, hydroxyl radicals and purinergic peptides can provoke modulations in both centripetal and centrifugal processing of intrathoracic nervous system [20]. Ischemia-induced cardio-cardiac reflexes and blunted baroreceptor reflexes result in dysfunction of cardiac autonomic nervous system and resulting deterioration of recovery of heart rate. Slower HRRI rates denote a significantly higher risk of death. HRRI, which is a simple and non-invasive method for evaluating sympathovagal balance, is not routinely evaluated in daily clinical practice even though it is an independent risk factor for cardiovascular diseases.
Because depressed vagal tone is associated with cardiovascular death, we hypothesized that decreased HRRI would be predictive of high SYNTAX score which is one of the major determinants of severity and mortality of CAD [21].
Airaksinen, et al. showed that cardiac vagal efferent activity is depressed in coronary artery disease even before the development of symptoms [22]. Also, a very rapid HR recovery immediately after exercise was found associated with lower risk of CHD and CVD events in a subanalysis of the Framingham Heart Study [1]. Vivekananthan, et al. [23] reported that HRR is an independent predictor of mortality along with angiographic severity of CAD, left ventricle function and exercise capacity.
Cole, et al. [24] studied 2428 adults who had been referred for exercise nuclear perfusion scans and they found that less than 12 bps In the high SYNTAX group(≥33); HRRI 1,2 and 3 showed a moderate predictive ability for high SYNTAX score according to the area under the ROC curve (AUC=0.713, 0.636-0.790, with 75.9% sensitivity and 66.7% specificity, cut-off:22.5 bps), (AUC=0.673, 0.593-0.752, with 74.7% sensitivity and 52.9% specificity, cut-off:45.5 bps) and (AUC=0.674, 0.594-0.764, with 93.1% sensitivity and 41.4% specificity, cut-off:60.5 bps) respectively with CI 95% (Figure 4). decrease in heart rate in the 1st minute after exercise was associated with a relative risk of 2.0 for death from any cause over a 6-year period after adjustment of other cardiac risk factors (95% confidence interval, 1.5-2.7; p<0.001). Current guidelines suggest that rapid supine patient placement after the exercise test offer advantages over other methods and should be considered [25]. A protocol with a staged cool-down walk was used in the aforementioned study. In our study, a protocol with prompt supine placement was used and ROC curve analysis determined the best cut-off value of HRRI1 as 22.5 bpm, with 75.9% sensitivity and 66.7% specificity to predict severe (Syntax >33) score. In a study of Akyuz, et al. no correlation was found between HRRI and Gensini scores and the number of coronary artery involvements, the only relationship was denoted between HRRI and the presence of CAD [26]. In our study, HRRI1 was directly correlated with high SYNTAX score which primarily defines severe CAD. Both after ROC curve analyses and partial correlation analyses adjusted for other possible confounders; HRRI1 and HRRI3 had significant, HRRI2 had moderate predictive values for high SYNTAX scores in patients with CAD. The most likely mechanism by which severe CAD affects HRRI is most likely due to the strong association between myocardial ischemia and cardiac autonomic dysfunction [18,20].

Conclusion
Heart rate recovery can be defined as the rate at which the HR declines from either maximal or submaximal exercise to resting levels and has been identified as a powerful and independent predictor of cardiovascular and all-cause mortality in healthy adults. Our results point to the importance of the heart rate recovery in the identification of high CV risk patients and provides additional support for routine incorporation of heart rate recovery, at first minute, into standard risk stratification assessments among patients with suspected or documented CAD. Nevertheless, this is an observational and nonrandomized study; our results should be confirmed prospectively to confirm value of HRRI to predict CAD severity.