Europace Advance Access published January 6, 2015
CLINICAL RESEARCH
Europace doi:10.1093/europace/euu338
Determinants of heart rate turbulence in individuals without apparent heart disease and in patients with stable coronary artery disease Gaetano Pinnacchio, Gaetano Antonio Lanza*, Alessandra Stazi, Giulia Careri, Ilaria Coviello, Roberto Mollo, and Filippo Crea Institute of Cardiology, Catholic University of the Sacred Heart, Rome, Italy Received 7 June 2014; accepted after revision 3 November 2014
Aims
----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords
Heart rate turbulence † Healthy subjects † Coronary artery disease
Introduction Heart rate turbulence (HRT) represents a physiologic biphasic chronotropic response of sinus rhythm to a premature ventricular complex (PVC). This response is believed to be mainly mediated by the cardiac autonomic nervous system, and consists of an early heart rate acceleration, consistent with an immediate vagal withdrawal, and a late heart rate deceleration, reflecting a subsequent restoration of vagal activity.1 – 5 The two components of HRT are measured as turbulence onset (TO) and turbulence slope (TS), respectively.1 Several studies have shown that HRT impairment reflects cardiac autonomic dysfunction, in particular impaired vagal-mediated
baroreflex sensitivity.6,7 Moreover, several studies have shown that impaired HRT predicts an increased risk of fatal events and ventricular tachyarrhythmias in patients with acute myocardial infarction (MI)8 – 12 or congestive heart failure.13,14 Despite its prognostic value, to date only few studies have tried to define the physiological characteristics of HRT in apparently healthy individuals, i.e. without any evidence of cardiac disease.15 – 18 Specifically, how clinical features, such as cardiovascular risk factors (CVRFs), left ventricular (LV) function, and drug therapy, affect HRT parameters in people without cardiac disease has poorly been investigated. Thus, in this study, we aimed to define characteristics and determinants of HRT parameters in a large cohort of consecutive unselected
* Corresponding author. Istituto di Cardiologia, Universita` Cattolica del Sacro Cuore, Largo A. Gemelli, 8, 00168 Roma, Italy. Tel: +39 06 3015 4187; fax: +39 06 3055535. E-mail address [email protected] Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2015. For permissions please email: [email protected].
Downloaded from by guest on June 3, 2015
To assess the characteristics and determinants of heart rate turbulence (HRT) in individuals without any apparent heart disease and in patients with coronary artery disease (CAD). ..................................................................................................................................................................................... Methods and Heart rate turbulence parameters, turbulence onset (TO), and turbulence slope (TS) were calculated on 24 h electrocardiogram recordings in 209 individuals without any heart disease (group 1) and in 157 CAD patients (group 2). In group results 1, only age independently predicted abnormal TO (≥0%) [odds ratio (OR), 1.05; P , 0.001], while predictors of abnormal TS (≤2.5 ms/RR) were age (OR, 0.85; P , 0.001) and hypertension (OR, 0.19; P ¼ 0.028). In group 2 patients, only age independently predicted TO (OR, 1.03; P ¼ 0.038), while age (OR, 0.90; P ¼ 0.001) and left ventricular ejection fraction (LVEF; OR, 1.07; P ¼ 0.008) predicted TS. Heart rate turbulence values were different in groups 1 and 2. Turbulence onset was (mean, standard deviation) 21.80 + 2.24 vs. 20.73 + 1.61%, respectively (P , 0.001), whereas TS was (median, interquartile interval) 5.83 (3.25–10.55) vs. 2.93 (1.73–5.81) ms/RR, respectively (P , 0.001). Coronary artery disease group, however, did not predict abnormal HRT parameters in multivariable analyses, both in the whole population and when comparing two subgroups matched for age and gender. Age and (for TS) LVEF, indeed, were the only independent predictors of abnormal HRT. ..................................................................................................................................................................................... Conclusions Age is a major HRT determinant both in subjects without any apparent heart disease and in stable CAD patients. Hypertension and LVEF contribute independently to HRT in these two groups, respectively. Coronary artery disease group was not by itself associated with abnormal HRT parameters in multivariable analyses.
Page 2 of 7
What’s new? † Our data show that, in subjects without any apparent cardiac disease, only age and, in part, hypertension are significant independent predictors of heart rate turbulence (HRT). † Heart rate turbulence is largely independent of cardiovascular risk factors, as well as medical therapy, in apparently healthy individuals. † Age remains a major modulator of HRT in coronary artery disease (CAD) patients, in whom left ventricular ejection function also assumes a major role. † Although HRT showed significantly different values in CAD patients compared with apparently healthy subjects, stable CAD is by itself only a weak independent predictor of HRT.
subjects without any apparent heart disease undergoing routine 24 h Holter ECG recording. At the same time, we assessed whether factors influencing HRT were different in patients with documented coronary artery disease (CAD).
Population From September 2009 to May 2010, we prospectively studied consecutive subjects without any overt heart disease referred to our centre to undergo 24 h ambulatory Holter ECG recording. Subjects with a cardiac rhythm other than sinus, presence of pacemaker, a number ,5 PVCs suitable for HRT analysis, evidence of diabetes, and/or significant systemic disease (e.g. neoplasia, acute or chronic inflammatory disease, respiratory, hepatic, or renal failure) were excluded. Each subject underwent an interview with a physician, who collected detailed information about clinical history with the aid of a standardized questionnaire. Data acquired included age, gender, CVRFs, reason for Holter ECG recording, history of cardiovascular disease, and drug therapy. Hypertension was defined as referred blood pressure ≥140/ 90 mmHg or assumption of anti-hypertensive drugs; hypercholesterolaemia was defined as referred blood cholesterol levels .200 mg/dL or assumption of anti-cholesterolemic drugs; smoking was defined as any usual cigarette smoking in the last 6 months; obesity was defined as a body mass index (derived from weight and height) ≥30 kg/m2; familial history for CAD was defined as evidence of CAD in first- or seconddegree relatives, male ,55 years and female ,65 years. Clinical examination, ECG at rest, and bi-dimensional echocardiography were also obtained in all subjects to rule out any structural cardiac diseases, including LV hypertrophy. Left ventricular ejection fraction (LVEF) was calculated on echocardiographic four-chamber and two-chamber views using the modified Simpson’s method. Patients with a suspect of typical angina pectoris were excluded. Subjects with atypical chest pain underwent maximal exercise stress test and were included only when the test was normal. A consecutive group of patients with a documented history of CAD referred to undergo Holter ECG during the same period was also studied. Coronary artery disease was diagnosed in case of a documented history of previous MI, percutaneous and/or surgical coronary intervention or documented coronary stenosis (.50%) in at least one major
coronary artery at angiography. Only stable CAD patients were included. Accordingly, patients with a history of new onset (de novo) angina, recent destabilization of previously stable angina, or an acute coronary syndrome in the last 6 months were excluded. In addition, we excluded patients with any other cardiac and systemic disease. The study was approved by the Institutional Ethical Committee and written informed consent was obtained from each individual to participate in the study.
Heart rate turbulence All subjects underwent 24 h Holter ECG monitoring using three-channel tape recorders (Oxford Medilog FD5, Abingdon, UK). The recordings were analysed by two expert cardiologists using the system Oxford Excel 3.0. To calculate HRT parameters, RR intervals were exported as text files and analysed using a dedicated and validated software system.17 Heart rate turbulence was obtained from sequences of sinus RR intervals related to isolated PVCs that fulfilled the following criteria: (i) index PVC embedded into at least five preceding and 20 succeeding normal RR intervals; (ii) cycle length of all considered RR intervals .300 ms but ,2000 ms; (iii) beat-to-beat differences ,200 ms; and (iv) differences ,20% from the average of five preceding intervals.1,5 Accordingly, a single tachogram for HRT analysis consisted of a sequence of five RR intervals preceding and 20 RR intervals succeeding a PVC. Interpolated PVCs were excluded. Heart rate turbulence was assessed using the two standard parameters TO and TS. Turbulence onset, expressed as a percentage, was calculated using the formula [(RR1 + RR2) 2 (RR22 + RR21)]/(RR22 + RR21)×100, where RR1 and RR2 were the first and the second sinus RR intervals after the PVC and RR21 and RR22 were the first and the second sinus RR intervals preceding the PVC. Turbulence slope, expressed in ms/RR, was obtained as the maximal positive slope of all slopes of a series of regression lines obtained from all sequences (n ¼ 16) of five consecutive RR intervals included between the 1st and 20th RR interval after the compensatory pause that followed the PVC. Turbulence onset was calculated as the average of all TOs measured in the 24 h on individual tachograms of index PVCs, while TS was calculated on the averaged tachogram obtained averaging the RR intervals of individual tachograms preceding and following the index PVCs in the 24 h. Although ‘normal’ HRT values have not been defined yet, normal TO is expected to be negative, whereas normal TS is expected to be positive. Previous studies showed that TO ≥ 0% and TS ≤ 2.5 ms/RR were significantly associated with a worse prognosis in cardiac patients.1,8 – 11 Accordingly, we used these dichotomized values as cut-offs to consider HRT parameters as abnormal. A composite HRT variable was also obtained by combining TO and TS values, which results in three mutually exclusive categories: (i) TO ≤ 0% and TS ≥ 2.5 ms/RR interval (i.e. both parameters normal, category 0); (ii) either TO ≥ 0% or TS ≤ 2.5 ms/RR interval (i.e. only one parameter abnormal, category 1); and (iii) TO ≥ 0% and TS ≤ 2.5 ms/RR interval (i.e. both parameters abnormal, category 2).
Statistics Statistical analyses were done by SPSS 20.0 (SPSS, Inc.). Turbulence onset showed a Gaussian distribution, while TS showed a skewed distribution, as assessed by Kolmogorov –Smirnov test. Accordingly, between-group values of TO and TS were compared using unpaired t-test and Mann– Whitney test, respectively. Proportions were compared using x2 test. Pearson and Spearman rank-correlation tests, respectively, were used to assess the relation between HRT parameters with other variables.
Downloaded from by guest on June 3, 2015
Methods
G. Pinnacchio et al.
Page 3 of 7
Determinants of heart rate turbulence
Univariate and multivariable linear regression analysis was performed to identify variables predictive of HRT parameters, considered as continuous variables. Univariate and multivariable logistic regression was applied to identify variables associated with abnormal TO and TS values, as well as HRT category 1 or category 2, compared with category 0. All variables with a P-value≤ 0.1 on univariate analyses were included as independent covariates in the multivariable models. Furthermore, we also assessed differences in HRT parameters by performing comparisons between two subgroups of 65 subjects of groups 1 and 2, which were matched for age and gender. Data are reported as mean + SD for normally distributed variables and as median (with interquartile interval) for variables with skewed distribution. Differences were considered significant for P , 0.05.
Results
to an insufficient number of PVCs (n ¼ 287) or a history of diabetes (n ¼ 23). Thus, 209 individuals (age 61 + 16 years, males 49%) satisfied the inclusion criteria and formed the group without any apparent cardiac disease (group 1). In the same period, 392 patients with heart disease were screened. Overall, 235 patients were excluded, due to an insufficient number of PVCs (n ¼ 155) or evidence of a non-ischaemic disease of the heart (n ¼ 80). Thus, 157 patients (age 69 + 10 years, males 78%) met the inclusion criteria for the study and constituted the group of stable CAD patients (group 2). The main clinical characteristics of the two study groups are summarized in Table 1. Compared with group 2 patients, subjects of group 1 were younger, less frequently men, had a lower prevalence of CVRFs, and had a higher LVEF. Furthermore, they also had a lower number of PVCs.
General findings During the period of the study, 519 subjects without any apparent cardiac disease were examined, 310 of whom were excluded due
Table 1 Main clinical characteristics of subjects included in the study Group 1 (n 5 209)
Group 2 (n 5 157)
P-value
61 + 16 49
69 + 10 78
,0.001 ,0.001
1111 + 1942 149 + 186
1760 + 3445 170 + 202
0.02 0.32
70 + 10
67 + 9
0.005
1
19
,0.001
Family history of CVD (%)
32
54
,0.001
Hypertension (%)
54
80
,0.001
Dyslipidemia (%) Smoking (%)
38 17
65 13
,0.001 0.382
b-Blockers (%) ACE-inhibitors (%)
32 23
77 56
,0.001 ,0.001
Statins (%)
16
................................................................................ Age (years) Men (%) Holter data PVCs (Nr/24 h) Processed PVCs (Nr/24 h) Mean heart rate (b.p.m.) NSVT (%) CVRFs
Drug therapy
Diuretics (%) 20 Reason for Holter recording
77
,0.001
41
,0.001
Follow-up (%)
35
60
Palpitations (%) Chest pain (%)
32 12
12 13
Syncope (%)
14
10
7 62 + 5
5 51 + 12
Other (%) LVEF (%)
,0.001
,0.001
PVCs, premature ventricular complexes; NSVT, non-sustained ventricular tachycardia; CVRFs, cardiovascular risk factors; CVD, cardiovascular disease; ACE, angiotensin converting enzyme; LVEF, left ventricular ejection fraction.
Premature ventricular complexes suitable for HRT assessment in group 1 subjects were 149 + 186. Turbulence onset, in this group, was 21.80 + 2.24%, while TS was 5.83 (3.25– 10.55) ms/RR. At univariate analysis, a highly significant correlation was found between age and both HRT parameters (TO: r ¼ 0.38, P , 0.001; TS: r ¼ 20.63, P , 0.001; Figure 1). Furthermore, higher TO values and lower TS values were observed in subjects with, compared with those without, hypertension or dyslipidemia; female gender was also associated with lower TS values, compared with males (Table 2). At multivariable analysis, however, the only independent predictor of TO was age (OR, 1.05; CI: 1.02 – 1.07; P , 0.001), and the only independent predictors of TS were age (OR, 0.85; CI: 0.81 – 0.88; P , 0.001) and hypertension (OR, 0.19; CI: 0.04 – 0.83; P ¼ 0.028).
Heart rate turbulence in coronary artery disease patients In group 2 patients, the number of PVCs suitable of HRT analysis was 170 + 202. Turbulence onset in this group was 20.73 + 1.61%, while TS was 2.93 (1.73–5.81) ms/RR. At univariate analysis, a significant, although modest, correlation was found on both HRT parameters with age (r ¼ 0.24, P ¼ 0.002 and r ¼ 20.34, P , 0.001, for TO and TS, respectively; Figure 2), LVEF (r ¼ 20.23, P ¼ 0.012 and r ¼ 0.43, P , 0.001, for TO and TS, respectively; Figure 3), and, for TS only, heart rate (r ¼ 20.24, P ¼ 0.01). Furthermore, higher TO values were observed in patients with a history of previous coronary artery by-pass surgery (P ¼ 0.036) and in those treated with diuretic or antiarrhythmic drugs. Lower TS values were instead found in females and in patients with diabetes or hypertension (Table 3). At multivariable analysis, the only independent predictor of TO was age (OR, 1.03; CI: 1.002 –1.06; P ¼ 0.038), whereas independent predictors of TS were age (OR, 0.90; CI: 0.84 –0.95; P ¼ 0.001) and LVEF (OR, 1.07; CI: 1.02–1.13; P ¼ 0.008).
Downloaded from by guest on June 3, 2015
Variable
Heart rate turbulence in subjects without cardiac disease
Page 4 of 7
G. Pinnacchio et al.
24
6
r = 0.38; P < 0.001
4
18 TS (ms/RR)
2 To (%)
r = –0.63; P < 0.001
21
0 –2 –4
15 12 9
–6
6
–8
3 0 10
–10 10 20 30 40 50 60 70 80 90 100 Age
20
30
40
50 60 Age
70
80
90 100
Figure 1 Correlation of age with TO (left) and TS (right) in individuals without any apparent heart disease.
Table 2 Clinical predictors of TO and TS in group 1 subjects at univariate linear regression analysis TO
P-value
TS
P-value
Age Male gender
1.05 (1.034–1.071) 0.82 (0.44– 1.51)
,0.001 0.52
0.83 (0.80– 86) 6.47 (1.54– 27)
,0.001 0.011
Mean heart rate
0.99 (0.96– 1.03)
0.77
0.94 (0.88– 1.01)
0.068
CVRFs Hypertension
...............................................................................................................................................................................
0.001
0.01 (0.01– 0.04)
,0.001
2.14 (1.15– 3.99)
0.017
0.12 (0.03– 0.50)
0.004
Smoke Obesity
0.57 (0.25– 1.27) 1.03 (0.66– 1.59)
0.169 0.882
5.39 (0.79– 36) 0.52 (0.19– 1.45)
0.084 0.213
b-Blockers Ca2+-antagonists
1.34 (0.69– 2.58) 1.86 (0.82– 4.22)
0.381 0.135
0.19 (0.04– 0.92) 0.05 (0.01– 0.30)
0.040 0.002
ACE-inhibitors
1.87 (0.91– 3.88)
0.088
0.04 (0.01– 0.22)
,0.001
Statins Diuretics
1.73 (0.76– 3.91) 2.10 (0.99– 4.47)
0.190 0.052
0.04 (0.01– 0.24) 0.06 (0.01– 0.37)
0.001 0.002
0.96 (0.89– 1.03)
0.238
1.13 (0.94– 1.35)
0.194
Therapy
LVEF
CVRFs, cardiovascular risk factors; ACE, angiotensin converting enzyme; LVEF, left ventricular ejection fraction.
21
6 r = 0.24; P < 0.002 TS (ms/RR)
2 To (%)
r = –0.34; P < 0.001
18
4
0 –2 –4
15 12 9 6 3
–6
0 30
40
50
60 70 Age
80
90
30
40
50
60 70 Age
Figure 2 Correlation of age with TO (left) and TS (right) in patients with stable coronary artery disease.
80
90
Downloaded from by guest on June 3, 2015
2.78 (1.52– 5.05)
Dyslipidemia
Page 5 of 7
Determinants of heart rate turbulence
18
6
r = 0.43; P < 0.001
4
15
2
12
TS (ms/RR)
To (%)
r = –0.23; P = 0.012
0 –2 –4
9 6 3
–6
0 10
30
50 LVEF
70
10
30
50
70
LVEF
Figure 3 Correlation of LVEF with TO (left) and TS (right) in patients with stable coronary artery disease.
Table 3 Clinical predictors of TO and TS in group 2 patients at univariate linear regression analysis TO
P-value
TS
P-value
Age
1.04 (1.01–1.06)
0.002
0.85 (0.81–0.90)
,0.001
Male gender Mean heart rate
0.64 (0.35–1.18) 1.01 (0.98–1.04)
0.15 0.35
5.12 (1.15–22.8) 0.87 (0.82–0.92)
0.03 0.01
Diabetes Hypertension
1.54 (0.91–2.59) 1.48 (0.79–2.77)
0.10 0.22
0.15 (0.04–0.52) 0.06 (0.02–0.27)
0.003 ,0.001
Dyslipidemia
0.79 (0.46–1.35)
0.39
1.34 (0.36–4.99)
0.65
Family history of CVD Smoke
0.87 (0.52–1.45) 0.96 (0.45–2.02)
0.59 0.91
0.94 (0.27–3.29) 0.69 (0.11–4.29)
0.92 0.68
Obesity
0.94 (0.67–1.31)
0.70
1.43 (0.62–3.27)
0.39
b-Blockers Ca2+-antagonists
1.02 (0.55–1.86) 1.63 (0.88–3.01)
0.96 0.11
3.30 (0.76–14.3) 0.18 (0.04–0.81)
0.11 0.02
ACE-inhibitors
0.76 (0.46–1.26)
0.29
1.96 (0.56–6.86)
0.29
ARA Statins
1.74 (0.97–3.13) 0.75 (0.41–1.37)
0.06 0.35
0.21 (0.05–0.86) 1.96 (0.45–8.63)
0.03 0.37
...............................................................................................................................................................................
Therapy
Anti-platelets
1.39 (0.37–5.23)
0.62
3.91 (0.15–99)
0.41
Digoxin Nitrates
6.78 (0.71–64) 1.33 (0.65–2.73)
0.09 0.43
0.07 (0.01–19) 0.40 (0.07–2.32)
0.35 0.30
Antiarrhythmics
2.62 (1.22–5.64)
0.014
0.20 (0.03–1.34)
0.09
Diuretics LVEF
1.87 (1.12–3.04) 0.97 (0.95–0.99)
0.016 0.012
0.17 (0.05–0.59) 1.10 (1.04–1.16)
0.006 0.001
Previous AMI
0.80 (0.43–1.50)
0.48
1.52 (0.32–7.13)
0.59
Previous PCI Previous CABG
0.68 (0.41–1.13) 1.75 (1.04–2.95)
0.14 0.036
1.22 (0.34–4.33) 0.35 (0.10–1.28)
0.75 0.11
CVRFs, cardiovascular risk factors; CVD, cardiovascular disease; ACE, angiotensin converting enzyme; ARA, aldosterone receptor antagonists; LVEF, left ventricular ejection fraction; AMI, acute myocardial infarction; PCI, percutaneous coronary intervention; CABG, coronary artery bypass graft.
Between-group comparisons Values of both HRT parameters were significantly different in group 1 and group 2 patients, being worse in the latter (P , 0.001 for both TO and TS).
An abnormal TO (.0%) was found in 26 group 1 subjects (12.4%) and 42 group 2 patients (26.8%; P ¼ 0.001), whereas an abnormal TS (,2.5 ms/RR) was found in 31 group 1 subjects (14.8%) and 64 group 2 patients (40.8%; P ¼ 0.001), respectively. Furthermore, HRT
Downloaded from by guest on June 3, 2015
CVRFs
Page 6 of 7
G. Pinnacchio et al.
categories 0, 1, and 2 were detected in 160 (76.6%), 41 (19.6%), and 8 (3.8%) subjects of group 1, but in 78 (49.7%), 52 (33.1%), and 27 (17.2%) patients of group 2, respectively (P , 0.001). All these findings remained unchanged when only patients with frequent PVCs (≥10/h) were included in the analyses (data not shown). In multivariable linear regression analysis, presence of CAD (group 2) showed an independent association with TO (OR, 2.23; P ¼ 0.02), but not with TS (OR, 0.40; P ¼ 0.18). Moreover, in multivariable logistic regression analysis, the variable ‘group’ did not show a significant association with both ‘abnormal’ TO and TS values and with HRT category 1 or 2, whereas age (P , 0.001) and LVEF (P , 0.001) remained the only independent predictor of HRT parameters.
Comparison of matched groups Two samples of 65 subjects of the two groups could be matched for age and gender. As shown in Table 4, the two groups were also comparable with regard to CVRFs, whereas HRT parameters were significantly different between the two groups. In multivariable linear regression analysis, CAD group showed an independent association with TO (OR, 2.43; P ¼ 0.028), but not with TS (OR, 0.31; P ¼ 0.22). However, age (P , 0.05) and LVEF
Variable
Group 1 (n 5 65)
Group 2 (n 5 65)
P-value
Age (years)
67 + 6
67 + 9
1
Men (%) Holter data
50
50
1
PVCs (Nr/24 h)
1120 + 1969
1246 + 1816
0.70
Processed PVCs (Nr/24 h)
150 + 156
178 + 218
0.43
Mean heart rate (b.p.m.)
70 + 9
69 + 9
0.40
1
16
0.004
................................................................................
NSVT (%) CVRFs Diabetes (%)
0
0
1
Hypertension (%) Dyslipidemia (%)
69 58
77 72
0.42 0.13
Smoking (%)
19
19
1
Drug therapy b-Blockers (%)
42
76
ACE-inhibitors (%)
35
48
0.15
Statins (%)
28
77
,0.001
41 61 + 6
50 50 + 14
0.37 ,0.001
21.24 + 2.2 5.08 (2.95– 7.30)
–0.53 + 1.2 2.18 (1.19– 4.64)
0.03 ,0.001
Diuretics (%) LVEF (%)
,0.001
HRT parameters TO (%) TS (ms/RR)
PVCs, premature ventricular complexes; NSVT, non-sustained ventricular tachycardia; CVRFs, cardiovascular risk factors; ACE, angiotensin converting enzyme; LVEF, left ventricular ejection fraction; HRT, heart rate turbulence; TO, turbulence onset; TS, turbulence slope.
Discussion Two main findings emerge from our results. First, our data showed that, in subjects without any apparent cardiac disease, only age and, in part, hypertension significantly influence in an independent way HRT parameters. Second, age remains a major modulator of HRT in CAD patients, in whom LVEF also assumes a major role. Heart rate turbulence is an autonomically mediated phenomenon with a close relationship to baroreflex mechanisms. Premature ventricular complexes, indeed, produce a fall in blood pressure, which results in vagal withdrawal and sympathetic activation, with an initial acceleration of sinus node activity and return of blood pressure to normal values; this in turn results in baroreceptor stimulation with vagal recruitment and sympathetic inhibition, thus leading to a later deceleration of sinus node activity.2 Increasing age has consistently been shown to be associated with a decrease in cardiac vagal modulation, using various measures of cardiac autonomic function, as basal heart rate and heart rate variability, which increase and decrease with age, respectively.19 Accordingly, some previous studies, in restricted populations of subjects, showed that age was a significant determinant of HRT values.15 – 17 These findings are confirmed in our study, in which age was the most powerful independent predictor of HRT in subjects without cardiac disease, with increasing age being associated with progressively higher TO and lower TS values, respectively. Furthermore, our study is the first to show that HRT is largely independent of CVRFs, as well as medical therapy, in apparently healthy individuals. Until now, the relation between CVRFs and HRT parameters in healthy people had indeed poorly been investigated, with the few studies including only small samples of volunteers.15 – 18 Our study shows that only hypertension has an independent, although weak, association with abnormal HRT, in agreement with the wellknown impairment of baroreflex sensitivity observed in hypertensive patients.20 Of note, hypertension independently influenced TS but not TO, suggesting that the abnormality related to hypertension resides in the reflex mechanism that follows an increased, rather than in the one that follows a decreased, stimulation of baroceptors. A further main finding of our study is that an impaired LVEF is the only other major determinant of HRT in stable CAD patients. The latter result is not surprising, as a reduced LV function has consistently been found to be a major determinant of sympatho-vagal imbalance in several clinical settings, including acute and old MI and dilated cardiomyopathy,21,22 and a relation of HRT with LVEF was already shown in our previous study.17 Of note, the presence of a previous acute MI was instead by itself insufficient to significantly influence HRT values, suggesting that it may exert its influence on HRT only when it is large enough to impair LV function. Similarly, although HRT showed significantly different values in CAD compared with healthy subjects, CAD presence was only a weak predictor of HRT, being independently associated only with TO, when considered as a continuous variable, in multivariable analyses. Of note, the poor relation of obstructive CAD by itself with HRT parameters was confirmed when comparing two subgroups
Downloaded from by guest on June 3, 2015
Table 4 Main clinical characteristics of matched groups
(P , 0.05), but not CAD group, were independent predictors of abnormal TO and TS values and of HRT category 1 or 2 in these matched selected individuals.
Page 7 of 7
Determinants of heart rate turbulence
of non-cardiac subjects and of CAD patients selected for being similar in age and gender. In the latter subgroup differences were again found in HRT parameters, but statistical significance was lost when corrected for LVEF, in agreement with the independent association of LVEF with HRT parameters in the whole sample of CAD patients. Some limitations of our study should be acknowledged. First, we cannot be sure that all apparently ‘healthy’ individuals did not have any form of CAD, as most of them had some CVRF. However, the lack of any symptom and the evidence of normal ECG, echocardiography, and maximal exercise stress test makes unlikely that a sizeable proportion had significant CAD. Second, several subjects were being treated with cardiologic drugs, in particular b-blockers, which might have influenced HRT. However, drugs did not emerge as correlated with HRT in multivariable analyses, suggesting that their influence is, if any, limited. Our finding is, in fact, in agreement with the lack of significant effect of beta-blockade on HRT parameters found in previous studies of apparently healthy individuals.4
Conclusions
7.
8.
9.
10. 11.
12.
13.
14.
15.
16.
17.
Conflict of interest: none declared.
References 1. Schmidt G, Malik M, Barthel P, Schneider R, Ulm K, Rolnitzky L et al. Heart rate turbulence after ventricular premature beats as a predictor of mortality after acute myocardial infarction. Lancet 1999;353:1390 –96. 2. Wichterle D, Melenovsky V, Simek J, Malik J, Malik M. Hemodynamics and autonomic control of heart rate turbulence. J Cardiovasc Electrophysiol 2006;17:286 – 91. 3. Welch WJ, Smith ML, Rea RF, Bauernfeind RA, Eckberg DL. Enhancement of sympathetic nerve activity by single premature ventricular beats in humans. J Am Coll Cardiol 1989;13:69–75. 4. Lin LY, Lai LP, Lin JL, Du CC, Shau WY, Chan HL et al. Tight mechanism correlation between heart rate turbulence and baroreflex sensitivity: sequential autonomic blockade analysis. J Cardiovasc Electrophysiol 2002;13:427 –31. 5. Bauer A, Malik M, Schmidt G, Barthel P, Bonnemeier H, Cygankiewicz I et al. Heart rate turbulence: standards of measurement, physiological interpretation, and clinical
18.
19.
20. 21.
22.
Downloaded from by guest on June 3, 2015
In conclusion, our study shows that age is a major determinant of HRT values both in subjects without any apparent cardiac disease and in CAD patients. Only hypertension, among CVRFs, contributes significantly, although in a moderate way, to HRT in people without cardiac disease, whereas LVEF becomes a major predictor of HRT in CAD patients.
6.
use. International society for Holter and non invasive electrophysiology consensus. J Am Coll Cardiol 2008;52:1353 –65. Davies LC, Francis DP, Ponikowski P, Piepoli MF, Coats AJ. Relation of heart rate and blood pressure turbulence following premature ventricular complexes to baroreflex sensitivity in chronic congestive heart failure. Am J Cardiol 2001;87:737 – 42. Iwasaki M, Yuasa F, Yuyama R, Mimura J, Kawamura A, Motohiro M et al. Correlation of heart rate turbulence with sympathovagal balance in patients with acute myocardial infarction. Clin Exp Hypertens 2005;27:251 – 7. La Rovere MT, Bigger JT Jr, Marcus FI, Mortara A, Schwartz PJ. Baroreflex sensitivity and heart-rate variability in prediction of total cardiac mortality after myocardial infarction. Lancet 1998;351:478 –84. Barthel P, Schneider R, Bauer A, Ulm K, Schmitt C, Scho¨mig A et al. Risk stratification after acute myocardial infarction by heart rate turbulence. Circulation 2003;108: 1221 –26. Multicenter Post-Infarction Research Group. Risk stratification and survival after myocardial infarction. N Engl J Med 1983;309:331 –6. Julian DG, Camm AJ, Frangin G, Janse MJ, Munoz A, Schwartz PJ et al. Randomised trial of effect of amiodarone on mortality in patients with left-ventricular dysfunction after recent myocardial infarction: EMIAT. European Myocardial Infarct Amiodarone Trial Investigators. Lancet 1997;349:667 –74. Sulimov V, Okisheva E, Tsaregorodtsev D. Non-invasive risk stratification for sudden cardiac death by heart rate turbulence and microvolt T-wave alternans in patients after myocardial infarction. Europace 2012;14:1786 –92. Grimm W, Schmidt G, Maisch B, Sharkova J, Mu¨ller HH, Christ M. Prognostic significance of heart rate turbulence following ventricular premature beats in patients with idiopathic dilated cardiomyopathy. J Cardiovasc Electrophysiol 2003;14:819 –24. Marynissen T, Flore´ V, Heidbuchel H, Nuyens D, Ector J, Willems R. Heart rate turbulence predicts ICD-resistant mortality in ischaemic heart disease. Europace 2014;16:1069 – 77. Schwab JO, Eichner G, Shlevkov N, Schrickel J, Yang A, Balta O et al. Impact of age and basic heart rate on heart rate turbulence in healthy persons. PACE 2005;28: S198 –201. Grimm W, Sharkova J, Christ M, Schneider R, Schmidt G, Maisch B. Heart rate turbulence following ventricular premature beats in healthy controls. Ann Noninvasive Electrocardiol 2003;8:127 –31. Sestito A, Valsecchi S, Infusino F, Sgueglia GA, Bellocci F, Zecchi P et al. Differences in heart rate turbulence between patients with coronary artery disease and patients with ventricular arrhythmias but structurally normal hearts. Am J Cardiol 2004;93: 1114 –8. Lindgren KS, Makikallio TH, Seppanen T, Raatikainen MJ, Castellanos A, Myerburg RJ et al. Heart rate turbulence after ventricular and atrial premature beats in subjects without structural heart disease. J Cardiovasc Electrophysiol 2003;14:447 –52. Bonnemeier H, Richardt G, Potratz J, Wiegand UK, Brandes A, Kluge N et al. Circadian profile of cardiac autonomic nervous modulation in healthy subjects: differing effects of aging and gender on heart rate variability. J Cardiovasc Electrophysiol 2003; 14:791–9. Mussalo H, Vanninen E, Ika¨heimo R, Laitinen T, Laakso M, La¨nsimies E et al. Baroreflex sensitivity in essential and secondary hypertension. Clin Auton Res 2002;12:465 – 71. Graham LN, Smith PA, Stoker JB, Mackintosh AF, Mary DA. Time course of sympathetic neural hyperactivity after uncomplicated acute myocardial infarction. Circulation 2002;106:793 –7. Ajiki K, Murakawa Y, Yanagisawa-Miwa A, Usui M, Yamashita T, Oikawa N et al. Autonomic nervous system activity in idiopathic dilated cardiomyopathy and in hypertrophic cardiomyopathy. Am J Cardiol 1993;71:1316 –20.
CLINICAL RESEARCH
Europace doi:10.1093/europace/euu338
Determinants of heart rate turbulence in individuals without apparent heart disease and in patients with stable coronary artery disease Gaetano Pinnacchio, Gaetano Antonio Lanza*, Alessandra Stazi, Giulia Careri, Ilaria Coviello, Roberto Mollo, and Filippo Crea Institute of Cardiology, Catholic University of the Sacred Heart, Rome, Italy Received 7 June 2014; accepted after revision 3 November 2014
Aims
----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords
Heart rate turbulence † Healthy subjects † Coronary artery disease
Introduction Heart rate turbulence (HRT) represents a physiologic biphasic chronotropic response of sinus rhythm to a premature ventricular complex (PVC). This response is believed to be mainly mediated by the cardiac autonomic nervous system, and consists of an early heart rate acceleration, consistent with an immediate vagal withdrawal, and a late heart rate deceleration, reflecting a subsequent restoration of vagal activity.1 – 5 The two components of HRT are measured as turbulence onset (TO) and turbulence slope (TS), respectively.1 Several studies have shown that HRT impairment reflects cardiac autonomic dysfunction, in particular impaired vagal-mediated
baroreflex sensitivity.6,7 Moreover, several studies have shown that impaired HRT predicts an increased risk of fatal events and ventricular tachyarrhythmias in patients with acute myocardial infarction (MI)8 – 12 or congestive heart failure.13,14 Despite its prognostic value, to date only few studies have tried to define the physiological characteristics of HRT in apparently healthy individuals, i.e. without any evidence of cardiac disease.15 – 18 Specifically, how clinical features, such as cardiovascular risk factors (CVRFs), left ventricular (LV) function, and drug therapy, affect HRT parameters in people without cardiac disease has poorly been investigated. Thus, in this study, we aimed to define characteristics and determinants of HRT parameters in a large cohort of consecutive unselected
* Corresponding author. Istituto di Cardiologia, Universita` Cattolica del Sacro Cuore, Largo A. Gemelli, 8, 00168 Roma, Italy. Tel: +39 06 3015 4187; fax: +39 06 3055535. E-mail address [email protected] Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2015. For permissions please email: [email protected].
Downloaded from by guest on June 3, 2015
To assess the characteristics and determinants of heart rate turbulence (HRT) in individuals without any apparent heart disease and in patients with coronary artery disease (CAD). ..................................................................................................................................................................................... Methods and Heart rate turbulence parameters, turbulence onset (TO), and turbulence slope (TS) were calculated on 24 h electrocardiogram recordings in 209 individuals without any heart disease (group 1) and in 157 CAD patients (group 2). In group results 1, only age independently predicted abnormal TO (≥0%) [odds ratio (OR), 1.05; P , 0.001], while predictors of abnormal TS (≤2.5 ms/RR) were age (OR, 0.85; P , 0.001) and hypertension (OR, 0.19; P ¼ 0.028). In group 2 patients, only age independently predicted TO (OR, 1.03; P ¼ 0.038), while age (OR, 0.90; P ¼ 0.001) and left ventricular ejection fraction (LVEF; OR, 1.07; P ¼ 0.008) predicted TS. Heart rate turbulence values were different in groups 1 and 2. Turbulence onset was (mean, standard deviation) 21.80 + 2.24 vs. 20.73 + 1.61%, respectively (P , 0.001), whereas TS was (median, interquartile interval) 5.83 (3.25–10.55) vs. 2.93 (1.73–5.81) ms/RR, respectively (P , 0.001). Coronary artery disease group, however, did not predict abnormal HRT parameters in multivariable analyses, both in the whole population and when comparing two subgroups matched for age and gender. Age and (for TS) LVEF, indeed, were the only independent predictors of abnormal HRT. ..................................................................................................................................................................................... Conclusions Age is a major HRT determinant both in subjects without any apparent heart disease and in stable CAD patients. Hypertension and LVEF contribute independently to HRT in these two groups, respectively. Coronary artery disease group was not by itself associated with abnormal HRT parameters in multivariable analyses.
Page 2 of 7
What’s new? † Our data show that, in subjects without any apparent cardiac disease, only age and, in part, hypertension are significant independent predictors of heart rate turbulence (HRT). † Heart rate turbulence is largely independent of cardiovascular risk factors, as well as medical therapy, in apparently healthy individuals. † Age remains a major modulator of HRT in coronary artery disease (CAD) patients, in whom left ventricular ejection function also assumes a major role. † Although HRT showed significantly different values in CAD patients compared with apparently healthy subjects, stable CAD is by itself only a weak independent predictor of HRT.
subjects without any apparent heart disease undergoing routine 24 h Holter ECG recording. At the same time, we assessed whether factors influencing HRT were different in patients with documented coronary artery disease (CAD).
Population From September 2009 to May 2010, we prospectively studied consecutive subjects without any overt heart disease referred to our centre to undergo 24 h ambulatory Holter ECG recording. Subjects with a cardiac rhythm other than sinus, presence of pacemaker, a number ,5 PVCs suitable for HRT analysis, evidence of diabetes, and/or significant systemic disease (e.g. neoplasia, acute or chronic inflammatory disease, respiratory, hepatic, or renal failure) were excluded. Each subject underwent an interview with a physician, who collected detailed information about clinical history with the aid of a standardized questionnaire. Data acquired included age, gender, CVRFs, reason for Holter ECG recording, history of cardiovascular disease, and drug therapy. Hypertension was defined as referred blood pressure ≥140/ 90 mmHg or assumption of anti-hypertensive drugs; hypercholesterolaemia was defined as referred blood cholesterol levels .200 mg/dL or assumption of anti-cholesterolemic drugs; smoking was defined as any usual cigarette smoking in the last 6 months; obesity was defined as a body mass index (derived from weight and height) ≥30 kg/m2; familial history for CAD was defined as evidence of CAD in first- or seconddegree relatives, male ,55 years and female ,65 years. Clinical examination, ECG at rest, and bi-dimensional echocardiography were also obtained in all subjects to rule out any structural cardiac diseases, including LV hypertrophy. Left ventricular ejection fraction (LVEF) was calculated on echocardiographic four-chamber and two-chamber views using the modified Simpson’s method. Patients with a suspect of typical angina pectoris were excluded. Subjects with atypical chest pain underwent maximal exercise stress test and were included only when the test was normal. A consecutive group of patients with a documented history of CAD referred to undergo Holter ECG during the same period was also studied. Coronary artery disease was diagnosed in case of a documented history of previous MI, percutaneous and/or surgical coronary intervention or documented coronary stenosis (.50%) in at least one major
coronary artery at angiography. Only stable CAD patients were included. Accordingly, patients with a history of new onset (de novo) angina, recent destabilization of previously stable angina, or an acute coronary syndrome in the last 6 months were excluded. In addition, we excluded patients with any other cardiac and systemic disease. The study was approved by the Institutional Ethical Committee and written informed consent was obtained from each individual to participate in the study.
Heart rate turbulence All subjects underwent 24 h Holter ECG monitoring using three-channel tape recorders (Oxford Medilog FD5, Abingdon, UK). The recordings were analysed by two expert cardiologists using the system Oxford Excel 3.0. To calculate HRT parameters, RR intervals were exported as text files and analysed using a dedicated and validated software system.17 Heart rate turbulence was obtained from sequences of sinus RR intervals related to isolated PVCs that fulfilled the following criteria: (i) index PVC embedded into at least five preceding and 20 succeeding normal RR intervals; (ii) cycle length of all considered RR intervals .300 ms but ,2000 ms; (iii) beat-to-beat differences ,200 ms; and (iv) differences ,20% from the average of five preceding intervals.1,5 Accordingly, a single tachogram for HRT analysis consisted of a sequence of five RR intervals preceding and 20 RR intervals succeeding a PVC. Interpolated PVCs were excluded. Heart rate turbulence was assessed using the two standard parameters TO and TS. Turbulence onset, expressed as a percentage, was calculated using the formula [(RR1 + RR2) 2 (RR22 + RR21)]/(RR22 + RR21)×100, where RR1 and RR2 were the first and the second sinus RR intervals after the PVC and RR21 and RR22 were the first and the second sinus RR intervals preceding the PVC. Turbulence slope, expressed in ms/RR, was obtained as the maximal positive slope of all slopes of a series of regression lines obtained from all sequences (n ¼ 16) of five consecutive RR intervals included between the 1st and 20th RR interval after the compensatory pause that followed the PVC. Turbulence onset was calculated as the average of all TOs measured in the 24 h on individual tachograms of index PVCs, while TS was calculated on the averaged tachogram obtained averaging the RR intervals of individual tachograms preceding and following the index PVCs in the 24 h. Although ‘normal’ HRT values have not been defined yet, normal TO is expected to be negative, whereas normal TS is expected to be positive. Previous studies showed that TO ≥ 0% and TS ≤ 2.5 ms/RR were significantly associated with a worse prognosis in cardiac patients.1,8 – 11 Accordingly, we used these dichotomized values as cut-offs to consider HRT parameters as abnormal. A composite HRT variable was also obtained by combining TO and TS values, which results in three mutually exclusive categories: (i) TO ≤ 0% and TS ≥ 2.5 ms/RR interval (i.e. both parameters normal, category 0); (ii) either TO ≥ 0% or TS ≤ 2.5 ms/RR interval (i.e. only one parameter abnormal, category 1); and (iii) TO ≥ 0% and TS ≤ 2.5 ms/RR interval (i.e. both parameters abnormal, category 2).
Statistics Statistical analyses were done by SPSS 20.0 (SPSS, Inc.). Turbulence onset showed a Gaussian distribution, while TS showed a skewed distribution, as assessed by Kolmogorov –Smirnov test. Accordingly, between-group values of TO and TS were compared using unpaired t-test and Mann– Whitney test, respectively. Proportions were compared using x2 test. Pearson and Spearman rank-correlation tests, respectively, were used to assess the relation between HRT parameters with other variables.
Downloaded from by guest on June 3, 2015
Methods
G. Pinnacchio et al.
Page 3 of 7
Determinants of heart rate turbulence
Univariate and multivariable linear regression analysis was performed to identify variables predictive of HRT parameters, considered as continuous variables. Univariate and multivariable logistic regression was applied to identify variables associated with abnormal TO and TS values, as well as HRT category 1 or category 2, compared with category 0. All variables with a P-value≤ 0.1 on univariate analyses were included as independent covariates in the multivariable models. Furthermore, we also assessed differences in HRT parameters by performing comparisons between two subgroups of 65 subjects of groups 1 and 2, which were matched for age and gender. Data are reported as mean + SD for normally distributed variables and as median (with interquartile interval) for variables with skewed distribution. Differences were considered significant for P , 0.05.
Results
to an insufficient number of PVCs (n ¼ 287) or a history of diabetes (n ¼ 23). Thus, 209 individuals (age 61 + 16 years, males 49%) satisfied the inclusion criteria and formed the group without any apparent cardiac disease (group 1). In the same period, 392 patients with heart disease were screened. Overall, 235 patients were excluded, due to an insufficient number of PVCs (n ¼ 155) or evidence of a non-ischaemic disease of the heart (n ¼ 80). Thus, 157 patients (age 69 + 10 years, males 78%) met the inclusion criteria for the study and constituted the group of stable CAD patients (group 2). The main clinical characteristics of the two study groups are summarized in Table 1. Compared with group 2 patients, subjects of group 1 were younger, less frequently men, had a lower prevalence of CVRFs, and had a higher LVEF. Furthermore, they also had a lower number of PVCs.
General findings During the period of the study, 519 subjects without any apparent cardiac disease were examined, 310 of whom were excluded due
Table 1 Main clinical characteristics of subjects included in the study Group 1 (n 5 209)
Group 2 (n 5 157)
P-value
61 + 16 49
69 + 10 78
,0.001 ,0.001
1111 + 1942 149 + 186
1760 + 3445 170 + 202
0.02 0.32
70 + 10
67 + 9
0.005
1
19
,0.001
Family history of CVD (%)
32
54
,0.001
Hypertension (%)
54
80
,0.001
Dyslipidemia (%) Smoking (%)
38 17
65 13
,0.001 0.382
b-Blockers (%) ACE-inhibitors (%)
32 23
77 56
,0.001 ,0.001
Statins (%)
16
................................................................................ Age (years) Men (%) Holter data PVCs (Nr/24 h) Processed PVCs (Nr/24 h) Mean heart rate (b.p.m.) NSVT (%) CVRFs
Drug therapy
Diuretics (%) 20 Reason for Holter recording
77
,0.001
41
,0.001
Follow-up (%)
35
60
Palpitations (%) Chest pain (%)
32 12
12 13
Syncope (%)
14
10
7 62 + 5
5 51 + 12
Other (%) LVEF (%)
,0.001
,0.001
PVCs, premature ventricular complexes; NSVT, non-sustained ventricular tachycardia; CVRFs, cardiovascular risk factors; CVD, cardiovascular disease; ACE, angiotensin converting enzyme; LVEF, left ventricular ejection fraction.
Premature ventricular complexes suitable for HRT assessment in group 1 subjects were 149 + 186. Turbulence onset, in this group, was 21.80 + 2.24%, while TS was 5.83 (3.25– 10.55) ms/RR. At univariate analysis, a highly significant correlation was found between age and both HRT parameters (TO: r ¼ 0.38, P , 0.001; TS: r ¼ 20.63, P , 0.001; Figure 1). Furthermore, higher TO values and lower TS values were observed in subjects with, compared with those without, hypertension or dyslipidemia; female gender was also associated with lower TS values, compared with males (Table 2). At multivariable analysis, however, the only independent predictor of TO was age (OR, 1.05; CI: 1.02 – 1.07; P , 0.001), and the only independent predictors of TS were age (OR, 0.85; CI: 0.81 – 0.88; P , 0.001) and hypertension (OR, 0.19; CI: 0.04 – 0.83; P ¼ 0.028).
Heart rate turbulence in coronary artery disease patients In group 2 patients, the number of PVCs suitable of HRT analysis was 170 + 202. Turbulence onset in this group was 20.73 + 1.61%, while TS was 2.93 (1.73–5.81) ms/RR. At univariate analysis, a significant, although modest, correlation was found on both HRT parameters with age (r ¼ 0.24, P ¼ 0.002 and r ¼ 20.34, P , 0.001, for TO and TS, respectively; Figure 2), LVEF (r ¼ 20.23, P ¼ 0.012 and r ¼ 0.43, P , 0.001, for TO and TS, respectively; Figure 3), and, for TS only, heart rate (r ¼ 20.24, P ¼ 0.01). Furthermore, higher TO values were observed in patients with a history of previous coronary artery by-pass surgery (P ¼ 0.036) and in those treated with diuretic or antiarrhythmic drugs. Lower TS values were instead found in females and in patients with diabetes or hypertension (Table 3). At multivariable analysis, the only independent predictor of TO was age (OR, 1.03; CI: 1.002 –1.06; P ¼ 0.038), whereas independent predictors of TS were age (OR, 0.90; CI: 0.84 –0.95; P ¼ 0.001) and LVEF (OR, 1.07; CI: 1.02–1.13; P ¼ 0.008).
Downloaded from by guest on June 3, 2015
Variable
Heart rate turbulence in subjects without cardiac disease
Page 4 of 7
G. Pinnacchio et al.
24
6
r = 0.38; P < 0.001
4
18 TS (ms/RR)
2 To (%)
r = –0.63; P < 0.001
21
0 –2 –4
15 12 9
–6
6
–8
3 0 10
–10 10 20 30 40 50 60 70 80 90 100 Age
20
30
40
50 60 Age
70
80
90 100
Figure 1 Correlation of age with TO (left) and TS (right) in individuals without any apparent heart disease.
Table 2 Clinical predictors of TO and TS in group 1 subjects at univariate linear regression analysis TO
P-value
TS
P-value
Age Male gender
1.05 (1.034–1.071) 0.82 (0.44– 1.51)
,0.001 0.52
0.83 (0.80– 86) 6.47 (1.54– 27)
,0.001 0.011
Mean heart rate
0.99 (0.96– 1.03)
0.77
0.94 (0.88– 1.01)
0.068
CVRFs Hypertension
...............................................................................................................................................................................
0.001
0.01 (0.01– 0.04)
,0.001
2.14 (1.15– 3.99)
0.017
0.12 (0.03– 0.50)
0.004
Smoke Obesity
0.57 (0.25– 1.27) 1.03 (0.66– 1.59)
0.169 0.882
5.39 (0.79– 36) 0.52 (0.19– 1.45)
0.084 0.213
b-Blockers Ca2+-antagonists
1.34 (0.69– 2.58) 1.86 (0.82– 4.22)
0.381 0.135
0.19 (0.04– 0.92) 0.05 (0.01– 0.30)
0.040 0.002
ACE-inhibitors
1.87 (0.91– 3.88)
0.088
0.04 (0.01– 0.22)
,0.001
Statins Diuretics
1.73 (0.76– 3.91) 2.10 (0.99– 4.47)
0.190 0.052
0.04 (0.01– 0.24) 0.06 (0.01– 0.37)
0.001 0.002
0.96 (0.89– 1.03)
0.238
1.13 (0.94– 1.35)
0.194
Therapy
LVEF
CVRFs, cardiovascular risk factors; ACE, angiotensin converting enzyme; LVEF, left ventricular ejection fraction.
21
6 r = 0.24; P < 0.002 TS (ms/RR)
2 To (%)
r = –0.34; P < 0.001
18
4
0 –2 –4
15 12 9 6 3
–6
0 30
40
50
60 70 Age
80
90
30
40
50
60 70 Age
Figure 2 Correlation of age with TO (left) and TS (right) in patients with stable coronary artery disease.
80
90
Downloaded from by guest on June 3, 2015
2.78 (1.52– 5.05)
Dyslipidemia
Page 5 of 7
Determinants of heart rate turbulence
18
6
r = 0.43; P < 0.001
4
15
2
12
TS (ms/RR)
To (%)
r = –0.23; P = 0.012
0 –2 –4
9 6 3
–6
0 10
30
50 LVEF
70
10
30
50
70
LVEF
Figure 3 Correlation of LVEF with TO (left) and TS (right) in patients with stable coronary artery disease.
Table 3 Clinical predictors of TO and TS in group 2 patients at univariate linear regression analysis TO
P-value
TS
P-value
Age
1.04 (1.01–1.06)
0.002
0.85 (0.81–0.90)
,0.001
Male gender Mean heart rate
0.64 (0.35–1.18) 1.01 (0.98–1.04)
0.15 0.35
5.12 (1.15–22.8) 0.87 (0.82–0.92)
0.03 0.01
Diabetes Hypertension
1.54 (0.91–2.59) 1.48 (0.79–2.77)
0.10 0.22
0.15 (0.04–0.52) 0.06 (0.02–0.27)
0.003 ,0.001
Dyslipidemia
0.79 (0.46–1.35)
0.39
1.34 (0.36–4.99)
0.65
Family history of CVD Smoke
0.87 (0.52–1.45) 0.96 (0.45–2.02)
0.59 0.91
0.94 (0.27–3.29) 0.69 (0.11–4.29)
0.92 0.68
Obesity
0.94 (0.67–1.31)
0.70
1.43 (0.62–3.27)
0.39
b-Blockers Ca2+-antagonists
1.02 (0.55–1.86) 1.63 (0.88–3.01)
0.96 0.11
3.30 (0.76–14.3) 0.18 (0.04–0.81)
0.11 0.02
ACE-inhibitors
0.76 (0.46–1.26)
0.29
1.96 (0.56–6.86)
0.29
ARA Statins
1.74 (0.97–3.13) 0.75 (0.41–1.37)
0.06 0.35
0.21 (0.05–0.86) 1.96 (0.45–8.63)
0.03 0.37
...............................................................................................................................................................................
Therapy
Anti-platelets
1.39 (0.37–5.23)
0.62
3.91 (0.15–99)
0.41
Digoxin Nitrates
6.78 (0.71–64) 1.33 (0.65–2.73)
0.09 0.43
0.07 (0.01–19) 0.40 (0.07–2.32)
0.35 0.30
Antiarrhythmics
2.62 (1.22–5.64)
0.014
0.20 (0.03–1.34)
0.09
Diuretics LVEF
1.87 (1.12–3.04) 0.97 (0.95–0.99)
0.016 0.012
0.17 (0.05–0.59) 1.10 (1.04–1.16)
0.006 0.001
Previous AMI
0.80 (0.43–1.50)
0.48
1.52 (0.32–7.13)
0.59
Previous PCI Previous CABG
0.68 (0.41–1.13) 1.75 (1.04–2.95)
0.14 0.036
1.22 (0.34–4.33) 0.35 (0.10–1.28)
0.75 0.11
CVRFs, cardiovascular risk factors; CVD, cardiovascular disease; ACE, angiotensin converting enzyme; ARA, aldosterone receptor antagonists; LVEF, left ventricular ejection fraction; AMI, acute myocardial infarction; PCI, percutaneous coronary intervention; CABG, coronary artery bypass graft.
Between-group comparisons Values of both HRT parameters were significantly different in group 1 and group 2 patients, being worse in the latter (P , 0.001 for both TO and TS).
An abnormal TO (.0%) was found in 26 group 1 subjects (12.4%) and 42 group 2 patients (26.8%; P ¼ 0.001), whereas an abnormal TS (,2.5 ms/RR) was found in 31 group 1 subjects (14.8%) and 64 group 2 patients (40.8%; P ¼ 0.001), respectively. Furthermore, HRT
Downloaded from by guest on June 3, 2015
CVRFs
Page 6 of 7
G. Pinnacchio et al.
categories 0, 1, and 2 were detected in 160 (76.6%), 41 (19.6%), and 8 (3.8%) subjects of group 1, but in 78 (49.7%), 52 (33.1%), and 27 (17.2%) patients of group 2, respectively (P , 0.001). All these findings remained unchanged when only patients with frequent PVCs (≥10/h) were included in the analyses (data not shown). In multivariable linear regression analysis, presence of CAD (group 2) showed an independent association with TO (OR, 2.23; P ¼ 0.02), but not with TS (OR, 0.40; P ¼ 0.18). Moreover, in multivariable logistic regression analysis, the variable ‘group’ did not show a significant association with both ‘abnormal’ TO and TS values and with HRT category 1 or 2, whereas age (P , 0.001) and LVEF (P , 0.001) remained the only independent predictor of HRT parameters.
Comparison of matched groups Two samples of 65 subjects of the two groups could be matched for age and gender. As shown in Table 4, the two groups were also comparable with regard to CVRFs, whereas HRT parameters were significantly different between the two groups. In multivariable linear regression analysis, CAD group showed an independent association with TO (OR, 2.43; P ¼ 0.028), but not with TS (OR, 0.31; P ¼ 0.22). However, age (P , 0.05) and LVEF
Variable
Group 1 (n 5 65)
Group 2 (n 5 65)
P-value
Age (years)
67 + 6
67 + 9
1
Men (%) Holter data
50
50
1
PVCs (Nr/24 h)
1120 + 1969
1246 + 1816
0.70
Processed PVCs (Nr/24 h)
150 + 156
178 + 218
0.43
Mean heart rate (b.p.m.)
70 + 9
69 + 9
0.40
1
16
0.004
................................................................................
NSVT (%) CVRFs Diabetes (%)
0
0
1
Hypertension (%) Dyslipidemia (%)
69 58
77 72
0.42 0.13
Smoking (%)
19
19
1
Drug therapy b-Blockers (%)
42
76
ACE-inhibitors (%)
35
48
0.15
Statins (%)
28
77
,0.001
41 61 + 6
50 50 + 14
0.37 ,0.001
21.24 + 2.2 5.08 (2.95– 7.30)
–0.53 + 1.2 2.18 (1.19– 4.64)
0.03 ,0.001
Diuretics (%) LVEF (%)
,0.001
HRT parameters TO (%) TS (ms/RR)
PVCs, premature ventricular complexes; NSVT, non-sustained ventricular tachycardia; CVRFs, cardiovascular risk factors; ACE, angiotensin converting enzyme; LVEF, left ventricular ejection fraction; HRT, heart rate turbulence; TO, turbulence onset; TS, turbulence slope.
Discussion Two main findings emerge from our results. First, our data showed that, in subjects without any apparent cardiac disease, only age and, in part, hypertension significantly influence in an independent way HRT parameters. Second, age remains a major modulator of HRT in CAD patients, in whom LVEF also assumes a major role. Heart rate turbulence is an autonomically mediated phenomenon with a close relationship to baroreflex mechanisms. Premature ventricular complexes, indeed, produce a fall in blood pressure, which results in vagal withdrawal and sympathetic activation, with an initial acceleration of sinus node activity and return of blood pressure to normal values; this in turn results in baroreceptor stimulation with vagal recruitment and sympathetic inhibition, thus leading to a later deceleration of sinus node activity.2 Increasing age has consistently been shown to be associated with a decrease in cardiac vagal modulation, using various measures of cardiac autonomic function, as basal heart rate and heart rate variability, which increase and decrease with age, respectively.19 Accordingly, some previous studies, in restricted populations of subjects, showed that age was a significant determinant of HRT values.15 – 17 These findings are confirmed in our study, in which age was the most powerful independent predictor of HRT in subjects without cardiac disease, with increasing age being associated with progressively higher TO and lower TS values, respectively. Furthermore, our study is the first to show that HRT is largely independent of CVRFs, as well as medical therapy, in apparently healthy individuals. Until now, the relation between CVRFs and HRT parameters in healthy people had indeed poorly been investigated, with the few studies including only small samples of volunteers.15 – 18 Our study shows that only hypertension has an independent, although weak, association with abnormal HRT, in agreement with the wellknown impairment of baroreflex sensitivity observed in hypertensive patients.20 Of note, hypertension independently influenced TS but not TO, suggesting that the abnormality related to hypertension resides in the reflex mechanism that follows an increased, rather than in the one that follows a decreased, stimulation of baroceptors. A further main finding of our study is that an impaired LVEF is the only other major determinant of HRT in stable CAD patients. The latter result is not surprising, as a reduced LV function has consistently been found to be a major determinant of sympatho-vagal imbalance in several clinical settings, including acute and old MI and dilated cardiomyopathy,21,22 and a relation of HRT with LVEF was already shown in our previous study.17 Of note, the presence of a previous acute MI was instead by itself insufficient to significantly influence HRT values, suggesting that it may exert its influence on HRT only when it is large enough to impair LV function. Similarly, although HRT showed significantly different values in CAD compared with healthy subjects, CAD presence was only a weak predictor of HRT, being independently associated only with TO, when considered as a continuous variable, in multivariable analyses. Of note, the poor relation of obstructive CAD by itself with HRT parameters was confirmed when comparing two subgroups
Downloaded from by guest on June 3, 2015
Table 4 Main clinical characteristics of matched groups
(P , 0.05), but not CAD group, were independent predictors of abnormal TO and TS values and of HRT category 1 or 2 in these matched selected individuals.
Page 7 of 7
Determinants of heart rate turbulence
of non-cardiac subjects and of CAD patients selected for being similar in age and gender. In the latter subgroup differences were again found in HRT parameters, but statistical significance was lost when corrected for LVEF, in agreement with the independent association of LVEF with HRT parameters in the whole sample of CAD patients. Some limitations of our study should be acknowledged. First, we cannot be sure that all apparently ‘healthy’ individuals did not have any form of CAD, as most of them had some CVRF. However, the lack of any symptom and the evidence of normal ECG, echocardiography, and maximal exercise stress test makes unlikely that a sizeable proportion had significant CAD. Second, several subjects were being treated with cardiologic drugs, in particular b-blockers, which might have influenced HRT. However, drugs did not emerge as correlated with HRT in multivariable analyses, suggesting that their influence is, if any, limited. Our finding is, in fact, in agreement with the lack of significant effect of beta-blockade on HRT parameters found in previous studies of apparently healthy individuals.4
Conclusions
7.
8.
9.
10. 11.
12.
13.
14.
15.
16.
17.
Conflict of interest: none declared.
References 1. Schmidt G, Malik M, Barthel P, Schneider R, Ulm K, Rolnitzky L et al. Heart rate turbulence after ventricular premature beats as a predictor of mortality after acute myocardial infarction. Lancet 1999;353:1390 –96. 2. Wichterle D, Melenovsky V, Simek J, Malik J, Malik M. Hemodynamics and autonomic control of heart rate turbulence. J Cardiovasc Electrophysiol 2006;17:286 – 91. 3. Welch WJ, Smith ML, Rea RF, Bauernfeind RA, Eckberg DL. Enhancement of sympathetic nerve activity by single premature ventricular beats in humans. J Am Coll Cardiol 1989;13:69–75. 4. Lin LY, Lai LP, Lin JL, Du CC, Shau WY, Chan HL et al. Tight mechanism correlation between heart rate turbulence and baroreflex sensitivity: sequential autonomic blockade analysis. J Cardiovasc Electrophysiol 2002;13:427 –31. 5. Bauer A, Malik M, Schmidt G, Barthel P, Bonnemeier H, Cygankiewicz I et al. Heart rate turbulence: standards of measurement, physiological interpretation, and clinical
18.
19.
20. 21.
22.
Downloaded from by guest on June 3, 2015
In conclusion, our study shows that age is a major determinant of HRT values both in subjects without any apparent cardiac disease and in CAD patients. Only hypertension, among CVRFs, contributes significantly, although in a moderate way, to HRT in people without cardiac disease, whereas LVEF becomes a major predictor of HRT in CAD patients.
6.
use. International society for Holter and non invasive electrophysiology consensus. J Am Coll Cardiol 2008;52:1353 –65. Davies LC, Francis DP, Ponikowski P, Piepoli MF, Coats AJ. Relation of heart rate and blood pressure turbulence following premature ventricular complexes to baroreflex sensitivity in chronic congestive heart failure. Am J Cardiol 2001;87:737 – 42. Iwasaki M, Yuasa F, Yuyama R, Mimura J, Kawamura A, Motohiro M et al. Correlation of heart rate turbulence with sympathovagal balance in patients with acute myocardial infarction. Clin Exp Hypertens 2005;27:251 – 7. La Rovere MT, Bigger JT Jr, Marcus FI, Mortara A, Schwartz PJ. Baroreflex sensitivity and heart-rate variability in prediction of total cardiac mortality after myocardial infarction. Lancet 1998;351:478 –84. Barthel P, Schneider R, Bauer A, Ulm K, Schmitt C, Scho¨mig A et al. Risk stratification after acute myocardial infarction by heart rate turbulence. Circulation 2003;108: 1221 –26. Multicenter Post-Infarction Research Group. Risk stratification and survival after myocardial infarction. N Engl J Med 1983;309:331 –6. Julian DG, Camm AJ, Frangin G, Janse MJ, Munoz A, Schwartz PJ et al. Randomised trial of effect of amiodarone on mortality in patients with left-ventricular dysfunction after recent myocardial infarction: EMIAT. European Myocardial Infarct Amiodarone Trial Investigators. Lancet 1997;349:667 –74. Sulimov V, Okisheva E, Tsaregorodtsev D. Non-invasive risk stratification for sudden cardiac death by heart rate turbulence and microvolt T-wave alternans in patients after myocardial infarction. Europace 2012;14:1786 –92. Grimm W, Schmidt G, Maisch B, Sharkova J, Mu¨ller HH, Christ M. Prognostic significance of heart rate turbulence following ventricular premature beats in patients with idiopathic dilated cardiomyopathy. J Cardiovasc Electrophysiol 2003;14:819 –24. Marynissen T, Flore´ V, Heidbuchel H, Nuyens D, Ector J, Willems R. Heart rate turbulence predicts ICD-resistant mortality in ischaemic heart disease. Europace 2014;16:1069 – 77. Schwab JO, Eichner G, Shlevkov N, Schrickel J, Yang A, Balta O et al. Impact of age and basic heart rate on heart rate turbulence in healthy persons. PACE 2005;28: S198 –201. Grimm W, Sharkova J, Christ M, Schneider R, Schmidt G, Maisch B. Heart rate turbulence following ventricular premature beats in healthy controls. Ann Noninvasive Electrocardiol 2003;8:127 –31. Sestito A, Valsecchi S, Infusino F, Sgueglia GA, Bellocci F, Zecchi P et al. Differences in heart rate turbulence between patients with coronary artery disease and patients with ventricular arrhythmias but structurally normal hearts. Am J Cardiol 2004;93: 1114 –8. Lindgren KS, Makikallio TH, Seppanen T, Raatikainen MJ, Castellanos A, Myerburg RJ et al. Heart rate turbulence after ventricular and atrial premature beats in subjects without structural heart disease. J Cardiovasc Electrophysiol 2003;14:447 –52. Bonnemeier H, Richardt G, Potratz J, Wiegand UK, Brandes A, Kluge N et al. Circadian profile of cardiac autonomic nervous modulation in healthy subjects: differing effects of aging and gender on heart rate variability. J Cardiovasc Electrophysiol 2003; 14:791–9. Mussalo H, Vanninen E, Ika¨heimo R, Laitinen T, Laakso M, La¨nsimies E et al. Baroreflex sensitivity in essential and secondary hypertension. Clin Auton Res 2002;12:465 – 71. Graham LN, Smith PA, Stoker JB, Mackintosh AF, Mary DA. Time course of sympathetic neural hyperactivity after uncomplicated acute myocardial infarction. Circulation 2002;106:793 –7. Ajiki K, Murakawa Y, Yanagisawa-Miwa A, Usui M, Yamashita T, Oikawa N et al. Autonomic nervous system activity in idiopathic dilated cardiomyopathy and in hypertrophic cardiomyopathy. Am J Cardiol 1993;71:1316 –20.
