Escolar Documentos
Profissional Documentos
Cultura Documentos
Heart Rate Turbulence in Post-MI Patients with DM. Background: Previous studies have
described the clinical utility of heart rate turbulence (HRT) as an autonomic predictor in risk-stratifying
patients after myocardial infarction (MI). Some reports showed that diabetes mellitus (DM) affects the
prognostic value of autonomic markers. We assessed the utility of HRT as a risk marker in post-MI patients
with DM and without DM.
Methods: We prospectively enrolled 231 consecutive DM patients and 300 non-DM patients after acute
MI. HRT was measured using an algorithm based on 24-hour Holter electrocardiograms (ECGs), assessing
2 parameters: turbulence onset (TO) and turbulence slope (TS). HRT was considered positive when both
TO 0% and TS 2.5 ms/R-R interval were met. The endpoint was defined as cardiac mortality.
Results: Of patients with DM, 9 patients (4%) were not utilized for HRT assessment because of frequent
ventricular contractions or presence of atrial fibrillation. Forty-two of 222 patients (19%) were HRT
positive. During follow-up of 876 424 days, 26 patients (22%) reached the endpoint. Several factors
including left ventricular ejection fraction (LVEF), renal dysfunction, documentation of nonsustained
ventricular tachycardia (VT), and a HRT-positive outcome had significant association with the endpoint.
Multivariate analysis determined that renal dysfunction and a positive HRT outcome had significant value
with a hazard ratio (HR) of 4.7 (95%CI, 1.911.5; P = 0.0008) and 3.5 (95%CI, 1.48.8; P = 0.007),
respectively. In non-DM patients, only a positive HRT outcome had significant value.
Conclusions: This study reveals that HRT detected by 24-hour Holter ECG can predict cardiac mortality
in post-MI patients whether DM is present or not. (J Cardiovasc Electrophysiol, Vol. 22, pp. 1135-1140,
October 2011)
cardiac mortality, diabetes mellitus, heart rate turbulence, myocardial infarction, ventricular tachycardia
Background
Numerous reports have described risk stratification techniques for serious cardiac events in patients following myocardial infarction (MI). At present, a reduced left ventricular ejection fraction (LVEF),1 induction of sustained ventricular tachyarrhythmias by electrophysiologic testing,2 and
the presence of nonsustained ventricular tachycardia (VT)
on ambulatory electrocardiogram (ECG)3 have been widely
used for identifying patients at risk for cardiac mortality
and/or arrhythmic events in post-MI patients. Heart rate turbulence (HRT)4 is a measure of autonomic response to perThis manuscript was supported in part by a Grant-in-Aid (21590909) for
Scientific Research from the Ministry of Education, Culture, Sports, Science
and Technology of Japan and by a grant for clinical research from Kyorin
University School of Medicine (Dr. Ikeda).
No disclosures.
Address for correspondence: Takanori Ikeda, M.D., Ph.D., F.A.C.C.,
Department of Cardiovascular Medicine, Toho University Medical Center,
6-11-1 Omorinishi, Ota-ku, Tokyo 143-8541, Japan. Fax: +81-3-3766-7810;
E-mail: ikety3@gmail.com
Manuscript received 11 January 2011; Revised manuscript received 1 March
2011; Accepted for publication 22 March 2011.
doi: 10.1111/j.1540-8167.2011.02082.x
Methods
Patient Population
Between July 2007 and December 2009 we prospectively
enrolled 231 consecutive patients (age 71 11 years, 175
male) with a diagnosis of acute MI and underlying DM.
Patients were excluded from HRT assessment if they had
chronic or persistent atrial tachyarrhythmias, an implanted
permanent pacemaker, cardiac resynchronization therapy
1136
TABLE 1
Baseline Characteristics of Post-MI Patients With and Without DM
No. of Study
Population
DM Patients
(n = 231)
Non-DM
Patients
(n = 300)
71 11
175/56
46 11
231 (100%)
32 (14%)
193 (84%)
140 (61%)
68 13
230/70
48 11
0 (0%)
18 (6%)
211 (70%)
187 (62%)
68 (29%)
89 (39%)
11(5%)
226 (98%)
191 (83%)
168 (73%)
173 (75%)
25 (11%)
6 (3%)
0 (0%)
0 (0%)
0 (0%)
277 (92%)
205 (68%)
201 (67%)
256 (85%)
25 (8%)
5 (2%)
P Value
0.003
NS
0.03
0.002
0.0004
NS
0.005
0.0002
NS
0.003
NS
NS
Miwa et al.
The endpoint was prospectively defined as cardiac mortality. In patients who died, the causes were verified from
the hospital and autopsy records, and from either the primary
physicians or those who had witnessed the death. Patients
who died of noncardiac causes such as stroke and cancer
were not included in the endpoint and were excluded from
analysis. In patients with an ICD, an appropriate episode of
defibrillation therapy was included in the endpoint. In other
words, antitachycardia pacing or cardioversion for sustained
tachyarrhythmias was not included. The records were verified from monitoring ECGs taken in an emergency room,
ambulatory ECGs captured out-of-hospital, or by checking
the ICD memory.
1137
Statistical Analysis
Numeric data were expressed as mean standard deviation. Comparisons of clinical variables in HRT-positive and
-negative patients were evaluated using a chi-square test. For
analysis of the association between the endpoints and the
clinical factors, univariate and multivariate Cox regression
analyses were performed. Results of mortality-free analyses
were presented with hazard ratios (HR) and 95% confidence
intervals (CI). Sensitivity, specificity, positive and negative
predictive values, and the predictive accuracy of a mortalityfree prediction were also evaluated. A difference in mortalityfree rates was shown using the KaplanMeier method and the
log rank test. A P value of < 0.05 was considered statistically
significant.
Results
Outcome of HRT and Other Risk Factors
Although HRT measurements were performed on 24-hour
Holter ECGs in all 231 DM patients and 300 non-DM patients, values obtained from 9 patients (4%) of DM group, and
9 patients (3%) of non-DM group were not utilized for the assessment because they had frequent VPCs (3 and 4 patients,
respectively) such as bigeminy or trigeminy or paroxysmal
atrial fibrillation (6 and 5 patients, respectively). Therefore,
clinical data from 231 DM patients and 291 non-DM patients
were assessed.
DM patients
Average values for TO and TS were 0.16 1.91% (abnormal in 73 patients [33%]) and 4.22 5.67 ms/RR interval
(abnormal in 95 patients [43%]), respectively. According to
these data, HRT was determined as positive in 42 patients
(19%) and negative in 180 patients (81%). Fifty-nine patients had an LVEF 40% (27%). The average SDNN was
67.8 14.2 ms and nonsustained VT was documented on
Holter ECGs in 41 patients (18%).
Non-DM patients
Average values for TO and TS were 0.35 2.42% (abnormal in 82 patients [28%]) and 5.81 6.18 ms/RR interval
(abnormal in 71 patients [24%]), respectively. According to
these data, HRT was determined as positive in 32 patients
(11%) and negative in 259 patients (89%). Thirty-nine patients had an LVEF 40% (20%). The average SDNN was
100.1 42.0 ms and nonsustained VT was documented on
Holter ECGs in 46 patients (16%).
1138
TABLE 2
Comparison of Clinical Features in HRT-Positive and -Negative Patients in Both Groups
DM Patients
Variables
Median age (years)
Age > 70 years
Gender (male / female)
Mean LVEF (%)
LVEF < 40%
Hypertension
Hypercholesterolemia
Renal dysfunction
ICD implantation
Medical treatment
Insulin
Antidiabetic drugs
Insulin + antidiabetic drugs
Aspirin
-blockers
Statins
ACE-inhibitors/ARB
Class III antiarrhythmic drugs
Holter ECG findings
No. of PVC
Document of nonsustained VT
Mean SDNN
Positive HRV
HRT parameters
Median TO (%)
Median TS (ms/RRI)
Non-DM Patients
HRT Positive
(n = 42)
HRT Negative
(n = 180)
P Value
HRT Positive
(n = 32)
HRT Negative
(n = 259)
P Value
75 9
29 (69%)
32/10
42.8 12.0
24 (57%)
39 (93%)
24 (57%)
22 (52%)
1 (2%)
70 11
99 (55%)
134/46
47.4 10.7
35 (19%)
145 (81%)
112 (62%)
29 (16%)
1 (1%)
0.01
NS
NS
0.016
<0.0001
NS
NS
<0.0001
NS
73 12
22 (69%)
19/13
43.2 14.7
13 (41%)
25 (78%)
18 (56%)
3 (9%)
1 (3%)
67 13
88 (34%)
206/53
49.1 10.8
36 (14%)
179 (69%)
163 (63%)
14 (5%)
3 (1%)
0.01
0.0001
0.01
0.006
0.0001
NS
NS
NS
NS
10 (24%)
11 (26%)
3 (7%)
41 (98%)
38 (90%)
29 (69%)
531 (74%)
8 (19%)
57 (32%)
78 (43%)
8 (4%)
176 (98%)
144 (80%)
134 (74%)
137 (76%)
9 (5%)
NS
NS
NS
NS
NS
NS
NS
31 (97%)
28 (88%)
22 (69%)
27 (84%)
3 (9%)
239 (92%)
172 (66%)
174 (67%)
222 (86%)
20 (8%)
NS
0.015
0.03
0.04
NS
1,078
15 (36%)
66.4 10.2
25 (60%)
640
26 (14%)
73.7 10.2
58 (32%)
NS
0.002
0.003
0.001
742
7 (22%)
83.7 42.5
16 (50%)
675
39 (15%)
102.1 41.2
57 (22%)
NS
NS
0.02
0.0006
1.11 1.20
1.12 0.90
0.53 1.92
5.11 6.14
<0.0001
<0.0001
0.83 0.93
1.05 0.77
0.55 2.54
6.62 6.33
0.003
<0.0001
ECG = electrocardiogram; HRV = heart rate variability; PVC = premature ventricular complex; RRI = RR interval; SDNN = standard deviation of normal
to normal beat interval; TO = turbulence onset; TS = turbulence slope; VT = ventricular rachycardia.
Non-DM patients
On univariate analysis, median age, age > 70 years,
LVEF 40%, documentation of nonsustained VT, an HRVpositive outcome, and an HRT-positive outcome were associated with the endpoint (P = 0.017, P = 0.0009, P =
0.016, P = 0.011, P = 0.036, P < 0.0001, respectively). The
HR of an HRT-positive outcome was 31.3 (95% CI, 8.1125).
TABLE 3
Comparisons of Clinical Features in Patients Who Did or Did Not Meet the Endpoint in Both Groups
DM Patients
Variables
Median age (years)
Age >70 years
Gender (male)
Mean LVEF (%)
LVEF <40%
Hypertension
Hypercholesterolemia
Renal dysfunction
Holter ECG findings
PVC>3,000
Document of
nonsustained VT
HRV positive
HRT positive
Univariate
Analysis
1.1 (1.1-1.1)
2.6 (1.1-6.5)
P Value
7.3 (3.2-16.4)
0.021
0.04
0.52
<0.0001
0.026
0.18
0.37
<0.0001
3.2 (1.4-7.0)
0.27
0.004
6.9 (3.2-14.9)
0.14
<0.0001
1.1 (1.1-1.2)
2.4 (1.1-5.2)
Multivariatte
Analysis
4.7 (1.9-11.5)
3.5 (1.4-8.8)
Non-DM Patients
P Value
Univariate
Analysis
0.13
0.14
1.1 (1.0-1.1)
7.7 (1.7-35.7)
0.56
0.60
4.3 (1.3-14.3)
0.0008
P Value
Multivariatte
Analysis
0.017
0.0091
0.088
0.30
0.016
0.15
0.27
0.27
0.16
4.6 (1.4-15.2)
0.063
0.011
0.007
3.6 (1.1-11.8)
31.3 (8.1-125)
0.036
<0.0001
P Value
0.054
0.065
0.19
0.21
22.7 (5.7-90.9)
0.27
<0.0001
Miwa et al.
1139
TABLE 4
Predictive Values Associating the Presence of HRT and Other Electrocardiographic Variables with the Endpoint in DM Patients
Sensitivity (%)
Specificity (%)
PPV (%)
NPV (%)
PA (%)
12/26 (46)
10/26 (38)
13/26 (50)
15/26 (58)
146/196 (75)
165/196 (84)
126/196 (64)
169/196 (86)
12/62 (19)
10/41 (24)
13/83 (16)
15/42 (36)
146/160 (91)
165/181 (91)
126/139 (91)
169/180 (94)
158/222 (71)
175/222 (79)
139/222 (64)
184/222 (83)
PA = predictive accuracy; PPV = positive predictive value; NPV = negative predictive value.
Figure 1. The association between the presence of HRT and cardiac mortality in post-MI patients with and without DM.
Serious ventricular tachyarrhythmias are one of the mechanisms responsible for cardiac mortality in patients after MI.
In the era of treatment of fatal ventricular tachyarrhythmias
with an ICD, it is now possible to prevent sudden cardiac
death. Cost-efficient primary prevention of sudden cardiac
death using an ICD in post-MI patients requires risk stratification and identification of high-risk subgroups. As is well
known, MI patients with DM are at high risk of sudden cardiac death.8 DM not only causes coronary artery disease and
cardiac neuropathy,14,15 but also increases the risk of sudden
cardiac death in post-MI patients.
It is well known that heart rate measurements such as HRV,
which reflects autonomic imbalance, have been introduced
as a useful technique in identifying patients at risk for cardiac
mortality, particularly in patients after MI.13 HRV has also
been introduced as a first-line tool for the diagnosis of cardiac autonomic neuropathy due to DM, in a recent statement
by the American Diabetes Association.14 However, Whang
et al. have shown that HRV has less prognostic value for
identifying patients at risk, if DM is present.9
As an alternative method for assessment of autonomic
activity, HRT has been introduced as a measure of the autonomic response to perturbations of arterial blood pressure
after single VPCs. Several prospective studies support the
contention that this marker is a significant predictor of cardiac mortality in patients after MI and in those with heart
failure.4-6,16,17
Several studies also demonstrated that impaired HRT was
found in diabetic patients with previous MI. Stein et al.18
showed that abnormal HRT predicted cardiovascular mortality in high-risk patients with acute MI and left ventricular
dysfunction, in whom 24% of the patients had DM. Jeron
et al.19 revealed a higher percentage of abnormal values for
both HRT parameters in patients with DM than without DM
in subgroup analysis. Bauer et al.20 showed that in addition
to HRT, DM was also associated with late mortality after MI
in multivariate analysis. Barthel et al.21 showed that the combined assessment of HRT, LVEF 30%, age 65 years, and
DM can predict high-risk population at risk. Thus, abnormal
HRT is associated with DM in post-MI patients. HRT can
detect cardiac neuropathy due to DM. In patients with type 2
DM, Balcioglu et al.10 investigated the detection of diabetic
neuropathy by autonomic markers that included HRT. They
concluded that HRT was the most powerful index to detect
cardiac diabetic neuropathy, among other autonomic indexes
that included HRV.
In this study, we assessed whether HRT can detect highrisk MI patients with DM and without DM at risk for cardiac
mortality. We found that HRT was significantly associated
with cardiac mortality in post-MI patients whether DM is
present or not. HRT could be strong predictor of cardiac
1140