1135

Heart Rate Turbulence Can Predict Cardiac Mortality Following

Myocardial Infarction in Patients with Diabetes Mellitus
YOSUKE MIWA, M.D., MUTSUMI MIYAKOSHI, M.D., KYOKO HOSHIDA, M.D.,
RYOJI YANAGISAWA, M.D., ATSUKO ABE, M.D., TAKEHIRO TSUKADA, M.D.,
HARUHISA ISHIGURO, M.D., HISAAKI MERA, M.D., SATORU YUSU, M.D.,
HIDEAKI YOSHINO, M.D., and TAKANORI IKEDA, M.D.
From the Second Department of Internal Medicine, Kyorin University School of Medicine, Tokyo; and
Department of Cardiovascular Medicine, Toho University Medical Center, Tokyo

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).

turbations of arterial blood pressure after single ventricular

premature contractions (VPCs) and has been introduced as an
autonomic risk stratification marker for cardiac events after
MI or ischemic cardiomyopathy.4-7
Diabetes mellitus (DM) is a well-established risk factor
for ischemic heart disease. Prognosis after MI is worse in patients with DM as compared to patients without DM.8 Some
studies showed that DM diminishes the prognostic value of
autonomic markers such as heart rate variability (HRV).9 In
contrast, a recent study demonstrated that a reduced HRT is
associated with cardiac autonomic neuropathy due to DM.10
No prospective study has described the prognostic value of
HRT as a risk stratification marker for cardiac mortality in
post-MI patients with DM.
In this study, we prospectively assessed the utility of HRT,
measured from 24-hour Holter ECGs, in identifying the risk
of cardiac mortality in post-MI patients with DM.

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

Journal of Cardiovascular Electrophysiology

Vol. 22, No. 10, October 2011

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)

Median age (years)

Gender (male/female)
Mean LVEF (%)
Diabetes mellitus (type 2)
Renal dysfunction
Hypertension
Hypercholesterolemia
Medical treatment
Insulin
Antidiabetic drugs
Insulin + antidiabetic drugs
Aspirin
-blockers
Statins
ACE-inhibitors/ARB
Class III antiarrhythmic drugs
ICD implantation

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

ACE = angiotensin-converting enzyme; ARB = angiotensin receptor

blocker; DM = diabetes mellitus; ICD = implantable cardioverter
defibrillator; LVEF = left ventricular ejection fraction.

(biventricular pacing), or if they were in the acute phase

of MI (usually < 2 weeks after the onset). These exclusion
criteria were based on the inability to perform a HRT determination or reliable data analysis. In this study, we also included
300 consecutive post-MI patients (age 68 13 years, 230
male) without DM in the same study period as control subjects. Clinical characteristics of both post-MI patients with
and without DM are listed in Table 1. Informed consent was
obtained from each patient. The study was approved by the
Ethics Committee of our institute.
The diagnosis of acute MI was based on clinical course,
serum creatine kinase activity, and ST-segment elevation on
ECG. In DM patients, revascularization of 1 or more coronary
lesions in the acute phase consisted of percutaneous coronary
intervention in 214 patients (93%) and coronary artery bypass
graft in 17 patients (7%). One hundred and fifty-one patients
(65%), 57 patients (25%), and 23 patients (10%) had anterior,
inferior, and lateral wall MI, respectively. All patients had a
New York Heart Association (NYHA) functional class of I
or II at HRT assessment. Thus, most patients had relatively
preserved cardiac function in this study. The mean LVEF was
46 11%. In this study, revascularization procedure such as
percutaneous coronary intervention was performed before
enrollment in both DM and non-DM patients, if necessary.
After that, we recorded 24-hour Holter ECGs and acquired
HRT data.
As regards DM, all patients had type 2 DM for 1 year or
more. In other words, no patients had type 1 DM. Type 2 DM
was defined based on medical history, a fasting blood glucose 126 mg/dL and a HbA1c 6.5%, no prior episodes
of diabetic ketoacidosis, diagnosis of DM after the age of
30, and no need for insulin in the first 5 years after diagnosis.11 Of DM patients, 168 patients (83%) were on antidiabetic medication including insulin (68 patients [29%]),
oral antidiabetic drugs (89 patients [39%]), and both forms
of treatment (11 patients [5%]). Adjuvant medication consisted of aspirin in 226 patients (98%), -blockers in 191
patients (83%), angiotensin-converting enzyme (ACE) in-

hibitors or angiotensin II receptor blockers (ARBs) in 173

patients (75%), and class III antiarrhythmic drugs in 25 patients (11%). For enrollment of nondiabetic patients, fasting
glucose level and HbA1c level were measured in all. When
patients had both a fasting glucose level < 126 mg/dL and
a HbA1c level < 6.5%, they regard as nondiabetic patients.
This study included 6 patients (3%) with an implantable
cardioverter-defibrillator (ICD) as a means of primary prevention of sudden cardiac death, inserted at the request of
these patients.
Measurement of HRT
HRT was automatically measured using an algorithm applied to 24-hour Holter ECGs obtained using the MARS
Holter system (GE Healthcare Inc., Milwaukee, WI, USA).
HRT parameters included turbulence onset (TO) and turbulence slope (TS), which were determined according to a previously published method.4 Briefly, the TO was calculated as
the percentage change between the mean of the first 2 sinus
RR intervals after a VPC and the last 2 sinus rhythm RR
intervals before the VPC, as follows: TO = ([RR 1 +RR 2 ]
[RR 2 +RR 1 ]) / (RR 2 +RR 1 ), where RR 1 is the ith sinus
beat following (i > 0) the compensatory pause of the VPC
or preceding (i < 0) the coupling interval of the VPC. These
measurements were performed for each single VPC and subsequently averaged. The TS was calculated as the maximum
positive slope of a regression line assessed over any sequence
of 5 subsequent RR intervals within the tachogram RR 1 , RR 2 ,
RR 3 , . . ., RR 15 , where RR i is the average of the ith sinus RR
intervals after the compensatory pause of a singular VPC.
The TO and TS were dichotomized at predefined cut
points (TO = 0%, TS = 2.5 ms/RR interval).4 In this study,
patients were defined as HRT positive when both TO and TS
were abnormal (TO 0% and TS 2.5 ms/RR interval),
and as HRT negative when TO and/or TS were normal (TO
< 0% and/or TS > 2.5 ms/RR interval) or when HRT cannot
be calculated because of no or too few suitable VPCs.7,12
Assessment of Other Risk Factors
In addition to HRT, the following risk factors were
prospectively assessed: median age and gender of the patients, a history of hypertension, hypercholesterolemia, renal
dysfunction, LVEF, documentation of nonsustained VT, and
the standard deviation of N-N intervals (SDNN) on timedomain analysis of HRV.
LVEF was measured by the modified Simpson method using echocardiography or cardiac magnetic resonance imaging. In this study, the LVEF was dichotomized at a cut point
of 40% (i.e., LVEF 40% and LVEF > 40%). The presence
of nonsustained VT was determined from routine 24-hour
Holter ECGs with recording of daily activities, or detected
during ECG monitoring. Nonsustained VT was defined as
the documentation of 3 consecutive VPCs at a rate of 100
beats/min. HRV was assessed from 24-hour Holter ECGs.
Positive HRV was determined when the SDNN was less than
70 ms.13
Follow-Up and Study Endpoints
All patients were followed as outpatients at our institute.
Regular follow-up contact was obtained at a hospital visit
every 2 or 4 weeks. When patients did not visit for more than
8 weeks, our medical staff conducted telephonic interviews.

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.

Heart Rate Turbulence in Post-MI Patients with DM

1137

Comparisons of Clinical Variables between HRT-Positive

and -Negative Patients in Both Groups
The results are shown in Table 2. In patients with DM,
median age, mean LVEF, LVEF 40%, renal dysfunction,
documented presence of nonsustained VT, and positive HRV
attained significant values (P = 0.01, P = 0.016, P < 0.0001,
P < 0.0001, P = 0.002, and P = 0.001, respectively).
In non-DM patients, median age, age > 70 years, gender, mean LVEF, LVEF 40%, -blockers, statins, ACEinhibitor/ARB, mean SDNN, and positive HRV had significant values (P = 0.01, P = 0.0001, P = 0.01, P = 0.006,
P = 0.0001, P = 0.015, P = 0.03, P = 0.04, P = 0.02, and
P = 0.0006, respectively).

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%).

Events During Follow-Up

DM patients
During a mean follow-up of 876 424 days, cardiac death
occurred in 25 patients (11%; ventricular tachyarrhythmias
in 8, sudden death in 2, pump failure in 13, and cardiac rupture in 2 patients after a further MI). Six patients suffered
noncardiac death (i.e., stroke, lung cancer, colon cancer, thyroid cancer, interstitial pneumonia, and chronic lymphocytic
leukemia) and were excluded from analysis. In patients with
an ICD, 1 patient (1%) had defibrillation therapy for ventricular fibrillation and this event was included in the endpoint.
Therefore, a total of 26 patients (14%) reached the defined
endpoint.
Non-DM patients
During a mean follow-up of 885 358 days, cardiac death
occurred in 11 patients (4%; ventricular tachyarrhythmias in
2, pump failure in 8, and cardiac rupture in 1 patient after
a further MI). Three patients suffered noncardiac death (i.e.,
lung cancer, malignant lymphoma, septic shock) and were
excluded from analysis. Therefore, a total of 11 patients (5%)
reached the defined endpoint.
Cumulative incidence of arrhythmic death or sudden death
was 4.3% (n = 10) in DM patients. On the contrary, the
incidence was 0.7% (n = 2) in non-DM patients. This type
of death in DM patients was significantly more than that in
non-DM patients (P = 0.005).
Association Between Risk Variables and Endpoints
Comparisons of clinical features in patients who did or
did not meet the endpoint in both groups are shown in
Table 3.
DM patients
On univariate analysis, median age, age > 70 years, mean
LVEF, LVEF 40%, renal dysfunction, documentation of
nonsustained VT, and a HRT-positive outcome were significantly associated with the endpoint (P = 0.021, P = 0.04, P
< 0.0001, P = 0.026, P < 0.0001, P = 0.004, P < 0.0001,
respectively). The HR of an HRT-positive outcome was 6.9
(95% CI, 3.214.9). Other variables such as positive HRV
and the number of PVCs did not attain statistical significance. Figure 1A shows KaplanMeier event-free curves for
cardiac mortality according to HRT outcome in this study
population. To test which factor was the most significant,

1138

Journal of Cardiovascular Electrophysiology

Vol. 22, No. 10, October 2011

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.

we performed multivariate analysis. Renal dysfunction and

an HRT-positive outcome had significant value for the endpoint, with an HR of 4.7 (95% CI, 1.911.5; P = 0.0008) and
3.5 (95% CI, 1.48.8; P = 0.007). Predictive values associating the presence of HRT and other electrocardiographic
risk variables with the endpoint are shown in Table 4. The
best positive predictive and negative predictive values were
attained in HRT positive.

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)

CI = confidence interval; HR = hazard ratio.

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.

Heart Rate Turbulence in Post-MI Patients with DM

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)

LVEF < 40%

Document of nonsustained VT
HRV positive
HRT positive

PA = predictive accuracy; PPV = positive predictive value; NPV = negative predictive value.

Figure 1B shows KaplanMeier event-free curves for cardiac

mortality according to HRT outcome in this study population. On multivariate analysis, an HRT-positive outcome had
the most significant value for the endpoint, with an HR of
22.7 (95% CI, 5.7-90.9; P < 0.0001), and only this marker
significantly predicted cardiac mortality.
Discussion
This is the first study to show that HRT is a more powerful
predictor than other factors, including a reduced LVEF, the
documentation of nonsustained VT, and abnormal HRV on
24-hour Holter ECG, in predicting cardiac mortality in postMI patients with DM.

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

Journal of Cardiovascular Electrophysiology

Vol. 22, No. 10, October 2011

mortality in diabetic patients suffered from MI as well as

nondiabetic post-MI patients.
HRT has 2 components (i.e., a biphasic response)TO
and TS. In this study, HRT was defined as positive when both
TO and TS were abnormal. A baroreceptor reflex thought to
be vagally mediated is usually used to explain this biphasic
response. The initial component (TO) is regulated by sympathetic activation or parasympathetic depression after PVCs,
and TO may not be affected in cardiac autonomic neuropathy,
which itself is a state of relatively predominant sympathetic
tone. In contrast, the second component of the biphasic response (TS) is disturbed in cardiac autonomic neuropathy
because this response is driven by the parasympathetic system. Thus, HRT that was affected by 2 different components
might be a more sensitive indicator of cardiac autonomic abnormality in post-MI patients with DM because these patients
have disturbance of both parasympathetic and sympathetic
tones.
Although the American College of Cardiology/American
Heart Association/European Society of Cardiology 2006
guidelines for management of patients with ventricular arrhythmias and prevention of sudden cardiac death22 stated
that a reduced LVEF is the most useful risk predictor for
overall-cause mortality, this may be not the case in this study
population. HRT appears to be a better marker than a reduced
LVEF in post-MI patients with DM.
Study Limitations
Because of the limited scope of our study, we did not include T-wave alternans,23 ventricular late potentials obtained
from signal-averaged ECGs,24 baroreflex sensitivity,13 and
other electrocardiographic markers, all of which have been
shown to be significant predictors in other populations. We
do not know the value of these markers in predicting cardiac
death or arrhythmic events in this study population.
References
1. Moss AJ, Hall WJ, Cannom DS, Daubert JP, Higgins SL, Klein H,
Levine JH, Saksena S, Waldo AL, Wilber D, Brown MW, Heo M:
Improved survival with an implanted defibrillator in patients with coronary disease at high risk for ventricular arrhythmia. Multicenter Automatic Defibrillator Implantation Trial Investigators. N Engl J Med
1996;335:1933-1940.
2. Denniss AR, Richards DA, Cody DV, Russell PA, Young AA, Cooper
MJ, Ross DL, Uther JB: Prognostic significance of ventricular tachycardia and fibrillation induced at programmed stimulation and delayed
potentials detected on the signal-averaged electrocardiograms of survivors of acute myocardial infarction. Circulation 1986;74:731-745.
3. Cheema AN, Sheu K, Parker M, Kadish AH, Goldberger JJ: Nonsustained ventricular tachycardia in the setting of acute myocardial
infarction: Tachycardia characteristics and their prognostic implications.Circulation 1998;98:2030-2036.
4. Schmidt G, Malik M, Barthel P, Schneider R, Ulm K, Rolnitzky L,
Camm AJ, Bigger JT Jr, Schomig A: Heart-rate turbulence after ventricular premature beats as a predictor of mortality after acute myocardial
infarction. Lancet 1999;353:1390-1396.
5. Ghuran A, Reid F, La Rovere MT, Schmidt G, Bigger JT Jr, Camm AJ,
Schwartz PJ, Malik M; ATRAMI Investigators: Heart rate turbulencebased predictors of fatal and nonfatal cardiac arrest (the Autonomic
Tone and Reflexes After Myocardial Infarction substudy). Am J Cardiol
2002;89:184-190.
6. Miwa Y, Ikeda T, Sakaki K, Miyakoshi M, Ishiguro H, Tsukada T,
Abe A, Mera H, Yusu S, Yoshino H: Heart rate turbulence as a predictor of cardiac mortality and arrhythmic events in patients with dilated cardiomyopathy: A prospective study. J Cardiovasc Electrophysiol 2009;20:788-795.

7. Bauer A, Malik M, Schmidt G, Barthel P, Bonnemeier H, Cygankiewicz

I, Guzik P, Lombardi F, Muller A, Oto A, Schneider R, Watanabe M,
Wichterle D, Zareba W: Heart rate turbulence: Standards of measurement, physiological interpretation, and clinical use: International Society for Holter and Noninvasive Electrophysiology Consensus. J Am
Coll Cardiol 2008;52:1353-1365.
8. Junttila MJ, Barthel P, Myerburg RJ, Makikallio TH, Bauer A, Ulm
K, Kiviniemi A, Tulppo M, Perkiomaki JS, Schmidt G, Huikuri HV:
Sudden cardiac death after myocardial infarction in patients with type
2 diabetes. Heart Rhythm 2010;7:1396-1403.
9. Whang W, Bigger JT Jr: Comparison of the prognostic value of RRinterval variability after acute myocardial infarction in patients with
versus those without diabetes mellitus. Am J Cardiol 2003;92:247251.
10. Balcioglu S, Arslan U, Turkoglu S, Ozdemir M, Cengel A: Heart rate
variability and heart rate turbulence in patients with type 2 diabetes mellitus with versus without cardiac autonomic neuropathy. Am J Cardiol
2007;100:890-893.
11. Report of the Expert Committee in the Diagnosis and Classification of
Diabetes Mellitus. Diabetes Care 1997;20:1183-1197.
12. Barthel P, Schneider R, Bauer A, Ulm K, Schmitt C, Schomig A,
Schmidt G: Risk stratification after acute myocardial infarction by heart
rate turbulence. Circulation 2003;108:1221-1226.
13. 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. ATRAMI (Autonomic
Tone and Reflexes After Myocardial Infarction) Investigators. Lancet
1998;351:478-484.
14. Maser RE, Lenhard MJ: Cardiovascular autonomic neuropathy due
to 2288; diabetes mellitus: Clinical manifestations, consequences, and
treatment. J Clin Endocrinol Metab 2005;90:5896-5903.
15. Boulton AJ, Vinik AI, Arezzo JC, Bril V, Feldman EL, Freeman R,
Malik RA, Maser RE, Sosenko JM, Ziegler D; American Diabetes Association: Diabetic neuropathies: A statement by the American Diabetes
Association. Diabetes Care 2005;28:956-962.
16. Koyama J, Watanabe J, Yamada A, Koseki Y, Konno Y, Toda S, Shinozaki T, Miura M, Fukuchi M, Ninomiya M, Kagaya Y, Shirato K:
Evaluation of heart rate turbulence as a new prognostic marker in patients with chronic heart failure. Circ J 2002;66:902-907.
17. Cygankiewicz I, Zareba W, Vazquez R, Vallverdu M, GonzalezJuanatey JR, Valdes M, Almendral J, Cinca J, Caminal P, de Luna
AB: Muerte Subita en Insuficiencia Cardiaca Investigators: Heart rate
turbulence predicts all-cause mortality and sudden death in congestive
heart failure patients. Heart Rhythm 2008;5:1095-1102.
18. Stein PK, Deedwania P: Usefulness of abnormal heart rate turbulence
to predict cardiovascular mortality in high-risk patients with acute myocardial infarction and left ventricular dysfunction (from the EPHESUS
study). Am J Cardiol 2009;103:1495-1499.
19. Bauer A, Malik M, Barthel P, Schneider R, Watanabe MA, Camm
AJ, Schomig A, Schmidt G: Turbulence dynamics: An independent
predictor of late mortality after acute myocardial infarction. Int J Cardiol
2006;107:42-47.
20. Jeron A, Kaiser T, Hengstenberg C, Lowel H, Riegger GA, Holmer
S: Association of the heart rate turbulence with classic risk stratification parameters in postmyocardial infarction patients. Ann Noninvasive
Electrocardiol 2003;8:296-301.
21. Barthel P, Schneider R, Bauer A, Ulm K, Schmitt C, Schomig A,
Schmidt G: Risk stratification after acute myocardial infarction by heart
rate turbulence. Circulation 2003;108:1221-1226.
22. ACC/AHA/ESC 2006 guidelines for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death: A
report of the American College of Cardiology/American Heart Association Task Force and the European Society of Cardiology Committee
for Practice Guidelines (writing committee to develop guidelines for
management of patients with ventricular arrhythmias and the prevention
of sudden cardiac death). J Am Coll Cardiol 2006;48, e247e346.
23. Ikeda T, Saito H, Tanno K, Shimizu H, Watanabe J, Ohnishi Y,
Kasamaki Y, Ozawa Y: T-wave alternans as a predictor for sudden
cardiac death after myocardial infarction. Am J Cardiol 2002;89:7982.
24. El-Sherif N, Denes P, Katz R, Capone R, Mitchell LB, Carlson M,
Reynolds-Haertle R: Definition of the best prediction criteria of the
time domain signal-averaged electrocardiogram for serious arrhythmic
events in the postinfarction period. The Cardiac Arrhythmia Suppression Trial/Signal-Averaged Electrocardiogram (CAST/SAECG) Substudy Investigators. J Am Coll Cardiol 1995;25:908-914.