European Heart Journal (2002) 23, 1301–1308
doi:10.1053/euhj.2001.3078, available online at http://www.idealibrary.com on
Defining the optimum upper heart rate limit during
exercise
A study in pacemaker patients with heart failure
M. Kindermann1, B. Schwaab2, N. Finkler1, S. Schaller1, M. Böhm1 and
G. Fröhlig1
1
Department of Internal Medicine, Division of Cardiology and Angiology, Universitätskliniken des Saarlandes,
Homburg/Saar, Germany; 2Curschmann-Clinic for Cardiovascular Diseases, Timmendorfer Strand, Germany
Aims There is no non-invasive method to determine the
individual optimum of maximum exercise heart rate.
Knowledge of this value is of particular interest in patients
with structural heart disease who are prone to tachycardia
intolerance. The purpose of this study was to define the
optimal maximum heart rate using cardiopulmonary exercise testing and exercise Doppler echocardiography and to
compare the results of both approaches.
Methods and Results In 49 pacemaker patients with
chronotropic incompetence, the optimum upper heart rate
limit was determined using cardiopulmonary exercise testing and exercise Doppler echocardiography. The optimum
upper rate limit was given by the highest pacing rate which
still produced an increase in oxygen consumption, or by
that pacing rate which was linked to the lowest value for the
Doppler-derived myocardial performance index. In patients
with normal left ventricular ejection fraction (d55%) the
optimum upper rate limit was 86% of age-predicted maximum heart rate, in patients with left ventriuclar dysfunction
(ejection fraction c45%) it was 75% of the age-predicted
maximum rate (P=0·004). The optimum upper rate limit, as
Introduction
Dynamic exercise capacity is a function of stroke volume, heart rate and arteriovenous oxygen difference,
with the latter two factors accounting for about 70% of
the increase in oxygen consumption[1]. In patients
with severe chronotropic incompetence, as defined by a
maximum exercise heart rate of c90 beats . min 1,
pacemaker therapy with rate-adaptive devices leads to a
Revision submitted 30 October 2001, accepted 7 November 2001,
and published online 31 January 2002.
Correspondence: Michael Kindermann, MD, Innere Medizin III
(Kardiologie/Angiologie), Universitätskliniken des Saarlandes,
Kirrberger Straße, D 66421 Homburg/Saar, Germany.
0195-668X/02/$35.00
defined by cardiopulmonary exercise testing and exercise
Doppler echocardiography, were closely correlated
(P<0·0001) with a mean deviation of 66 beats . min 1.
Conclusion Cardiopulmonary exercise testing and exercise
Doppler echocardiography are valuable tools which help to
determine the optimum upper rate limit in order to avoid
excess heart rates in heart failure patients. The application
of these methods is not limited to pacemaker patients
but may be helpful in therapeutic interventions with
chronotropic drugs.
(Eur Heart J, 2002; 23: 1301–1308, doi:10.1053/euhj.2001.
3078)
2002 The European Society of Cardiology. Published by
Elsevier Science Ltd. All rights reserved.
Key Words: Heart rate, exercise, heart failure, cardiac
pacing.
See page 1239, doi:10.1053/euhj.2002.3180 for the Editorial
comment on this article
highly significant increase in aerobic capacity and maximum workload performed[2]. However, there is no
agreement upon the haemodynamically optimal upper
heart rate limit and methodological approaches to define
it in the individual patient. The maximum age-predicted
heart rate, as calculated by the well-known formula
(220-age)[3], is no more than a calibration standard to
estimate the level of physical stress during exercise
testing. It is not necessarily identical with the heart
rate that produces maximum cardiac output during
peak exercise. This holds true particularly in patients
with structural heart disease, who are prone to variable degrees of tachycardia intolerance. Induction
of ischaemia, impairment of diastolic filling[4] and a
blunted or inversed force-frequency relationship[5–7] are
the main factors that limit cardiac output augmentation
2002 The European Society of Cardiology. Published by Elsevier Science Ltd. All rights reserved.
1302
M. Kindermann et al.
during exercise-induced tachycardia in patients with
heart disease.
Knowledge of the haemodynamically optimal limit
of exercise heart rate is not only useful for patients
with bradycardia who are treated with rate adaptive
pacemakers, it may also serve to titrate the dosage of
negative chronotropic agents in patients with inadequate
tachycardia. As tachycardia is considered to be related
to an adverse outcome[8], excess heart rates in heart
failure patients could be avoided by an optimal setting
of the target heart rate during exercise. Since a practical
approach for determination of the individual optimum for maximum exercise heart rate is missing, it
was the object of this study to propose practical and
non-invasive methods to define it.
Methods
Patients
Forty-nine patients with permanent cardiac pacemakers
were investigated. All patients gave informed consent for
participation in the study, which followed institutional
guidelines. All patients revealed an inappropiate heart
rate increase during exercise, when the rate response
function of their pacing system was turned off. Chronotropic incompetence was defined by a heart rate increase
of less than 2 beats . min 1 per oxygen consumption
ml . kg 1 min 1[9,10]. During the study, five patients
with atrial fibrillation were paced in VVI mode, two
patients ran in AAI mode and the majority of 42
patients had their pacemakers programmed to DDD
mode. In these patients the programmable atrioventricular delay of the pulse generator was individually
optimized prior to the study. Methods used for atrioventricular delay optimization were Doppler echocardiography of transmitral[11] or transaortic[12] blood flow or
impedance cardiography[11].
To analyse the impact of left ventricular function on
the optimum upper heart rate limit, the patient population was dichotomized into two groups, one with left
ventricular dysfunction defined by an ejection fraction
c45% (group I) and the other with normal left ventricular function (ejection fraction d55%, group II).
Patients with only minor degrees of left ventricular
dysfunction (ejection fraction 46–54%) were not
included in the study. Ejection fraction was estimated by
echocardiography using the Teichholz method or — if
left heart catheterization was performed for another
clinical indication — by angiography. As the majority of
patients from group I had a clinical indication for
angiography, ejection fraction was measured by levocardiography in 15 out of 18 patients (83%) from
group I, whereas levocardiography was performed in
only eight patients (35%) from group II.
Cardiopulmonary exercise testing
All patients underwent symptom-limited cardiopulmonary exercise testing with breath-by-breath gas
Eur Heart J, Vol. 23, issue 16, August 2002
Table 1
Time (min)
0
1
2
3
4
5
6
7
8
9
10
Heart rate increments during exercise
Workload (Watt)
Pacing rate
(beats . min 1)
0
15
30
45
60
75
90
105
120
135
150
70
85
90
100
105
110
120
130
140
150
160
exchange analysis using a MedGraphics CPX/D spiroergometry system (Medical Graphics Corporation,
St. Paul, MN, U.S.A.). Patients performed bicycle exercise in a 45 semisupine position lying on an ergometrics
900EL reclining ergometer (Ergoline cardio-systems,
Bitz, Germany). After adaptation to the mouthpiece and
resting conditions, exercise workload was increased continuously by 15 Watt . min 1 (ramp protocol). During
exercise, the rate response function of the pulse generators was turned off and the paced heart rate was
increased by manual programming according to a fixed
protocol (Table 1). The protocol was based on the
finding[9,10] that an increase in oxygen uptake of
1 ml . kg 1 min 1 is related to a heart rate increase
between 2 and 4 beats . min 1 in patients without
chronotropic incompetence. Thus, the exercise workloads (Watt) were converted into increments of oxygen
uptake (ml . kg 1 min 1) which were multiplied by
pre-defined heart rate slopes to obtain the heart rate
increment for each stage. At the beginning of exercise a
low heart rate slope of 2·4 beats . min 1 per ml . kg 1
min 1 was chosen which increased to a maximum slope
of 4·8 beats . min 1 per ml . kg 1 min 1 with increasing workload. Due to the limited programming options
of the pacemaker models, not all heart rates could be
programmed and the target pacing rates in Table 1
represent a compromise between the heart rates required
by the pre-defined heart rate slope and the spectrum of
programmable pacing rates. In each patient, maximum
workload achieved, peak oxygen consumption and oxygen uptake at the ventilatory anaerobic threshold were
measured. The anaerobic threshold was determined
according to the V-slope method[13].
The optimum upper heart rate limit during exercise
was defined as the highest pacing rate which was still
producing an increase in oxygen consumption (V
~ O2).
The time course of V
~ O2 at the end of exercise was
assumed to reflect the time course of cardiac output[14].
This assumption is supported by Fick’s principle where
oxygen consumption is the product of cardiac output
and arteriovenous oxygen difference. As the increase
in arteriovenous oxygen difference during exercise in
relation to exercise workload indicates curvilinear
behaviour[15], arteriovenous oxygen difference near peak
Optimum heart rate limit during exercise
a
LV outflow
FT
IRT
ET
ICT
exercise is virtually constant so that oxygen consumption is directly correlated to cardiac output. A significant
increase in V
~ O2 was defined by a cutoff value of d10
ml . min 1 Watt 1[16]. It was hypothesized that patients
who are overpaced beyond their optimum upper heart
rate limit would show a plateau or even a decrease in
oxygen consumption.
1303
Doppler echo examination
All studies were performed with an Ultramark 9 HDI
(Advanced Technology Laboratory, Inc., Bothell, WA,
U.S.A.) ultrasound system and a 3·5 MHz phased array
transducer used for pulsed wave Doppler recordings.
Mitral and aortic flow velocities were obtained with the
sample volume placed between the valve leaflets to
record valve opening and closure. Recordings were
performed at rest and during exercise after each increase
in pacing rate and stored on videotape. The following
parameters were measured using the flow velocity
recordings (Fig. 1): left ventricular filling time, isovolumic contraction time, left ventricular ejection time
and isovolumic relaxation time. All measurements were
done in triplicate with the mean value used for further
analysis. Two Doppler echo indices were calculated: the
Tei index[17] and the Z ratio[18]. The Tei index is the sum
of isovolumic contraction and relaxation time divided by
left ventricular ejection time. It has been shown to be
inversely correlated to both left ventricular systolic as
well as diastolic function and is supposed to be an
overall myocardial performance index[17,19]. The Z ratio
is the sum of left ventricular filling and ejection time
given as a percentage of cardiac cycle length. Its relationship to systolic and diastolic parameters of left
ventricular function is less well established but it has
been shown that the Z ratio is highest in patients with
normal left ventricular function and a normal QRS
duration, while it is minimal in patients with dilated
cardiomyopathy and left bundle branch block[18]. The
optimum upper heart rate limit, as measured by Doppler
echocardiography, was defined by the exercise heart rate
where the Tei index reached its minimum and the Z ratio
was at its maximum.
Mitral inflow
CL
AOC
MVO
CL
Atrial pacing spike
Ventricular pacing spike
Figure 1 Doppler echo time intervals given in the study.
FT=left ventricular filling time; ICT=isovolumic contraction time; ET=left ventricular ejection time; IRT=
isovolumic relaxation time; CL=cycle length. a=interval
between mitral valve closure and opening; AOC,
MVO=intervals between ventricular pacemaker spike (or
beginning of QRS) and aortic valve closure and mitral
valve opening, respectively. ET, a, AOC and MVO were
directly measured. CL was given by the pacing rate
(CL=6000/pacing rate [ms]). The other parameters were
calculated using the following formulae: IRT=MVO
AOC; ICT=aIRTET; FT=CLa; TEI index=
(aET)/ET; Z ratio=(FT+ET)/CL.
Results
Statistics
Values are expressed as meanSD. Differences between
both groups of patients were compared using the U-test
of Mann and Whitney. Differences between different
stages of exercise were evaluated with Friedman’s test.
Frequency distributions were compared with Fisher’s
exact test. Spearmans rank order correlation and
Wilcoxon’s signed rank test were used to analyse the
relationship between the different approaches to determine optimum upper heart rate limit. A probability
value of P<0·05 was considered significant.
Cardiopulmonary exercise testing
During spiroergometry, 41 patients (83·7%) reached the
anaerobic threshold. Data of the remaining eight
patients, who did not exercise until anaerobic threshold,
were not considered for further data analysis. Hence,
18 patients were left in group I and 23 in group II. Both
groups were comparable with respect to age (Table 2).
Gender was not equally distributed as four out of the
five female patients of this study were included in group
II. Maximum workload achieved, maximum oxygen
consumption and oxygen consumption at the anaerobic
threshold, the optimum upper heart rate limit in
Eur Heart J, Vol. 23, issue 16, August 2002
1304
M. Kindermann et al.
Table 2
Results of cardiopulmonary exercise testing
Group I
Group II
P
Number, n
18
23
—
Age, years
70·09·3
68·512·0
ns
Max. workload, Watt
88·218·7 112·722·4 0·002
9·42·4
13·02·8 0·001
V
~ O2AT, ml . min 1 kg 1
14·23·7
19·65·4 0·001
V
~ O2max, ml . min 1 kg 1
1
122·811·3 135·414·2 0·004
HRmax, beats . min
111·910·0 129·615·4 0·001
OURL, beats . min 1
OURL/MHR, %
75·09·0
85·911·2 0·004
1
OURL d130 beats . min , n
2
17
0·001
~ O2max =
V
~ O2AT =oxygen consumption at anaerobic threshold; V
maximum oxygen consumption; HRmax =maximum pacing rate
achieved; OURL=optimum upper rate limit; MHR=age-predicted
maximum heart rate (220-age); OURL/MHR=OURL expressed as
a percentage of MHR. ns=not significant. Results are given as
mean valuestandard deviation.
Table 3 Results of cardiopulmonary exercise testing.
Patients with V~O2-plateau
Group I
Group II
P
Number, n
15
14
—
Age, years
70·98·7
69·75·3
ns
Max. workload, Watt
90·315·4 111·921·9 0·015
1
1
kg
9·52·6
12·62·0 0·006
V
~ O2AT, ml . min
14·43·9
18·53·3 0·003
V
~ O2max, ml . min 1 kg 1
HRmax, beats . min 1
124·010·6 133·214·1 0·049
1
110·99·1 123·613·2 0·021
OURL, beats . min
OURL/MHR, %
74·78·0
82·28·7 0·034
1
9
0·002
OURL d130 beats . min 1, n
For abbreviations see Table 2.
absolute terms and as a percentage of maximum agepredicted heart rate were all significantly lower in group
I than in group II. The majority (70%) of patients
in group II had an optimum upper rate limit of
d130 beats . min 1, while this was exceptional (11%) in
group I patients (P<0·001).
The formation of a V
~ O2-plateau was observed in 29
out of 41 patients (71%); it was more common in group
I (83%) than in group II (61%), but this difference did
not reach significance. Due to their limited exercise
capacity, patients in group I achieved significantly
(P<0·004) lower maximum pacing rates as compared
to group II patients (12311 beats . min 1 vs
13514 beats . min 1). To rule out that this was the
main reason for a lower optimum upper rate limit, this
parameter was also compared for patients with a V
~ O2plateau (Table 3). Again, absolute and relative values for
the optimum upper rate limit were significantly lower in
patients with left ventricular dysfunction. Table 4 gives
an overview of the individual relations between the
maximum age-predicted heart rate and the optimum
upper heart rate limit in both groups. A typical example
of normal oxygen kinetics in a patient without left
ventricular dysfunction and the early formation of a
Eur Heart J, Vol. 23, issue 16, August 2002
Table 4 Individual maximum age predicted and optimal
maximum heart rates as assessed by cardiopulmonary
exercise testing
Group I
MHR
146
155
146
143
144
149
162
141
140
169
169
148
147
151
144
156
153
138
Group II
OURL
120
90*
120*
105*
110*
130
110*
105*
130*
110*
100
110*
120*
110*
110*
110*
120*
104*
26
65
26
38
34
19
52
36
10
59
69
38
27
41
34
46
33
34
1509 11210†
MHR
147
158
147
148
149
136
137
150
148
157
152
143
200
157
151
150
152
158
148
146
152
147
151
OURL
160
130*
130*
140*
130*
140
120*
130*
140
140*
130*
130
150
110
105*
150
130*
110*
105*
100*
130*
140
130
+13
28
17
8
19
+4
17
20
8
17
22
13
50
47
46
0
22
48
43
46
22
7
21
3815‡ 15112 13015
2218
MHR=age-predicted maximum heart-rate (220-age); OURL=
optimum upper heart rate limit; =difference (OURLMHR).
~ O2All values are expressed in beats . min 1. Patients with a V
plateau are marked by an asterisk. Mean valuesSD are given in
the last line. †P<0·001 vs group II. ‡P<0·006 vs group II.
V
~ O2-plateau in a patient with impaired left ventricular
function is depicted in Fig. 2.
Doppler examination
Doppler echo examinations were performed in 33 out
of the 41 patients whose data could be analysed by
cardiopulmonary exercise testing. In general, all
Doppler time intervals decreased significantly with
exercise (Fig. 3), but differences between both groups
were limited to diastolic filling time and isovolumic
contraction time during maximum workload or optimum upper pacing rate and thus mainly due to the
higher exercise capacity and higher pacing rate achieved
in group II. Both Doppler echo indices changed significantly (P<0·001) with exercise: the Tei index decreased
from 0·720·21 (group I) and 0·670·12 (group II)
to an average minimum of 0·380·16 (group I) and
0·340·13 (group II), respectively. The Z-ratio
increased from 77·65·2% (group I) and 78·23·2%
(group II) to an average maximum of 83·94·5%
(group I) and 84·14·9% (group II), respectively. In 21
out of 33 cases (64%) the Tei index approached to a
Optimum heart rate limit during exercise
4
200
Exercise-end #1
MHR (#1)
Exercise-end #2
160
OURL (#2)
150
3
∆ = +13 bpm
140
∆ = –52 bpm
150
MHR (#2)
–1
130
·
–1
VO2 (l.min )
120
100
2
85
105
110
OURL (#1)
100
90
HR
70
Heart rate (beats.min )
Start of exercise
Patient #2
1
50
·
VO2
0
1305
1
2
Patient #1
3
4
5
6
7
8
Time (min)
9
10
11
12
13
14
15
0
Figure 2 Different oxygen kinetics in relation to the increase in pacing rate in a patient from group I
(no. 1, solid line) and group II (no. 2, dotted line). Patient no. 1 was a 58-year-old man with an ejection
fraction of 30%. Despite ongoing exercise and a further increment in pacing rate, V
~ O2 failed to increase
beyond a pacing rate of 110 beats . min 1 (V
~ O2-plateau). The resulting difference between the maximum
age-predicted heart rate (MHR) of 162 beats . min 1 and the optimum upper rate limit (OURL) of
110 beats . min 1 was 52 beats . min 1. Interestingly, V
~ O2 peaked shortly after the end of exercise when
pacing rate sharply decreased. In contrast to patient no. 1, in patient no. 2 (73 years, ejection fraction 80%)
oxygen consumption was continuously increasing up to the end of exercise without formation of a plateau. The
optimum upper rate limit was estimated at 160 beats . min 1, 13 beats . min 1 above the theoretical
maximum heart rate of 147 beats . min 1.
minimum value during the course of exercise and increased again until the end of exercise. The Z ratio first
increased and then decreased during exercise in 25 out of
33 patients (76%). Despite a slightly lower mean Tei
index and a slightly higher Z ratio at rest and during
maximum workload in group II, these differences were
far from being significant.
Although neither the Tei index nor the Z ratio nor
any other Doppler parameter was able to differentiate
between patients with and without left ventricular dysfunction, the intra-individual time course of the Doppler
indices allowed for the identification of the optimum
upper heart rate limit in the majority of patients. In
Table 5, the optimum upper rate limits as determined by
cardiopulmonary exercise testing, by the minimum of
the Tei index and by the maximum of the Z ratio are
given for each of the 33 patients who underwent a
combined examination of ventilatory and Doppler echo
parameters. The correlation for the relationship between
the optimum upper rate limits as determined by spiroergometry and by the two Doppler indices was highly
significant (P<0·0001). As indicated by the rank corre-
lation coefficient, the correlation was better for the Tei
index (r=0·81, see Fig. 4) than for the Z ratio (r=0·66).
The mean deviation between spiroergometry and
Doppler echo was significantly (P<0·008) smaller for the
Tei index (deviation 66 beats . min 1) than for the
Z ratio (deviation 1213 beats . min 1).
Discussion
This is the first study that presents an approach for the
determination of the individual optimum for maximum
exercise heart rate in patients with heart failure. Pacemaker stimulation in chronotropic incompetence was
used as an experimental model to study the effects of
heart rate increase on cardiac function during exercise.
According to the results presented, both methods, cardiopulmonary exercise testing with evaluation of oxygen
uptake kinetics as well as exercise Doppler echocardiography with successive measurements of the Tei
index lead to similar results.
Eur Heart J, Vol. 23, issue 16, August 2002
1306
M. Kindermann et al.
ns
ns
350
300
ns
p < 0·001
250
p < 0·04
200
150
100
Ejection time (ms)
300
400
ns
p < 0·05
250
200
150
100
50
50
0
Isovolumic contraction time (ms)
350
ns
450
Rest
120
100 bpm
ns
ns
100
ns
80
p < 0·03
60
40
20
0
Rest
100 bpm
0
Max. WL Opt. URL
Isovolumic relaxation time (ms)
Diastolic filling time (ms)
500
Max. WL Opt. URL
180
160
Rest
100 bpm
Max. WL Opt. URL
ns
140
120
ns
100
ns
ns
80
60
40
20
0
Rest
100 bpm
Max. WL Opt. URL
Figure 3 Doppler echo time intervals at rest, during exercise with a pacing rate of 100 beats . min 1, at
maximum workload (Max. WL) and at the optimum upper rate limit (Opt. URL) as determined by
cardiopulmonary exercise testing. Comparisons between groups I and II are given in the figure (ns, not
significant). Intra-group comparisons can be summarized as follows. Group I: ( ) for filling and ejection time,
all differences were significant with the exception of 100 beats . min 1 vs Opt. URL. For isovolumic
contraction time, only Rest vs Max. WL and Rest vs Opt. URL were significant. For isovolumic relaxation
time, all comparisons but Rest vs 100 beats . min 1 and Max. WL vs Opt. URL were significant. Group II:
( ) all differences were significant with the following exceptions: Max. WL vs Opt. URL for filling time and
isovolumic relaxation time and Rest vs 100 beats . min 1 for ejection time.
Optimum upper rate limit – Tei index
–1
(beats.min )
The findings of this study which reported significantly
lower optimum upper rate limits for patients with left
ventricular dysfunction are in accordance with previous
Table 5 Comparison of optimum upper rate limits as
defined by cardiopulmonary exercise testing and Doppler
echo
CPx
Tei
Z
150
130
130
140
130
120
140
120
130
90
140
140
130
150
120
105
110
150
130
140
120
130
120
140
130
130
85
120
140
140
140
120
120
100
150
130
140
120
130
110
80
130
130
70
120
140
110
130
120
120
80
CPx
Tei
Z
105
150
130
110
130
105
140
130
110
110
120
110
110
110
120
104
105
150
140
120
115
110
140
120
105
110
120
120
120
110
130
104
105
150
140
120
115
110
140
120
105
70
120
120
120
80
110
104
160
150
Spearman's R = 0·81
140
130
120
110
100
Line of equality
90
80
90 100 110 120 130 140 150 160
–1
Optimum upper rate limit – CPx (beats.min )
Figure 4 Correlation between the optimum upper heart
rate limits as measured by cardiopulmonary exercise
testing (CPx) and by Doppler echo (Tei index). Eightyeight percent of all points are within a 10 beats . min 1
frequency band around the line of equality.
Eur Heart J, Vol. 23, issue 16, August 2002
K.Ch.
I.S.
G.R.
F.Lu.
H.Bl.
A.L.
S.Ko.
S.B.
H.E.
J.A.
E.K.
W.F.
D.U.
J.S.
K.A.
K.S.
S.N.
M.W.
H.Be.
H.M.
S.M.
K.Co.
H.W.
A.S.
K.H.
D.R.
K.P.
L.M.
F.B.
S.Kl.
P.T.
L.G.
F.Lo.
CPx/Tei/Z=optimum upper rate limits (in beats . min 1) as
defined by cardiopulmonary exercise testing (CPx) or Doppler
echocardiography using the Tei-index or Z-ratio criterion.
Optimum heart rate limit during exercise
studies: Sowton[20] using indicator-dilution cardiac
output measurements, was the first to observe that
optimal heart rates on exercise were considerably lower
in patients with myocardial disease. However, only
two studies[21,22] systematically investigated the influence
of the upper rate limit in pacemaker patients with
and without heart failure. Using cardiopulmonary
exercise testing and fixed upper rate limits, they found
that patients with advanced heart failure increased
their aerobic capacity by rate adaption up to
110 beats . min 1 but had no additional benefit with
heart rates between 110 and 130 beats . min 1.
One of the main results of this study is that the
majority (71%) of patients investigated had optimum
upper heart rate limits that were below the maximum
pacing rates programmed at peak exercise. These
patients were characterized by a levelling off in oxygen
uptake, a failure to increase oxygen consumption despite
ongoing exercise with increasing workloads and increasing pacing rate. As oxygen consumption is the product
of heart rate, stroke volume and arteriovenous oxygen
difference and since the latter cannot decrease if workload is still increasing, the lack of an increase in oxygen
uptake despite increasing pacing rate can only be
explained by a fall in stroke volume.
The decrease in stroke volume may be understood as
an impairment of cardiac pump performance — a systolic or diastolic or combined malfunction — induced by
inadequate tachycardia. Three possible mechanisms may
explain tachycardia related decreases in pump function:
induction of ischaemia[22], impairment of left ventricular
diastolic filling[23,24] and a loss or an inversion of the
positive force–frequency relationship[5–7,23]. Although
far more frequent in severe structural heart disease, none
of these mechanisms is exclusively linked to low ejection
fractions: ischaemia may limit exercise and heart rate
tolerance even in patients with normal pump function at
rest[22], a predominantly disturbed diastolic function
with maintained contractility is typical for left ventricular hypertrophy and different degrees of an impaired
force–freqency relationship have been observed in various types of heart disease[7]. Thus, the optimum upper
heart rate limit is not a function of a single parameter of
cardiac haemodynamics but the result of a complex
interaction of systolic and diastolic properties. This is
the main reason why both groups of this study included
patients with a plateau of oxygen uptake kinetics despite
a significant group difference for the optimum upper
rate limit.
Although it can be derived from this study that in
patients with a normal ejection fraction the upper pacing
rate should be limited to 85% of age-predicted maximum
heart rate and in patients with a reduced ejection fraction it should not exceed 75%, these are just rules of
thumb. The inability of ejection fraction to predict the
optimum upper rate limit in the individual patient is
demonstrated by the wide scatter in each study group.
Two subjects in the left ventricular dysfunction group
had rather high upper rate optima of 130 beats . min 1,
which corresponded to 87% and 92% of the age-
1307
predicted maximum heart rate, whereas in the group
with a normal left ventricular ejection fraction there
were six subjects with optimum upper rate limits of only
68%–75%. This emphasizes the importance of an individualized approach which may be crucial in young,
active people who suffer from impaired cardiac function.
In these patients the potential span between the physiological optimum of maximum pacing rate and the
maximum age-predicted heart rate is particularly large.
A maximum pacing rate that is programmed too low
will cause an additional impairment of aerobic capacity,
while an excess of overpacing during exercise does not
contribute to cardiac output and might be harmful. It
has been shown that tachycardia at rest[8] and an attenuated heart rate recovery after exercise[25] are predictive of
mortality. Whether the same is true for an inadequately
high heart rate during exercise is not known and requires
further investigations.
The methods proposed in this study have been applied
to pacemaker patients with chronotropic incompetence
for two reasons: firstly, there is a clinical need for a tool
to optimize pacing haemodynamics in these patients and
secondly, investigations of the heart rate-related impact
on haemodynamics can more easily be performed in
pacemaker patients because heart rate is under the
control of the investigator. However, the approaches of
this study may have wider clinical applications. This
could be the identification of inappropiate exercise
tachycardia in patients with impaired left ventricular
function. Basically both methods, cardiopulmonary
exercise testing with evaluation of oxygen uptake
kinetics as well as repetitive measurements of the Tei
index, should permit not only the identification of the
optimum upper heart rate limit, they may additionally
serve as tools for the titration of drugs with negative
chronotropic effects. In patients with inadequate
exercise tachycardia due to atrial fibrillation, these
methods may be used to control exercise heart rate on a
haemodynamically based rationale. In heart failure
patients, the information that a given exercise heart rate,
which might be far below the maximum age-predicted
rate, is haemodynamically useless, may be valuable for
tailoring therapy with beta-blockers. Further studies are
needed to evaluate the clinical value of these approaches
with regard to morbidity and mortality.
Interestingly, there was no significant difference in the
Tei index between patients with normal and impaired
left ventricular function. Tei et al.[17] demonstrated that
this index discriminates between normals and patients
with different degrees of left ventricular dysfunction.
The finding of this study that the Tei index was a good
indicator of intra- but not inter-individual cardiac performance may be explained by different patients studied.
All patients in the current study had cardiac pacemakers
and, with the exception of two patients, were paced in
the right ventricle. As right ventricular pacing by itself
reduces left ventricular performance[26] this might have
worsened the Tei index in both groups resulting in a loss
in discriminative power. This hypothesis is supported by
the rather high value for the resting Tei index in our
Eur Heart J, Vol. 23, issue 16, August 2002
1308
M. Kindermann et al.
‘normal’ population (0·670·12) as compared to
0·390·05 which was reported for normal subjects
without pacemaker by Tei et al.[17]. The inability of the Z
ratio to distinguish between pacemaker patients with
intact and impaired left ventricular function was not too
surprising, because this parameter seems to be more
susceptible to changes in ventricular activation sequence
than to changes in systolic parameters[18]. As the vast
majority of patients in both groups had an altered
ventricular activation pattern due to right ventricular pacing this might have levelled out the impact of
different ejection fractions.
Limitations
The assumption that arteriovenous oxygen difference at
the end of exercise is virtually constant was not confirmed by direct measurements and the influence of
different exercise modalities and heart rate protocols was
not investigated. The definition of the optimum upper
rate limit was guided by the intention to maximize
cardiac output. This method of defining optimal rate,
however, requires a reminder that cardiac integrity
should never be compromised by maximizing output.
Thus, in case of ischaemia, hypoxia or arrhythmias, high
heart rates are not optimal for the patient even if they
produce maximum cardiac output.
Conclusions
Cardiopulmonary exercise testing and exercise Doppler
echocardiography are valuable tools to define the optimum upper heart rate limit in individual patients with
rate adaptive cardiac pacemakers. In patients with left
ventricular dysfunction, the optimum upper heart
rate limit is usually lower than in patients with intact
left ventricular function. However, an individualized
approach to determine the optimum heart rate limit is
mandatory and specifically recommended in young and
active patients with severe heart disease to avoid excess
heart rates which might be harmful.
References
[1] Higginbotham MB, Morris KG, Williams RS, McHale PA,
Coleman RE, Cobb FR. Regulation of stroke volume during
submaximal and maximal upright exercise in normal man.
Circ Res 1986; 58: 281–91.
[2] Alt EU, Schlegl MJ, Matula MM. Intrinsic heart rate
response as a predictor of rate-adaptive pacing benefit. Chest
1995; 107: 925–30.
[3] Tanaka H, Monahan KD, Seals DR. Age-predicted maximal
heart rate revisited. J Am Coll Cardiol 2001; 37: 153–6.
[4] Brutsaert DL, Sys SU, Gillebert TC. Diastolic failure: pathophysiology and therapeutic implications. J Am Coll Cardiol
1993; 22: 318–25.
[5] Böhm M, La Rosée K, Schmidt U, Schulz C, Schwinger
RHG, Erdmann E. Force-frequency relationship and
inotropic stimulation in the nonfailing and failing human
myocardium: implications for the medical treatment of heart
failure. Clin Investig 1992; 70: 421–5.
Eur Heart J, Vol. 23, issue 16, August 2002
[6] Pieske B, Kretschmann B, Meyer M et al. Alterations in
intracellular calcium handling associated with the inverse
force-frequency relation in human dilated cardiomyopathy.
Circulation 1995; 92: 1169–78.
[7] Alpert NR, Leavitt BJ, Ittleman FP, Hasenfuss G, Pieske B,
Mulieri LA. A mechanistic analysis of the force-frequency
relation in non-failing and progressively failing human
myocardium. Basic Res Cardiol 1998; 93: 23–32.
[8] Kjekshus J, Gullestad L. Heart rate as a therapeutic target in
heart failure. Eur Heart J 1999; 1 (Suppl H): 64–9.
[9] McElroy PA, Janicki JS, Weber KT. Physiologic correlates of
the heart rate response to upright isotonic exercise: relevance
to rate-responsive pacemakers. J Am Coll Cardiol 1988; 11:
94–9.
[10] Treese N, MacCarter D, Akbulut O et al. Ventilation and
heart rate response during exercise in normals: relevance for
rate variable pacing. PACE 1993; 16: 1693–700.
[11] Kindermann M, Fröhlig G, Doerr T, Schieffer H. Optimizing
the AV delay in DDD pacemaker patients with high
degree AV block: mitral valve Doppler versus impedance
cardiography. PACE 1997; 20: 2453–62.
[12] Mehta D, Gilmour S, Ward DE, Camm AJ. Optimal atrioventricular delay at rest and during exercise in patients with
dual chamber pacemakers: a non-invasive assessment by
continuous wave Doppler. Br Heart J 1989; 61: 161–6.
[13] Beaver WL, Wasserman K, Whipp BJ. A new method for
detecting anaerobic threshold by gas exchange. J Appl Physiol
1986; 60: 2020–7.
[14] McElroy PA, Janicki JS, Weber KT. Cardiopulmonary exercise testing in congestive heart failure. Am J Cardiol 1988; 62:
35A–40A.
[15] Ekelund LG, Holmgren A. Central hemodynamics during
exercise. Circ Res 1967; 20/21: 33–43.
[16] Hansen JE, Sue DY, Wasserman K. Predicted values
for clinical exercise testing. Am Rev Respir Dis 1984; 129:
S49–S55.
[17] Tei C, Ling LH, Hodge DO et al. New index of combined
systolic and diastolic myocardial performance: a simple
and reproducible measure of cardiac function — a study in
normals and dilated cardiomyopathy. J Cardiol 1995; 26:
357–66.
[18] Zhou Q, Henein M, Coats A, Gibson D. Different effects
of abnormal activation and myocardial disease on left
ventricular ejection and filling times. Heart 2000; 84: 272–6.
[19] Tei C, Nishimura RA, Seward JB, Tajik AJ. Noninvasive
Doppler-derived myocardial performance index : correlation
with simultaneous measurements of cardiac catheterization
measurements. J Am Soc Echocardiogr 1997; 10: 169–78.
[20] Sowton E. Haemodynamic studies in patients with artificial
pacemakers. Br Heart J 1963; 26: 737–46.
[21] Haltern G, Kempa L, Ochs JG, Hanrath P, Sigmund M.
Chronic frequency-adaptive pacemaker therapy in patients
with heart failure. Z Kardiol 1995; 84: 834–43.
[22] Krüger S, Stellbrink C, Frielingsdorf J, Hermanns E, Sigmund
M, Hanrath P. Value of spiro-ergometry and stress echocardiography on optimized programming of the upper frequency
limit of dual chamber cardiac pacemakers. Z Kardiol 1998; 87:
817–25.
[23] Feldman MD, Alderman JD, Aroesty JM. Depression of
systolic and diastolic myocardial reserve during atrial pacing
tachycardia in patients with dilated cardiomyopathy. J Clin
Invest 1988; 82: 1661–9.
[24] Ng KS, Gibson DG. Impairment of diastolic function by
shortened diastolic filling period in severe left ventricular
disease. Br Heart J 1989; 62: 246–52.
[25] Cole CR, Blackstone EH, Pashkow FJ, Snader CE, Lauer
MS. Heart-rate recovery immediately after exercise as a
predictor of mortality. N Engl J Med 1999; 341: 1351–7.
[26] Owen CH, Esposito DJ, Davis JW, Glower DD. The effects of
ventricular pacing on left ventricular geometry, function,
myocardial oxygen consumption, and efficiency of contraction
in conscious dogs. PACE 1998; 21: 1417–29.