Influence of Heart Rate and Inhibition of
Autonomic Tone on the QT Interval
S. AHNVE, M.D.,
AND
H. VALLIN, M.D.
SUMMARY To evaluate whether heart-rate-induced changes of the QT interval are dependent on autonomic tone, we studied 13 healthy subjects, mean age 67.5 years. The maximal uncorrected QT from leads I,
II, V1 and V,, was determined during atrial pacing at 90 beats/min and 130 beats/min before and after i.v. administration of propranolol, 0.1 mg/kg, and atropine, 0.02 mg/kg.
Significant reductions (p < 0.01) of QT were induced by the paced increases in heart rate before drugs
(10%), after propranolol (10%) and after the combination of atropine and propranolol (9%). Propranolol
caused no significant change in the QT interval when heart rate was held constant by pacing. In contrast,
atropine produced rate-independent reductions of QT interval (5%) in subjects with ,3-adrenergic blockade (p
< 0.05). Bazett's formula for heart-rate correction of the QT interval (QTc) was not applicable for atrial overdrive pacing, as it gave proportionately longer QTc values at higher heart rates.
These results show that heart rate is a major determinant of the duration of the QT interval and that paced
changes in heart rate induce QT-interval responses that are essentially uninfluenced by autonomic tone. The
rate-independent effect of the QT interval produced by elimination of cholinergic tone suggests a direct influence of cholinergic activity on the repolarization of ventricular myocardium.
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
INTEREST in obtaining clinical information from
the QT interval of the ECG has recently increased.
One reason for this is the reported association between prolonged QT intervals and serious ventricular
arrhythmiasl'6 and the identification of QT prolongation as a prognostic factor for sudden death.' 7 8
Several factors influence the QT interval, including
electrolyte disturbances" and drugs.9""' The effect of
sympathetic nervous activity on the QT interval is well
established.'2-'5 In addition, the autonomic nervous
system affects the ventricular fibrillation threshold.'6, 17 Although cholinergic activity has been found
to enhance ventricular electrical stability in dog experiments and the presence of cholinergic nerve fiber
supplying the ventricular conducting system has been
demonstrated in man,'8 an effect of cholinergic activity on the QT interval has not been reported before.
Heart rate is a major determinant of the duration of
the QT interval and this relation has been studied by
many workers.9' 19-22 In those studies, data were collected and the QT interval was measured at the spontaneous heart rate of each subject. Consequently, the
data probably include individual differences both in intrinsic properties of the ventricular myocardium and
in the autonomic balance at the time of measurement
and thus do not clarify whether QT is strictly ratedependent or whether it is also directly influenced by
autonomic tone.
We undertook the present study to determine the
variation of the QT interval in the same subjects (1) at
spontaneous and atrially paced heart rates; (2) before
and after inhibition of autonomic activity by propranolol and atropine; (3) to evaluate to what extent
rate-related QT changes are dependent on autonomic
neural tone; and (4) to determine whether Bazett's
formula is applicable to changes in heart rate produced by atrial pacing.
Material and Methods
Thirteen healthy volunteers, mean age 67.5 years
(range 54-76 years), were included in the study, which
was part of an electrophysiologic investigation to
define reference values for sinus node and atrioventricular (AV) node function.23 Requirements for participation were absence of a history of heart disease
and absence of previous or present therapy with cardioactive drugs. Other exclusion criteria were signs of
functionally relevant cardiac disease, ECG findings
suggesting cardiac conduction disturbances and cardiac enlargement by chest x-ray.24 Five patients in the
original group were excluded from the present study
because clearly defined T waves were not present in all
parts of the investigation protocol or 1: 1 AV conduction or 2:1 AV block occurred after propranolol during pacing at 130 beats/min. Patients gave informed
consent after the study was approved by the local
ethics committee.
The electrophysiologic investigation was performed with patients in a nonsedated, postabsorptive
state. For atrial pacing, an electrode catheter was introduced through a femoral vein and advanced to the
superior lateral part of the right atrium. ECG leads I,
II, V, and V6 were recorded on an ink-jet recorder
(Siemens-Elema) at a paper speed of 100 mm/sec during spontaneous heart rate and at the end of 30 seconds of atrial pacing at 90 and 130 beats/min. This
protocol was repeated 4 minutes after the administration of i.v. propranolol, 0.1 mg/kg, over 30 seconds
and again 5 minutes after i.v. atropine, 0.02 mg/kg,
over 30 seconds. Each repetition of the protocol was
completed within 15 minutes. Thus, during the last
From the Department of Medicine, Serafimerlasarettet,
Karolinska Institute, Stockholm, Sweden.
Supported by grants from the Swedish National Association
Against Heart and Chest Diseases and the Serafimer Hospital
Foundation.
Dr. Vallin's present address: Department of Medicine, Huddinge
University Hospital, S-141 86 Huddinge, Sweden.
Dr. Ahnve's present address and address for correspondence:
Department of Medicine, Danderyd's Hospital, S-182 88 Danderyd,
Sweden.
Received July 18, 1980; revision accepted June 4, 1981.
Circulation 65, No. 3, 1982.
435
CI RCULATION
436
part of the study both propranolol and atropine were
active and dual autonomic inhibition was achieved.
QT and RR intervals were measured from five consecutive beats when possible. However, during atrial
pacing at 130 beats/min, the pacing stimulus sometimes fell at the end of the preceding T wave and a QT
interval could be determined only in the last paced
beat. The corrected QT interval (QTc) was calculated
according to Bazett's formula.9" 19 Measurements were
made in the four recorded leads and the longest QT interval was chosen. The QT interval was measured
from the beginning of the QRS complex to the end of
the T wave. No U waves were present in these subjects. The QT-interval measurements were made without knowledge of whether drugs had been given.
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Statistical Methods
Conventional statistical methods were applied. The
t test for paired data was used to compare QT values
at different heart rates during the same treatment, and
vice versa. Rate-induced QT differences between the
various treatments were compared by applying
Scheff&'s procedure for linear contrast, which takes
individual covariation into account when defining
confidence intervals.25 The 5% confidence level was
chosen to indicate a statistically significant difference.
Results
Uncorrected QT Interval in Relation to Heart Rate
The QT interval was significantly shorter during
atrial pacing at 90 beats/min than at spontaneous
heart rate (p < 0.01). A significant additional reduction in QT resulted when heart rate was further increased by pacing at 130 beats/min (p < 0.01). Significant decreases of the QT interval were also seen
after administration of propranolol and again after
the combination of atropine and propranolol (fig. 1,
table 1).
Drug Effect on QT Interval
No significant change in the QT interval was
recorded after propranolol when heart rate was kept
constant at 90 and 130 beats/min by atrial pacing. In
contrast, when atropine was added, the QT interval
shortened significantly at both paced heart rates. It
was also shorter than the value before both drugs at 90
beats/min, while a lesser, nonsignificant difference
was present at a paced heart rate of 130 beats/min
(fig. 1, table 1).
Influence of Autonomic Tone on Rate-related QT Changes
The difference between the QT interval at the two
paced rates was 10% before drugs, 10% after propranolol and 9% after both propranolol and atropine administration. Thus, the magnitude of the changes was
uninfluenced by autonomic tone (fig. 1, table 1).
Changes in Heart Rate and QTc Interval
Relation to Heart Rate and Drugs
in
Propranolol caused
no
significant change in heart
rate, in contrast to the effect of atropine (p < 0.01).
During pacing, heart rate correction of QT (QTc) ac-
VOL 65, No 3, MARCH 1982
QT
sec
0.5000.480
0.4600.440 0.420 0.400 0.380 0.360-
>1 QTC
.~~~/-'1
c
Q
rT
00
.s
0.3400.320 -
..~~~~~~..
I
-4
0.300-0 ,I
I
I
I
I
I
,
I
6570 80 90 100 110 120 130
HR beats/ min
FIGURE 1. Uncorrected QT and corrected QT (QT,) in
relation to heart rate (HR) and drugs. Bars indicate SEM;
dotted line = atropine plus propranolol; dashed line = propranolol; dot-dash line = no drugs.
cording to Bazett's formula resulted in inversion of the
relationship, giving longer values at higher paced
heart rates (fig. 1, table 2). The resulting differences
were
significant.
Discussion
The relation between heart rate and the duration of
the QT interval is well recognized. Higher heart rates
are associated with shorter QT intervals.19-22 The
mathematical definition of this relationship has been
derived from population studies with each individual
contributing one pair of values (i.e., the QT interval
recorded at the spontaneous heart rate). However, this
relationship is not necessarily the same under different physiologic conditions and at different heart rates
in the same subject. Studies of exercise-induced
changes in the QT interval26128 indicated that the workinduced increase in heart rate was generally paralleled by a shortening of the QT interval. The purpose
of these studies was to assess whether QT intervals are
useful in predicting the presence of coronary artery
disease. Some authors found that QT interval measurement may be useful in this respect,28 while others
reported a lack of specificity.26 27 Further, in patients
with long-QT syndromes, exercise may produce
prolongation of the QT interval and alterations of the
T wave, which may trigger syncopal attacks."3
Corresponding changes can also be produced by
437
HEART RATE, AUTONOMIC TONE AND QT INTERVAL/Ahnve and Vallin
TABLE 1. Mean Uncorrected QT in Relation to Heart Rate and Drugs Administered
Drug
influence
Mean heart
rate
(beats/min)
No drug
Spontaneous rhythm
Pacing
Pacing
(Diff 90-130)
65 ± 5*
90
130
Mean QT
(sec)
Drug-induced
difference
p
Rate-induced
difference
p
on ratedependent
difference
p
0.386 ± 0.023*
0.356 ± 0.019
0.322 ± 0.024
0.035 ± 0.026
NS
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Drugs
Propr
Propr/pacing
Propr/pacing
(Diff 90-130)
Propr + Atrop
Propr + Atrop/pacing
Propr + Atrop/pacing
(Diff 90-130)
NS
62 ± 7*
90
130
77 ± 5*
90
130
0.392 ± 0.021
0.365 ± 0.021
0.328 ±0.010
0.038 ± 0.024
0.368 ± 0.020
0.343 ± 0.019
0.312 ± 0.020
0.031 ± 0.014
0
001
0.05
NS
NS
0.01
0.01
0.05
NS
0.01
*Values are mean ± SD.
Abbreviations: Propr = propranolol; Atrop = atropine.
asymmetric alterations of sympathetic nervous tone.
Yanowitz et al.29 lengthened the QT interval in dogs
either'by stimulating the left stellate ganglion or by
ablating the right stellate ganglion. Both prolongation and shortening of the QT interval can be produced by catecholamine administration. Abildskov'3
found that brief nerve stimulation or rapid catecholamine injection in dogs was followed by transient
prolongation of the QT interval, whereas slow catecholamine infusion or prolonged sympathetic nerve
stimulation resulted in no alterations or a reduction of
the QT interval. Sympathomimetic drugs have also
been used therapeutically to decrease liability for ventricular arrhythmias and to shorten a prolonged QT
interval produced by toxic quinidine treatment.4
Whether this beneficial effect is mainly a result of the
acceleration of heart rate or a direct action on the
repolarization of ventricular myocardium is not
known.
Thus, both heart rate and adrenergic activity affect
the duration of the QT interval, but it is not clear to
what extent these effects are interrelated or whether
the heart rate effects are independent of autonomic
tone. Nor, to our knowledge, has the importance of a
direct effect of cholinergic activity been demonstrated.
Rate-related Changes in QT
We used atrial pacing to achieve changes in heart
independent of autonomic tone. The heart rate
was increased in two steps, first to 90 and then to 130
beats/min, each associated with significant shortening of the QT interval. These reductions in QT were
rate
mainly rate-dependent and not significantly influenced
by possible secondary reflex changes in autonomic
tone, because the magnitudes were of the same after 3adrenergic blockade with propranolol and after combined autonomic inhibition by 'propranolol and
atropine. Thus, the dependence of the QT interval on
heart rate appears to be an intrinsic electrophysiologic property of the ventricular myocardium;
the shortening is directly related to the reduction in
ventricular cycle length.
A covariation between the duration of the QT interval and the refractoriness of ventricular myocardium
is well established. The present findings agree with a
shortening of the action potential at higher driving
rates.30 In addition, a reduction of the effective refractory period of the human ventricular myocardium at
higher heart rates has been shown.30 81 In the study of
Guss et al.,30 who also determined QT intervals, higher
pacing rates gave a shortening of the same magnitude
as in the present investigation.
Our'data allow us t o evaluate whether the relation
between spontaneous heart rate and QT: is also valid
for artificially increased heart rates produced by atrial
pacing. A lack of agreement is evident from o'ur finding that calculating QT0 values did not abolish heartrate-dependent differences in QT during atrial pacing
(table 2). Instead, correction produced an overadjustment and in fact reversed the differences between the
uncorrected values. Thus, Bazett's formula cannot be
applied when 'heart rate is increased by short-term
pacing. Corresponding results were obtained in a recent study by Milne et al.32 One reason for the overcorrection may lie in the short-term nature of the
VOL 65, No 3, MARCH 1982
CIRCULATION
438
TABLE 2. Mean Corrected QT (QT,) in Relation to Heart Rate and Drugs Administered
Drug
influence
on rate-
Mean heart
rate
(beats/min)
No drug
Spontaneous rhythm
Pacing
Pacing
(Diff 90-130)
65 ±5*
90
130
Mean QT,
Rate-induced
difference
p
0.402 ± 0.022*
0.439 ± 0.023
0.476 ± 0.034
0.037 ± 0.036
0.01__
(sec)
Drug-induced
difference
p
dependent
difference
p
0.01
NS
Drugs
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Propr
Propr/pacing
Propr/pacing
(Diff 90-130)
NS
62 ± 7*
90
130
0.399 ± 0.026
0.450 ± 0.028
0.486 ± 0.016
0.035 ± 0.031
0.05
0.01
NS
0.01
NS
NS
0.01
Propr + Atrop
Propr + Atrop/pacing
Propr + Atrop/pacing
(Diff 90-130)
*Values are mean ± SD.
Abbreviations: See table 1.
77 ± 5*
90
130
0.415 ± 0.021
0.423 ± 0.022
0.462 ± 0.032
0.039 ± 0.026
effects, as pacing was applied for only 30 seconds. It is
possible that long- and short-term effects do differ, in
keeping with the influence of catecholamine infusion
on the QT interval reported by Abildskov,15'but an
alternative explanation seems more likely. The normal
relation between the'RR and QT intervals is dependent on the parallel influence of autonomic tone on
heart rate and ventricular repolarization. An artificial
increase-in heart rate by atrial pacing abolishes this
natural relationship.
Effect of 3-adrenergic Blockage
In the present study, no significant change in either
QT or QTc occurred after propranolol when heart rate
was held constant at 90 or 130 beats/min. This
absence of a significant change in QT is in agreement
with other studies using acute administration of ,Badr'energic blockade33 84 as well as= long-term
therapy.35 Milne et al.32 investigated the effect of propranolol at various constant heart rates (atrial pacing).
QT was increased in 10 of 15 subjects. In the present
investigation, QT interval tended to be prolonged, but
not significantly. This alteration of the QT interval by
propranolol administration is probably explained by
low initial sympathetic tone. In a study on the effect of
long-term fl-blockade therapy (metoprolol) in myocardial infarct patients,36 heart rate showed a significant change and the QT, interval was significantly
shortened.
Effect of Cholinergic Drug Inhibition
With the heart rate held constant at 90 or 130
beats/min, atropine caused a significant shortening of
0.05
0.01
0.05
NS
the QT interval. This effect was seen when padrenergic blockade had already been instituted. At
the lower paced heart rate, the autonomic inhibition of
QT also differed significantly from that before drugs.
At the higher paced heart rate there was a lesser,
nonsignificant change in the same direction. Thus,
cholinergic inhibition produced a rate-independent
effect on the QT'interval, suggesting a direct influence
on the duration of the combined depolarization and
repolarization of the ventricular myocardium. These
findings support reports that cholinergic receptors are
not limited to the sinus node, AV node and atrial myocardium, but may also be found in the human ventricular myocardium.13 8 The former report also included
a demonstration of enhanced ventricular electrical
stability, an increase in the fibrillation threshold,
mediated by cholinergic fibers in dogs. The reason
why a rate-independent effect on QT was produced'by
atropine but not by propranolol in the present study
probably mainly reflects the autonomic balance at
rest, with cholinergic activity dominating over sympathetic tone.23 37
We conclude that in healthy subjects, the relation
between the RR interval and the QT interval is dependent on the integrated action of several factors.
Alterations in heart rate and inhibition of cholinergic
tone independently affected the duration of the QT interval, but no rate-independent effect of f,-adrenergic
blockade could be demonstrated. Atrial pacing'at constant driving rates provides a sensitive method for
demonstrating direct drug effects on the combined
ventricular depolarization and repolarization time.
The results also confirm that atrial overdrive pacing is
HEART RATE, AUTONOMIC TONE AND QT INTERVAL/Ahnve and Vallin
an efficient way to shorten the QT interval, which
agrees well with the positive experience of therapeutic
overdrive pacing to suppress ventricular tachyarrhythmias that accompany a prolonged QTc interval,
as seen in quinidine intoxication.4 Whether atropine
offers an additional therapeutic effect in these conditions must be investigated.
18.
19.
20.
Acknowledgment
We thank Andreas Sjogren, M.D., for valuable criticism of the
manuscript.
References
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
1. Jervell A, Lange-Nielsen F: Congenital deaf-mutism, functional heart disease with prolongation of the Q-T interval and
sudden death. Am Heart J 54: 59, 1957
2. Romano C, Gemme G, Pongiglione R: Aritmie cardiache rare
dell' eta pediatrica. II. Accessi sincopali per fibillazione ventricolare parossistica. Clin Pediatr 45: 656, 1963
3. Ward OC: A new familial cardiac syndrome in children. J Irish
Med Assoc 54: 103, 1964
4. Kirkler DM, Curry PVL: Torsade de pointes, an atypical ventricular tachycardia. Br Heart J 38: 117, 1976
5. Haynes RE, Hallstrom AP, Cobb LA: Repolarization abnormalities in survivors of out-of-hospital ventricular fibrillation.
Circulation 57: 654, 1978
6. Ahnve S, Lundman T, Shoaleh-var M: The relationship
between QT interval and ventricular arrhythmias in acute myocardial infarction. Acta Med Scand 204: 17, 1978
7. Schwartz PJ, Wolf S: QT interval prolongation as predictor of
sudden death in patients with myocardial infarction. Circulation 57: 1074, 1978
8. Ahnve S, Helmers C, Lundman T, Rehnqvist N, Sjbgren A:
QTc intervals in acute myocardial infarction: first-year prognostic implications. Clin Cardiol 3: 303, 1980
9. Lepeschkin E: Modern Electrocardiography, vol I. Baltimore,
Williams & Wilkins, 1951
10. Reynolds EW, Vander Ark CR: Quinidine syncope and the
delayed repolarization syndromes. Mod Concepts Cardiovasc
Dis 45: 117, 1976
11. Ahnve S, Helmers C, Lundman T: QT, intervals at discharge
after acute myocardial infarction and long-term prognosis.
Acta Med Scand 208: 55, 1980
12. Schwartz PJ, Malliani A: Electrical alternation of the T-wave:
clinical and experimental evidence of its relationship with the
sympathetic nervous system and with the long QT syndrome.
Am Heart J 89: 45, 1975
13. Schwartz PJ, Periti M, Malliani A: The long Q-T syndrome.
Am Heart J 89: 378, 1975
14. Moss AJ, McDonald J: Unilateral cervicothoracic sympathetic
ganglionectomy for the treatment of long Q-T interval syndrome. N Engl J Med 285: 903, 1971
15. Abildskov JA: Adrenergic effects on the QT interval of the electrocardiogram. Am Heart J 92: 210, 1976
16. Kliks BR, Burgess MJ, Abildskov JA: Influence of sympathetic
tone on ventricular fibrillation threshold during experimental
coronary occlusion. Am J Cardiol 36: 45, 1975
17. Schwartz PJ, Snebold NG, Brown AM: Effects of unilateral
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
439
cardiac sympathetic denervation on the ventricular fibrillation
threshold. Am J Cardiol 37: 1034, 1976
Kent KM, Epstein SE, Cooper T, Jacobowitz DM: Cholinergic
innervation of the canine and human ventricular conducting
system: anatomic and electrophysiologic correlations. Circulation 50: 948, 1974
Bazett HC: An analysis of the time-relations of electrocardiograms. Heart 7: 353, 1920
Fridericia LS: Die Systolendauer im Elektrokardiogramm bei
normalen Menschen und bei Herzkranken. Acta Med Scand
53: 469, 1920
Ashman R: The normal duration of the QT interval. Am Heart
J 23: 522, 1942
Adams W: The normal duration of the electrocardiographic
ventricular complex. J Clin Invest 15: 335, 1936
Vallin HO: Autonomous influence on sinus node and AV node
function in the elderly without significant heart disease: assessment with electrophysiological and autonomic tests. Cardiovasc
Res 14: 206, 1980
Jonsell S: A method for the determination of the heart size by
teleroentgenography (a heart volume index). Acta Radiol 20:
325, 1939
Morrisson DF: Multivariate Statistical Methods. New York,
McGraw-Hill, 1967
Kemp GL: The Q-T ratio in treadmill stress testing. Angiology
22: 523, 1971
Roman L, Bellet S: Significance of the QX/QT ratio and the
QT ratio (QTr) in the exercise electrocardiogram. Circulation
32: 435, 1965
Greenberg PS, Friscia DA, Ellestad MH: Predictive accuracy
of Q-X/Q-T ratio, Q-Tc interval, S-T depression and R wave
amplitude during stress testing. Am J Cardiol 44: 18, 1979
Yanowitz F, Preston JB, Abildskov JA: Functional distribution of right and left stellate innervation to the ventricles.
Production of neurogenic electrocardiographic changes by unilateral alteration of sympathetic tone. Circ Res 18: 416, 1966
Guss SB, Kastor JA, Josephson ME, Scharf D: Human ventricular refractoriness: effects of cycle length, pacing site and
atropine. Circulation 53: 450, 1976
Olsson B, Brorson L, Harper R, Ryden L: Estimation of ventricular refractoriness in man by the extra stimulus method.
Cardiovasc Res 11: 31, 1977
Milne JR, Camm AJ, Ward DE, Spurrell RAJ: Effect of intravenous propranolol on QT interval: a new method of assessment. Br Heart J 43: 1, 1980
Stern S, Eisenberg S: The effect of propranolol (Inderal) on the
electrocardiogram of normal subjects. Am Heart J 77: 192,
1969
Seides SF, Josephson ME, Batsford WP, Weisfogel GM, Lau
SH, Damato AN: The electrophysiology of propranolol in man.
Am Heart J 88: 733, 1974
Nyberg G, Vedin A, Wilhelmsson C: QT time in patients
treated with alprenolol or placebo after myocardial infarction.
Br Heart J 41: 452, 1979
Ahnve S, Erhardt L, Lundman T, Rehnqvist N, Sj6gren A:
Effect of metoprolol on QT, intervals after actue myocardial infarction. Acta Med Scand 208: 223, 1980
Anderson M, del Castillo J: Cardiac innervation and synaptic
transmission in heart. In Electrical Phenomena in the Heart,
edited by De Mello WC. New York and London, 1972, p 235
Influence of heart rate and inhibition of autonomic tone on the QT interval.
S Ahnve and H Vallin
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Circulation. 1982;65:435-439
doi: 10.1161/01.CIR.65.3.435
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1982 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://circ.ahajournals.org/content/65/3/435
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally
published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further
information about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/