DEPRESSION
AND
ANXIETY 26:251–258 (2009)
Research Article
EFFECT OF MEDICATION AND PSYCHOTHERAPY ON
HEART RATE VARIABILITY IN PANIC DISORDER
Amir Garakani, M.D.,1 Jose M. Martinez, M.A.,1 Cindy J. Aaronson, M.S.W. Ph.D.,1 Andrei Voustianiouk, Ph.D.,2
Horacio Kaufmann, M.D.,2 and Jack M. Gorman, M.D.3
Background: Panic disorder (PD) patients have been shown to have reduced
heart rate variability (HRV). Low HRV has been associated with elevated risk
for cardiovascular disease. Our aim was to investigate the effects of treatment on
heart rate (HR) in patients with PD through a hyperventilation challenge.
Methods: We studied 54 participants, 43 with Diagnostic and Statistical
Manual of Mental Disorders (DSM-IV) PD and 11 controls. Subjects lay supine
with their heads in a plastic canopy chamber, resting for 15 min and then
breathing at a rate of 30 breaths per minute for 10 min. HRV was sampled for
spectral analysis. Clinical and behavioral measures of anxiety were assessed.
Treatment was chosen by patients: either 12 weeks of CBT alone or CBT with
sertraline. Results: All patients showed significant decrease on clinical measures
from baseline and 31 were treatment responders, 8 dropped out of the study
before completion of the 12-week treatment phase and 4 were deemed
nonresponders after 12 weeks of treatment. Although both treatments led to
significant clinical improvement, only CBT alone demonstrated a significant
reduction in HR and increase in HRV. Conclusions: Our study replicated the
finding that increased HR and decreased HRV occur in PD patients. Given the
evidence of cardiac risk related to HRV, CBT appears to have additional benefits
beyond symptom reduction. The mechanisms of this difference between CBT and
sertraline are unclear and require further study. Depression and Anxiety
26:251–258, 2009.
r 2008Wiley-Liss, Inc.
Key words: anxiety; cognitive behavioral therapy; SSRI; cardiac; life stress;
breathing
INTRODUCTION
Panic disorder (PD), with a lifetime prevalence of
4.7%, continues to remain a significant health problem.[1] Expert guidelines call for the use of either
selective serotonin reuptake inhibitors (SSRIs) or
cognitive behavioral therapy (CBT), as both have been
reported to be effective in the treatment of PD.[2] A
large-scale study found that combination treatment of
CBT plus medication was as effective when compared
to medication or CBT alone after acute treatment.[3] It
1
Department of Psychiatry, Mount Sinai School of Medicine,
New York, New York
2
Department of Neurology, New York University School of
Medicine, New York, New York
3
Comprehensive NeuroScience, Inc., White Plains, New York
r 2008 Wiley-Liss, Inc.
has been reported, however, that only 11% of patients
with PD receive an SSRI,[4] and not many more receive
some form of CBT.[5]
Contract grant sponsor: NIMH; Contract grant number: RO1
071827-01A2; Contract grant sponsor: Mount Sinai School of
Medicine, General Clinical Research Center; Contract grant
number: MO1-RR-00071
Correspondence to: Amir Garakani, M.D., Department of
Psychiatry, Mount Sinai Hospital, One Gustave L. Levy Place,
Box 1230, New York, NY 10029..
E-mail: amir.garakani@mssm.edu
Received for publication 23 May 2008; Revised 14 August 2008;
Accepted 19 August 2008
DOI 10.1002/da.20533
Published online 6 October 2008 in Wiley InterScience (www.
interscience.wiley.com).
252
Garakani et al
Another pressing concern is evidence that links
anxiety disorders, in particular PD, with an increased
risk of coronary artery disease and sudden cardiac
death.[6–11] A reduction in heart rate variability (HRV), a
measure of rate differences in the beat-to-beat time
interval, has been found to be an independent risk factor
for mortality in patients with coronary artery disease
(CAD).[12–18] Although partly due to effects of circulating catecholamines, centrally mediated direct innervation to the heart is the primary factor regulating HRV.
In general, parasympathetic input to the heart increases
HRV and sympathetic input reduces it. Parasympathetic
influences over cardiac function, carried via the vagus
nerve from brain to heart, are reflected in the highfrequency (HF) component of HRV.[19] Although lowfrequency (LF) component is considered as a parameter
of sympathetic and vagal influences, an increase in the
ratio of LF to HF signifies an increase in sympathetic
influences over cardiac function and is believed to put
the heart at risk for arrhythmia.[20,21] It has been
reported that patients with PD, in addition to having a
reduced HRV and HF power, demonstrate increased LF
power, and an increased LF/HF ratio.[22] Yeragani’s
group reported that panic subjects have higher standing
mid-frequency power and higher mid-frequency power
than normal comparison subjects when challenged by
yohimbine, an a-2 adrenergic antagonist,[23,24] and PD
patients had a higher sympathovagal ratio than controls
after challenge with isoproterenol, a b-adrenergic
agonist.[25] Contrary to the above findings, it has also
been reported that patients with PD and comorbid CAD
have a lower LF/HF ratio, suggesting a lower sympathetic modulation, although the utility of the LF/HF has
been a matter of some controversy.[22,26] It should also
be noted that respiratory rate can influence sympathetic
activation via chemoreflexes, and states of rapid breathing, such as hyperventilation, can cause a temporary
increase in HF power and a long-term reduction in
HRV.[27]
A study of 20 cardiac transplant patients and healthy
controls found that transplant patients, in whom the
graft heart is denervated and has virtually no HRV, had
a blunted heart rate response to a stressful mental
arithmetic test when compared to controls.[28] This
indicates the major role that direct central nervous
system innervation of the heart plays in controlling the
cardiac response to psychologically stressful stimuli. A
number of studies have reported reduced HRV in
patients with anxiety disorder, including PD.[24,26,29–33]
Studies of the effects of antidepressant medications
or CBT on HRV are limited. An open-labeled study
using mirtazapine in PD found that nonresponders to
treatment had lower HRV and reduced LF power, but
there was no difference in LF/HF ratio between the
groups.[34] Carney et al. [35] tested 30 patients with
CAD and depression, classified as mild, moderate, or
severe, who received 16 sessions of CBT. After CBT
treatment, there was a decrease in HR and an increase
in HRV in the severely depressed group but no effect
Depression and Anxiety
was seen in mildly or nondepressed groups. One study
compared HRV in 17 PD patients treated with
paroxetine to 16 unmedicated controls, and found an
increase in mean parasympathetic components of HRV
after 4 weeks on medication.[36]
Our group published the only study comparing the
effects of an SSRI medication with CBT on HRV in
patients with PD versus healthy volunteers,[37] which is
a subset of the sample presented in this article. In a
noncontrolled pilot sample of 16 patients with PD,
during a hyperventilation challenge patients had
decreased HRV and increased QT interval variability,
and that after treatment of a smaller sample of patients
with PD with sertraline plus CBT (n 5 4) or CBTalone
(n 5 4), there was increased HRV and decreased QT
interval variability in all patients.
The objective of our study is twofold: first, to
determine whether hyperventilation, a common occurrence during panic attacks, alters heart rate and/or
HRV (HR/HRV) differently in PD patients than
healthy control subjects and in a manner suggestive
of an increased cardiac risk; second, to determine
whether treatment alters HR/HRV toward normal
control values.
METHODS
SUBJECTS
This study began at Columbia University/New York State
Psychiatric Institute (NYSPI) and was completed at Mount Sinai
School of Medicine and was approved by the Institutional Review
Board of both institutions. Baseline heart rate data was obtained in 43
patients with PD (24 male, 19 female) and 11 healthy controls
(7 male, 4 female). Mean age for the patient group was 34.4 years
(SD 5 10.2) and 29.0 years (SD 5 8.1) for the control group. Of the
43 PD patients, 19 also had agoraphobia. Criteria for study
participation included: adults aged 18–65 with Diagnostic and
Statistical Manual of Mental Disorders (DSM-IV) PD (with or without
agoraphobia), at least one panic attack per week in the 4 weeks earlier
to enrollment; and/or a mobility inventory score of 3.3 (alone) or 2.5
(accompanied), and without an unstable medical condition. Exclusion
criteria included current posttraumatic stress disorder, substance
abuse or dependence within 6 months of study entry, Hamilton
Depression Scale (HAM-D) score415, lifetime history of schizophrenia, bipolar disorder, obsessive compulsive disorder, or any
eating disorder. Subjects were also excluded if they were effectively
medicated for PD. All enrolled subjects were psychotropic medication- free for at least 2 weeks (4 weeks for fluoxetine) before the
baseline visit. Healthy control subjects were screened for all Axis I
disorders and excluded if they had any history of psychiatric illness or
unstable medical condition.
PROCEDURE
ASSESSMENTS
All subjects, after explanation of study procedures,
signed informed consent. Diagnostic evaluation consisted of psychiatric assessment by the study psychiatrist and the administration of the SCID[38] by a trained
clinical interviewer. After passing screening interviews,
Research Article: Panic Treatment and Heart Rate Variability
all participants underwent a physical exam, including a
medical history, routine blood tests, and electrocardiogram.
Clinical measures administered at baseline and
posttreatment included (1) the Clinical Global Impression Scale (CGI),[39] (2) Panic Disorder Severity Scale
(PDSS),[40] a 7-item scale providing ratings of the core
features of PD and the degree of impairment, (3)
Hamilton Rating Scale for Anxiety (HAM-A),[41] and
(4) HAM-D.[42]
These assessments were administered during the
hyperventilation challenge: the Acute Panic Inventory
(API),[43] a 27-item inventory of physical sensations
frequently experienced during panic attacks; the Borg
Breathlessness Scale,[44] a 10-point measure of dyspnea;
and the Anxiety Scale,[45] a 10-point Likert-type scale
measuring anxious feelings.
The Mobility Inventory (MI) for Agoraphobia ,[46] a
self-report instrument, which assesses the level of
avoidance of many situations both alone and accompanied due to fear of a panic attack, was also collected
at baseline and posttreatment as was the Survey of
Recent Life Experiences Scale (SRLE),[47] which
assesses mundane stressors or ‘‘hassles.’’
EXPERIMENTAL PROCEDURES
Participants lay supine with their heads in a plastic
canopy chamber, allowing both visual and auditory
contact. Room air was vented through the chamber at a
rate of 20 l/min, to prevent the accumulation of CO2.
After resting quietly for 15 min in the canopy system,
participants were instructed to breathe at a rate of 30
breaths per minute for 10 min, paced by a metronome.
Continuous heart rate was sampled at 1,000 times per
second for spectral analysis throughout the procedure.
The API, Borg, and Anxiety Scales were rated at the
end of the 15-min rest period and again after the 10min hyperventilation period. After the 12-week treatment, patients were retested in the canopy system.
Controls were not retested.
TREATMENT
At the outset of the longitudinal study, after baseline
challenge, all patients were given both sertraline and
CBT. When several patients requested CBT without
medication, we elected to give patients the choice
between CBT alone or with sertraline.
Sertraline was initiated at 25 mg daily for the first
week and then increased to 50 mg daily. At the end of 4
weeks, if response was judged inadequate, dose was
raised by 50 mg increments weekly to a maximum of
200 mg. Pharmacotherapy was conducted by the study
psychiatrist with patients seen weekly for 30 min.
CBT was conducted by a doctoral level social worker
or psychologist weekly in sessions lasting 45 min for 12
weeks. The type of CBT used in this study was Panic
Control Therapy[48] and included psychoeducation,
breathing retraining, progressive muscle relaxation,
253
cognitive restructuring, interoceptive exposure, and
situational exposure.
DATA ANALYSIS
Electrocardiogram (ECG) data were analyzed with
custom-written software using PV-WAVE programming language [Visual Numerics, Houston, TX].
R-peaks in the ECG waveform were detected by an
automated detection algorithm with artifact elimination and subsequently verified by visual inspection. Fast
Fourier transformation was used on a series of interbeat intervals to provide information over a range of
frequencies. Frequency ranges for spectral analyses
were set according to the guidelines specified in the
report by the Task Force of the European Society of
Cardiology and the North American Society of Pacing
and Electrophysiology.[49] The HRV spectrum contains two major components: the LF range
(0.04–0.15 Hz) and the HF range (0.15–0.4 Hz).
Six heart rate variables were computed; mean R to R
wave interval (mean-RRI) measured in milli-seconds
the standard deviation of mean RRI; PNN50 (%;
proportion of adjacent R waves more than 50 msec);
different LF component of HRV; HF component of
HRV; and the ratio of LF/HF of HRV.
STATISTICAL ANALYSIS
To compare HR/HRV measures between patients
and controls, t tests were performed for both rest and
hyperventilation conditions for each cardiac measure.
To test whether there were gender differences in any
HR/HRV measure, t tests were performed for men and
women separately within group. We calculated Pearson
correlations between age and all HR/HRV measures
for each group separately to test whether age had any
effects in our groups. Repeated measures analysis of
variance (RM-ANOVA) was performed on each HR/
HRV measure with time (pretreatment and posttreatment) as the within-groups factor and treatment type
(CBT alone and CBT plus sertraline) as the betweengroups factor to test treatment differences in our
sample.
To assess the treatment effect on the three rating
scales obtained during the challenge, paired t-tests
were performed on scale scores comparing pretreatment to posttreatment scores. Pearson correlations
were calculated between pretreatment rating scale
scores and both the clinical measures and the baseline
HR/HRV measures. This would reveal if any relationships existed between the level of anxiety during the
canopy tests and any of the HR/HRV measures and/or
the baseline severity of clinical illness.
Paired t-tests were performed between pretreatment
and posttreatment clinical measures to test whether
there was significant improvement. To test for treatment differences, RM-ANOVA was performed on
each clinical measure between the two time points
Depression and Anxiety
254
Garakani et al
(pretreatment and posttreatment) by the two treatments (CBT and CBT plus medication).
RESULTS
(t 5 2.12, df 5 49, Po.04), and higher LF component
of HRV (t 5 3.69, df 5 50, Po.001) than controls
(Tables 1 and 2).
BEHAVIORAL ASSESSMENTS
Gender cross tabulation showed no gender distribution differences between the two groups, nor were
there significant gender differences in any HR/HRV
measure or any significant correlation between HR/
HRV measures and age. At pretreatment, during quiet
rest period, patients had significantly higher HR
(t 5 4.77, df 5 52, Po.001) and lower HRV
(t 5 3.14, df 5 52, Po.003) than healthy controls.
During the hyperventilation period, patients had
higher HR (t 5 4.60, df 5 50, Po.001), lower HRV
In the patient group, paired t-tests revealed significant treatment effects on all behavioral assessments
during both the baseline rest period (Table 3) and the
hyperventilation period (Table 4). RM-ANOVA revealed only a time effect with no time-by-treatment
interactions (results not presented). Pearson correlations revealed significant relationships between the LF
power component of HR and pretreatment baseline
API (r 5 .42, Po.001, n 5 41) and significant negative
relationships between both greater breathlessness score
TABLE 1. Heart rate and spectral measures during 15 min baseline rest period
Measure
Group
Na
Mean
Std. Deviation
Statistics
Mean_rri (msec)
PD
NC
PD
NC
PD
NC
PD
NC
PD
NC
PD
NC
43
11
43
11
43
11
42
11
42
11
42
11
877.90
1093.552
73.81
89.03
24.24
43.88
0.37
0.28
0.24
0.20
2.32
2.54
134.60
130.535
43.39
37.63
17.62
21.85
0.14
0.18
0.14
0.17
1.91
2.55
t 5 4.769; Po.001; df 5 52
SD_rri (msec)
PNN50 (%)
LF/Total
HF/Total
LF/HF
NS
t 5 3.141; Po.003; df 5 52
NS
NS
NS
Mean_rri (msec), mean r to r interval measured in milli-seconds; SD_rri (msec), standard deviation of r to r interval measured in milli-seconds;
PNN50 (%), the proportion of adjacent R waves more than 50 milli-seconds; LF/Total, low frequency (0.075–0.125 Hz) divided by total HR
power; HF/Total, high frequency (0.125–0.50 Hz) divided by total HR power; LF/HF, ratio of low-frequency to high-frequency power; NS, not
significant.
a
Some data from one subject with panic disorder could not be analyzed due to motion artifact.
TABLE 2. Heart rate and spectral measures during 10 min hyperventilation period
Measure
Group
Na
Mean
Std. Deviation
Statistics
Mean_rri(msec)
PD
NC
PD
NC
PD
NC
PD
NC
PD
NC
PD
NC
41
11
41
11
40
11
41
11
41
11
41
11
853.70
1055.79
65.52
75.45
20.70
35.38
0.38
0.24
0.16
0.18
3.82
2.74
131.04
122.76
62.04
58.80
20.07
21.24
0.11
0.13
0.12
0.12
2.69
3.26
t 5 4.600; Po.001; df 5 50
SD_rri(msec)
PNN50 (%)
LF/Total
HF/Total
LF/HF
NS
t 5 2.124; Po.039; df 5 49
t 5 3.689; Po.001; df 5 50
NS
NS
Mean_rri (msec), mean r to r interval measured in milli-seconds; SD_rri (msec), standard deviation of r to r interval measured in milli-seconds;
PNN50 (%), the proportion of adjacent R waves more than 50 milli-seconds; LF/Total, low frequency (0.075–0.125 Hz) divided by total HR
power; HF/Total, high frequency (0.125–0.50 Hz) divided by total HR power; LF/HF, ratio of low-frequency to high-frequency power; NS, not
significant.
a
Data from two subjects with panic disorder could not be analyzed due to motion artifact.
Depression and Anxiety
Research Article: Panic Treatment and Heart Rate Variability
255
TABLE 3. Behavioral assessment during 15 min rest
period pretreatment and posttreatment for all patients
(N 5 28)
Pretreatment
Posttreatment
Results
Measures
Mean
SD
Mean
SD
t
Df
P
API
BORG
Anxiety
6.89
0.88
2.11
7.68
1.32
1.62
1.68
0.41
1.29
4.32
1.13
0.01
3.96
2.28
3.07
27
27
27
.001
.031
.005
SD, standard deviation; API, Acute Panic Inventory; BORG, Borg
Breathlessness Scale; Anxiety, 10-point Anxiety Scale.
TABLE 4. Behavioral assessment during 10 min
hyperventilation period pretreatment and posttreatment
for all patients (N 5 29)
Pretreatment
Posttreatment
Results
Measures
Mean
SD
Mean
SD
t
df
P
API
BORG
Anxiety
13.41
2.33
3.41
10.66
2.15
2.11
5.48
1.43
2.07
6.31
1.73
1.75
4.70
3.38
2.99
28
28
28
.001
.002
.006
SD, standard deviation; API, Acute Panic Inventory; BORG, Borg
Breathlessness Scale; Anxiety, 10-point Anxiety Scale.
(r 5 .33, Po.04, n 5 40) and greater anxiety score
(r 5 .34, Po.03, n 5 40) and lower HRV (PNN50).
TREATMENT EFFECTS
Of the 43 patients, 20 chose CBT and 23 chose
combination CBT and sertraline. Of the 19 female
patients, more chose CBT (n 5 14), whereas of the 24
male patients, more chose the combined treatment
(n 5 18), a significant difference (w2 5 10.10, Po.002).
Eight patients dropped out of the study, four from the
CBT group and four from the combination CBT/
sertraline group. Thirty-five patients completed the 12
weeks of treatment; 31 were treatment responders at
postacute treatment, which was determined by a 40%
decrease in PDSS score and a CGI-S of 1 or 2 (not ill
or borderline ill). The response criteria we used were
similar to that of the Barlow et al. study.[3] Of these 31
responders, 25 had valid repeat posttreatment HR
measures. All patients had a significant decrease in all
clinical measures (HAM-A, HAM-D, PDSS, CGI-S,
MI, SRLE). RM-ANOVA for treatment effects on HR/
HRV measures were not significant for any group
effects comparing CBT alone to CBT plus sertraline,
but there was a trend toward a significant group-bytime interaction such that patients who chose CBT
treatment had an increase in their HRV whereas
patients who chose the combination treatment had
relatively no change (Po.071). However, this analysis
became significant when looking at treatment responders only (F 5 5.12; Po.037; df 5 1,17) (Fig. 1). RM-
Figure 1. Heart rate variability (PNN50) at baseline and after
cognitive behavioral therapy (CBT) or CBT and sertaline
treatment. Statistics: repeated measure ANOVA group by time:
F 5 4.730; Po.045; df 5 1,16.
ANOVA performed on the clinical scales revealed no
significant group effects or group-by-time interactions,
but there was a significant time effect for all the clinical
scales (Table 5).
Correlation coefficients between pretreatment clinical measures and pretreatment HR/HRV measures
revealed a negative relationship between PDSS scores
and mean-RRI. As a lower RRI represents higher HR
patients with higher pretreatment HR during both the
resting period (r 5 .32, Po.036, n 5 43) and the
hyperventilation period (r 5 .36, Po.020, n 5 41)
had higher PDSS scores. There was also a significant
negative correlation between RRI and the MI index
(when alone) at baseline (r 5 .44, Po.005, n 5 40) and
during hyperventilation (r 5 .43, Po.007, n 5 38).
Greater baseline CGI severity was also associated with
higher HR during hyperventilation (r 5 .34, Po.030,
n 5 41). Greater life stress scores were associated with
an increased ratio of the LF/HF component of HRV
during hyperventilation (r 5 .36, Po.035, n 5 34).
DISCUSSION
Consistent with previous findings, we found that
patients with PD have higher HR and lower HRV than
normal controls, both at rest and during hyperventilation. Furthermore, we found a number of associations
between clinical severity of PD and abnormal central
nervous system regulation of cardiac rhythm. The
relationship between higher life stressors and higher
anxiety symptoms at baseline with higher LF/HF and
LF component of HR may suggest a greater sympathetic tone in these patients. Similarly the significant
correlations between higher HR, lower HRV with
greater severity of illness and panic symptoms are
Depression and Anxiety
256
Garakani et al
TABLE 5. Clinical measures for CBT and CBT plus medication groups before and after treatment using M-ANOVA
CBT
Pretreatment
CBT
Posttreatment
Pretreatment
Results time effects
Posttreatment
Measure
N
Mean
SD
Mean
SD
N
Mean
SD
Mean
SD
F
P
CGI-S
MI Accomp. Alone
16
13
16
11
16
11
0.63
16.29
22.74
9.21
3.71
2.86
15.88
2.00
29.77
37.23
5.69
3.64
3.38
53.73
0.63
3.88
17.75
4.41
2.91
1.71
12.97
20
17
Ham-A
Ham-D
PDSS
SRLE
4.44
39.19
53.19
22.13
11.00
13.94
65.55
4.45
36.78
48.18
18.19
10.71
12.58
57.75
0.94
11.47
22.01
9.93
5.85
4.27
15.84
1.45
29.44
31.76
3.63
2.14
2.21
45.13
0.60
7.20
10.96
2.96
2.21
1.84
5.75
344.51
15.33
29.87
92.33
67.12
204.40
20.57
.001
.001
.001
.001
.001
.001
.001
16
14
19
16
CBT, Cognitive Behavioral Therapy; SD, standard deviation; CGI-S, Clinical Global Impression—Severity; MI, Mobility Inventory for
Agoraphobia, Accompanied, and Alone; HAM-A, Hamilton Anxiety Scale; HAM-D, Hamilton Depression Scale; PDSS, Panic Disorder Severity
Scale; SRLE, Survey of Recent Life Experiences.
consistent with other reports that have reported that
higher stressors and/or anxiety or worry were associated with higher HR and lower HRV.[50,51] Taken
together, these findings are consonant with previous
speculations that changes in the brain’s control over
heart rhythm may play a role in increasing the risk for
cardiac morbidity and mortality among patients with
PD.
We also found that treatment with CBT alone
resulted in a statistically significant decrease in HR
and increase in HRV. Treatment with combined CBT
and sertraline, however, did not produce a change in
HRV or HR despite the fact that both treatments
resulted in equivalent clinical improvement. We
suggest that CBTworks directly on symptom reduction
(i.e., palpitations and hyperventilation) through the
techniques of breathing retraining and progressive
muscle relaxation. In addition, cognitive restructuring,
the changing of overexaggerated fearful cognitions to
more realistic ones, has the effect of altering a pattern
of frequent engagement of the sympathetic response to
fear thoughts. When treated patients no longer
perceive harm or threat from benign situations, they
will not experience attacks of palpitations, difficulty
breathing, chest tightness, or other sensations (i.e.,
parathesia, sweating, abdominal distress). Interestingly,
one study reported that employees with work-related
stress showed significantly higher stress-related symptoms and higher LF component of HR than nonstressed workers and that these alterations were
reverted by a stress management program.[52] The
finding that the CBT group had improvement in HRV,
but the combination treatment did not, is unexpected
because the combined treatment group also experienced the same techniques of breathing retraining,
progressive muscle relaxation, and cognitive restructuring. This outcome may be explained by the findings of
Barlow et al.,[3] where after treatment discontinuation,
the combined imipramine and CBT group had the
equivalent relapse rates as the imipramine only group,
considerably higher than the CBT only and CBT plus
Depression and Anxiety
placebo groups. The medication blocks the panic
sensations before the patient has the opportunity to
learn how to control those sensations with the CBT
techniques. The inhibition of learning may interfere
with the physiological benefits of CBT in PD treatment. The mechanism by which sertraline may somehow block the effects of CBT on HRV is not obvious
and requires further study.
Siepmann et al. [53] investigated the effects of
sertraline on autonomic and cognitive function in a
randomized, placebo controlled, crossover study. They
gave 12 healthy male volunteers 50 mg of sertraline and
placebo, each for 14 days (with a 14-day washout
period in between). They found that sertraline caused a
significant reduction in HR, but not in HRV. The
reduction in HR was consistent with findings with
other SSRIs [54–56] and may be linked to an inhibition
of central sympathetic activity.
There are several limitations of our study. First, there
was no sertraline only group of PD patients to which
we could compare the combined and CBT alone data.
As patients were permitted to choose whether to take
medication in addition to CBT, this may lead to the
problem of selection bias. More men in our study chose
the combined treatment, although our statistical
analysis did not show gender to be a significant factor.
Whether patients who select psychological therapy
over medications have a higher internal locus of
control, and thereby are more likely to be more
responsive to CBT over medication, is an area of
ongoing investigation.[57] Second, the sample is relatively small, although obviously large enough to detect
fairly robust differences. Third, we did not retest the
healthy controls after the 3-month period to rule out
that the changes were attributable to time.
CONCLUSION
Our study replicates the finding of decreased HRV in
patients with PD compared to controls, and a
treatment-related increase in HRV. These findings
Research Article: Panic Treatment and Heart Rate Variability
suggest that the increased risk for cardiovascular
morbidity and mortality seen among patients with
mood and anxiety disorders may be mediated in part by
changes in CNS control of cardiac rhythm and that
these changes may be modifiable by psychosocial
interventions. Further investigation is necessary, comparing a medication alone to CBT alone on these
important cardiac measures.
Acknowledgments. This study was supported in
part by NIMH grant RO1 071827-01A2 (J.M.G.) and
grant MO1-RR-00071, Mount Sinai School of Medicine, General Clinical Research Center. Its contents
are solely the responsibility of the authors and do not
necessarily represent the official views of the NIMH.
We would like to thank Cari Kessler, Paul Toth, and
Jessica Sohl (all formerly of Mount Sinai School of
Medicine), and Jan Mohlman, Ph.D. (formerly of
Columbia University) for their tireless work and
contributions to this project, and extend gratitude to
all subjects for their participation in this study.
REFERENCES
1. Kessler RC, Chiu WT, Jin R et al. The epidemiology of panic
attacks, panic disorder, and agoraphobia in the National
Comorbidity Survey Replication. Arch Gen Psychiatry
2006;63:415–424.
2. Mitte K. A meta-analysis of the efficacy of psycho- and
pharmacotherapy in panic disorder with and without agoraphobia. J Affect Disord 2005; 88:27–45.
3. Barlow DH, Gorman JM, Shear MK, Woods SW. Cognitivebehavioral therapy, imipramine, or their combination for panic
disorder. A randomized controlled trial. JAMA 2000;283:
2529–2536.
4. Bruce SE, Vasile RG, Goisman RM et al. Are benzodiazepines
still the medication of choice for patients with panic disorder
with or without agoraphobia? Am J Psychiatry 2003;160:
1432–1438.
5. Goisman RM, Warshaw MG, Keller MB. Psychosocial treatment prescriptions for generalized anxiety disorder, panic
disorder, and social phobia, 1991–1996. Am J Psychiatry 1999;
156:1819–1821.
6. Coryell W, Noyes R, Clancy J. Excess mortality in panic
disorder: a comparison with primary unipolar depression. Arch
Gen Psychiatry 1982;39:701–703.
7. Haines AP, Imeson JD, Meade TW. Phobic anxiety and
ischaemic heart disease. Br Med J (Clin Res Ed) 1987;295:
297–299.
8. Weissman M, Markowitz JS, Ouellette R et al. Panic disorder
and cardiovascular/cerebrovascular problems: results from a
community survey. Am J Psychiatry 1990;147:1504–1508.
9. Gomez-Caminero A, Blumentals WA, Russo LJ et al. Does panic
disorder increase the risk of coronary heart disease? A cohort
study of a national managed care database. Psychosom Med
2005; 67:688–691.
10. Kawachi I, Sparrow D, Vokonas PS. Symptoms of anxiety and
risk of coronary heart disease: the Normative Aging Study.
Circulation 1994; 90:2225–2229.
11. Smoller JW, Pollack MH, Wassertheil-Smoller S et al. Panic
attacks and risk of incident cardiovascular events among
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
257
postmenopausal women in the Women’s Health Initiative
Observational Study. Arch Gen Psychiatry 2007;64:1153–1160.
Kleiger RE, Miller JP, Bigger JT, Moss AJ. Decreased heart rate
variability and its association with increased mortality after acute
myocardial infarction. Am J Cardiol 1987;59:256–262.
Molgaard H, Sorensen KE, Bjerregaard P. Attenuated 24-hr
heart rate variability in apparently healthy subjects, subsequently
suffering sudden cardiac death. Clin Auton Res 1991;1:233–237.
Bigger JT, Fleiss JL, Steinman RC et al. Frequency domain
measures of heart period variability and mortality after myocardial infarction. Circulation 1992;85:164–171.
Bigger JT, Fleiss JL, Steinman RC et al. Correlations among
time and frequency domain measures of heart period variability
two weeks after acute myocardial infarction. Am J Cardiol
1992;69:891–898.
Tsuji H, Larson MG, Venditti Jr FJ et al. Impact of reduced heart
rate variability on risk for cardiac events: the Framingham Heart
Study. Circulation 1996;94:2850–2855.
Liao D, Cai J, Rosamond WD et al. Cardiac autonomic function
and incident coronary heart disease a population-based casecohort study. The ARIC study. Atherosclerosis risk in communities study. Am J Epidemiol 1997;145:696–706.
Evrengul H, Tanriverdi H, Kose S et al. The relationship
between heart rate recovery and heart rate variability in
coronary artery disease. Ann Noninvasive Electrocardiol 2006;
11:154–162.
Pagani M, Lombardi F, Guzzetti S et al. Power spectral analysis
of heart rate and arterial pressure variabilities as a marker of
sympatho-vagal interaction in man and conscious dog. Circ Res
1986;59:178–193.
Malliani A, Lombardi F, Pagani M, Cerutti S. Power spectral
analysis of cardiovascular variability in patients at risk for sudden
cardiac death. J Cardiovasc Electrophysiol 1994;5:274–286.
Piccirillo G, Elvira S, Bucca C et al. Abnormal passive head-up
tilt test in subjects with symptoms of anxiety power spectral
analysis study of heart rate and blood pressure. Int J Cardiol
1997;60:121–131.
Lavoie KL, Fleet RP, Laurin C et al. Heart rate variability in
coronary artery disease patients with and without panic disorder.
Psychiatry Res 2004;128:289–299.
Yeragani VK, Berger R, Pohl R et al. Effects of yohimbine on
heart rate variability in panic disorder patients and normal
controls: a study of power spectral analysis of heart rate.
J Cardiovasc Pharmacol 1992;20:609–618.
Yeragani VK, Pohl R, Berger R et al. Decreased heart rate
variability in panic disorder patients: a study of power-spectral
analysis of heart rate. Psychiatry Res 1993;46:89–103.
Yeragani VK, Pohl R, Srinivasan K et al. Effects of isoproterenol
infusions on heart rate variability in patients with panic disorder.
Psychiatry Res 1995;56:289–293.
McCraty R, Atkinson M, Tomasino D, Stuppy WP. Analysis of
twenty-four hour heart rate variability in patients with panic
disorder. Biol Psychol 2001;56:131–150.
Guzik P, Piskorski J, Krauze T et al. Correlations between
the Poincaré plot and conventional heart rate variability
parameters assessed during paced breathing. J Physiol Sci 2007;
57:63–71.
Shapiro PA, Sloan RP, Bigger Jr JTet al. Cardiac denervation and
cardiovascular reactivity to psychological stress. Am J Psychiatry
1994;151:1140–1147.
Klein E, Cnaani E, Harel T et al. Altered heart rate variability in
panic disorder patients. Biol Psychiatry 1995;37:18–24.
Sloan EP, Natarajan M, Baker B et al. Nocturnal and daytime
panic attacks: comparison of sleep architecture, heart rate
Depression and Anxiety
258
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
Garakani et al
variability, and response to sodium lactate challenge. Biol
Psychiatry 1999;45:1313–1320.
Cohen H, Benjamin J, Geva AB et al. Autonomic dysregulation
in panic disorder and in post-traumatic stress disorder: application of power spectrum analysis of heart rate variability at rest
and in response to recollection of trauma or panic attacks.
Psychiatry Res 2000;96:1–13.
Gorman JM, Sloan RP. Heart rate variability in depressive and
anxiety disorders. Am Heart J 2000;140:77–83.
Yeragani VK, Tancer M, Uhde T. Heart rate and QT interval
variability: abnormal alpha-2 adrenergic function in patients with
panic disorder. Psychiatry Res 2003; 121:185–196.
Slaap BR, Boshuisen ML, van Roon AM, den Boer JA. Heart rate
variability as predictor of nonresponse to mirtazapine in panic
disorder: a preliminary study. Int Clin Psychopharmacol 2002;
17:69–74.
Carney RM, Freedland KE, Stein PK et al. Change in heart rate
and heart rate variability during treatment for depression in
patients with coronary heart disease. Psychosom Med 2000;62:
639–647.
Tucker P, Adamson P, Miranda R et al. Paroxetine increases heart
rate variability in panic disorder. J Clin Psychopharmacol 1997;
17:370–376.
Sullivan GM, Kent JM, Kleber M et al. Effects of hyperventilation on heart rate and QT variability in panic disorder pre-and
post-treatment. Psychiatry Res 2004;125:29–39.
First MB, Spitzer RL, Gibbon M, Williams JBW. Structured
Clinical Interview for DSM-IV-TR Axis I Disorders, Patient
Edition (SCID-IP, 11/2002 Revision). Washington, DC: American Psychiatric Press; 2002.
Guy W. ECDEU Assessment Manual for Psychopharmacology,
Revised (DHEW Publication No. ADM 76-338). Washington,
DC: US Government Printing Office; 1976.
Shear MK, Brown TA, Barlow DH et al. Multicenter collaborative panic disorder severity scale. Am J Psychiatry 1997;154:
1571–1575.
Hamilton M. The assessment of anxiety states by rating. Br J
Psychiatry 1959;32:50–55.
Hamilton M. A rating scale for depression. J Neurol Neurosurg
Psychiatry 1960;23:56–62.
Dillon DJ, Gorman JM, Liebowitz MR et al. Measurement of
lactate-induced panic and anxiety. Psychiatry Res 1987;20:
97–105.
Depression and Anxiety
44. Borg GA. Psychophysical bases of perceived exertion. Med Sci
Sports Exerc 1982;14:377–381.
45. Houtman IL, Bakker FC. The Anxiety Thermometer: a
validation study. J Pers Assess 1989;53:575–582.
46. Chambless DL, Caputo GC, Jasin SE et al. The Mobility
Inventory for Agoraphobia. Behav Res Ther 1985;23:35–44.
47. Kohn PM, Macdonald JE. The survey of recent life experiences:
a decontaminated hassles scale for adults. J Behav Med 1992;15:
221–236.
48. Barlow DH, Craske MG. Mastery of Your Anxiety and Panic. 3rd
ed. New York: Graywind Publications, Inc; 2000.
49. Task Force of the European Society of Cardiology and the North
American Society of Pacing and Electrophysiology. Heart rate
variability: standards of measurement, physiological interpretation and clinical use. Circulation 1996;93:1043–1065.
50. Brosschot JF, Van Dijk E, Thayer JF. Daily worry is related to
low heart rate variability during waking and the subsequent
nocturnal sleep period. Int J Psychophysiol 2007;63:39–47.
51. Martens EJ, Nyklı́cek I, Szabó BM, Kupper N. Depression and
anxiety as predictors of heart rate variability after myocardial
infarction. Psychol Med 2008;38:375–383.
52. Lucini D, Riva S, Pizzinelli P, Pagani M. Stress management at
the worksite: reversal of symptoms profile and cardiovascular
dysregulation. Hypertension 2007;49:291–297.
53. Siepmann M, Grossmann J, Muck-Weymann M, Kirch W.
Effects of sertraline on autonomic and cognitive functions
in healthy volunteers. Psychopharmacology (Berl) 2003;168:
293–298.
54. Yeragani VK, Pohl R, Balon R et al. Effect of imipramine
treatment on heart rate variability measures. Neuropsychobiology 1992;26:27–32.
55. Roose SP, Glassman AH, Attia E et al. Cardiovascular effects of
fluoxetine in depressed patients with heart disease. Am J
Psychiatry 1998;155:660–665.
56. Rasmussen SL, Overo KF, Tanghoj P. Cardiac safety of
citalopram: prospective trials and retrospective analyses. J Clin
Psychopharmacol 1999;19:407–415.
57. Dusseldorp E, Spinhoven P, Bakker A et al. Which panic
disorder patients benefit from which treatment: cognitive
therapy or antidepressants? Psychother Psychosom 2007;76:
154–161.