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.