Psychosomatics 41:311-320, August 2000
© 2000 The Academy of Psychosomatic Medicine
Non-Fearful Panic Disorder
A Variant of Panic in Medical Patients?
Richard P. Fleet, Ph.D.,
Jean-Pierre Martel, Ph.D.(C),
Kim L. Lavoie, M.A.,
Gilles Dupuis, Ph.D., and
Bernard D. Beitman, M.D.
Received July 21, 1999; revised December 1, 1999; accepted December 29, 1999. From the Research Center, Montreal Heart Institute, Department of Psychology, Université du Québec à Montréal, Montreal, Quebec, Canada; Department of Psychology, Concordia University, Montreal, Quebec, Canada; Anxiety Disorders Clinic, Sacré-Coeur Hospital, Montreal, Quebec, Canada; Department of Psychiatry and Neurology, University of Missouri-Columbia, Columbia, MO. Address reprint requests to Dr. Fleet, Research Center, Montreal Heart Institute, 5000 Bélanger Street East, Montreal, Quebec, H1T 1C8, Canada.
|
ABSTRACT
|
Is it possible to have panic attacks without fear? Beitman et al. reported that 32%– 41% of panic disorder (PD) patients seeking treatment for chest pain have non-fearful panic disorder (NFPD). To replicate and extend this work on NFPD, the authors compared NFPD patients (N= 48), PD patients (N=60), and No-PD patients (N=333) at the time of an emergency department visit and follow-up approximately 2 years later. The authors compared comorbid Axis I diagnoses, panic attack symptoms, and scores on self-report measures. A significantly greater proportion of PD patients had comorbid generalized anxiety disorder and agoraphobia than NFPD patients. NFPD patients had self-report scores that were between no-PD and PD patients or similar to no-PD patients, with the exception of the Beck Depression Inventory. At follow-up, NFPD patients, like PD patients, were still symptomatic and had either not improved or had worsened according to scores on all self-report measures. NFPD should be recognized as a variant of PD, both because of its high prevalence in medical settings and its poor prognosis.
Key Words: Panic Disorder
|
INTRODUCTION
|
Is it possible to have panic attacks without fear? It is according to at least three studies conducted in medical settings. These studies suggest that between 32% and 41% of patients with panic disorder (PD) seeking treatment for chest pain have non-fearful panic disorder.1–3 Over a decade ago, Beitman et al.1 observed that several patients with noncardiac chest pain met criteria for PD without reporting either "fear of dying" or "fear of going crazy or losing control." Therefore, Beitman et al. coined the term "non-fearful panic disorder"(NFPD) and defined it as intense episodes of discomfort without either fear of dying or fear of going crazy or losing control but including at least 4 of the remaining 11 symptoms of PD. According to Beitman et al.`s1 definition, patients with NFPD still meet Diagnostic and Statistical Manual-III-Revised (DSM-III-R) or DSM-IV criteria4 for PD. According to DSM-IV criteria, panic attacks are characterized by discrete periods of intense fear or discomfort in which 4 or more of the 13 symptoms of a panic attack develop abruptly and reach a peak within 10 minutes. Hence, patients reporting intense periods of discomfort accompanied by other symptoms of a panic attack, but without fear, still meet DSM-III-R and DSM-IV criteria for PD.
Besides differences in self-reported fear, Beitman et al.1,2 also found that patients with NFPD and PD are essentially indistinguishable in terms of demographic and other clinical characteristics (e.g., course and outcome of illness), suggesting that NFPD may simply be a variant of PD.
Although patients with NFPD meet DSM-IV criteria for PD, this concept remains understudied and somewhat controversial. Only five original studies1–3,5,6 were found through a recent MEDLINE and PSYCHLIT search covering 1980 to 1998 using "non-fearful PD" and "non-fear PD" as search terms. Finding only five studies illustrates the paucity of research on this concept. Besides the lack of research on NFPD, this concept is particularly controversial because it is virtually unheard of in mental health settings where most patients with PD are treated once avoidance behaviors begin. Furthermore, most of our knowledge on PD is derived from research generated almost exclusively in mental health settings. As we have previously reported, there are several differences between medical and psychiatric patients with this anxiety disorder that may have serious diagnostic and therapeutic implications.7 We compared emergency department (ED) and psychiatric setting patients with PD and found that the psychiatric PD patients were younger, consisted of proportionally more women, had significantly higher rates of comorbid agoraphobia, social phobia, specific phobia, and posttraumatic stress disorder, and scored consistently higher on self-report panic measures relative to ED PD patients. We also hypothesized that the differences among panic patients in these settings may be because of a significant proportion of patients in medical settings having NFPD. The high prevalence of NFPD in medical settings coupled with the fact that the empirical basis for this potential variant of PD rests on few studies with small samples (between 6 and 13 patients with NFPD) warrants further study on the subject. Hence, as part of the Montreal Heart Panic Project, we decided to reexamine this controversial yet important concept by comparing NFPD patients to patients with typical PD and patients without PD at presentation to an ED. We also examined the prognostic significance of a NFPD diagnosis by reevaluating patients 2 years after their initial ED visit.
|
METHODS
|
Patients
The sample comprised 441 consecutive patients that presented to the walk-in (ambulatory) emergency department of a major university teaching hospital with a chief complaint of chest pain. These patients (82%) consented to participate in the Montreal Heart Panic Project, one of the largest studies to date examining the prevalence of current Axis I disorders in ED chest pain patients. All subjects were White. There were 172 (39%) women and 269 (61%) men; age (mean±SD=56.8±12.4). Inclusion criteria for this study were complaint of chest pain, ability to comprehend the French language, and the capability to complete the evaluation in the ED. Exclusion criteria were significant cognitive impairment, current psychotic state, and intoxication.
We conducted the study between 9 A.M. and 5 P.M., Mondays through Fridays over a 16-month period. A total of 607 patients were approached. Fifty-one subjects were not included for language reasons. Eighteen patients had already participated. Ninety-seven (18%) of 538 eligible patients did not consent to participate. The following reasons were given for refusal to participate: lack of time (29%), not interested (16%), too tired or uncomfortable (10%), and no precise reason (45%). Nonconsenting patients did not differ significantly from consenting patients with respect to age, gender, cardiac history, or discharge diagnosis (all P>0.05).
Procedure
Patients with a chief complaint of chest pain were identified by the ED triage nurse who then paged in-house research assistants. Nurses were asked to avoid any other screening for inclusion in the study. Patients were asked to participate in a study evaluating the relationship between emotions and chest pain. Trained research assistants (two senior doctoral-level clinical-psychology graduate students and one licensed psychologist) conducted psychiatric interviews and had patients complete self-report questionnaires in the ED. Psychiatric assessments and questionnaires took between 45 and 90 minutes to complete. The on-duty ED cardiologists examined patients and completed the Chest Pain Qualities Scale (described later) for eligible patients. The ED is staffed by 28 certified cardiologists on rotating schedules (average age±SD=47.9±9.9 years; average number of years±SD experience in cardiology practice=18.4±9.6). All cardiologists participated in the study. Both cardiologists and nurses were blind to psychiatric diagnoses. Research assistants conducting psychiatric interviews were blind to chest pain and medical diagnoses. Additional study methods are published elsewhere.8
Study Groups
For the purpose of this report, we compared the following three groups: No-PD, NFPD, and PD at the time of their emergency visit and approximately 2 years later on the measures described below. Patients were diagnosed with PD if they met DSM-III-R criteria for panic disorder using the Anxiety Disorders Interview Schedule-Revised (ADIS-R). Patients were diagnosed with NFPD if they reported having recurrent, intense episodes of discomfort without either fear of dying or fear of going crazy or doing something uncontrolled but including at least 4 of the remaining 11 symptoms of a panic attack. Patients in the No-PD group did not meet diagnostic criteria for either PD or NFPD.
Measures
Structured Diagnostic Interview
The ADIS-R9 was used to diagnose Axis I mood, anxiety, and substance abuse disorders upon patients' presentation to the ED. Interviewers were two advanced doctoral-level clinical psychology students and one licensed psychologist specifically trained in the administration of the ADIS-R. The ADIS-R is based on DSM-III-R criteria, displays good interrater reliability for PD diagnosis,9 and is recommended for panic research.10 All interviews were audiotaped for the purpose of assessing interrater reliability.
Self-Report Measures
At the initial emergency visit and at follow-up, patients completed the following questionnaires: to measure depression— Beck Depression Inventory (BDI);11 general anxiety—State-Trait Anxiety Inventory12 (STAI); and panic-agoraphobia symptomatology—Agoraphobia Cognitions Questionnaire (ACQ),13 Body Sensations Questionnaire (BSQ),13 Mobility Inventory for Agoraphobia (MIA).14 These questionnaires are widely used in research and clinical practice and display good-to-excellent psychometric properties.10
Cardiac Assessment
A medical evaluation was conducted at the time of the ED visit. The medical tests ordered were at the discretion of the attending ED cardiologist and were consistent with standard clinical practice in Canada. Medical charts were reviewed approximately 3 months after the ED visit by a nurse who was blind to the patients' psychiatric diagnosis in order to include in- or outpatient medical testing. Medical charts were searched for history of coronary artery disease (CAD) and hospital discharge diagnoses after the ED visit. Positive history of CAD was indicated when the patient met one or more of the following conditions: a documented history of myocardial infarct (MI), coronary bypass surgery (CABG), percutaneous transluminal coronary angioplasty (PTCA), angiographic testing indicating at least 50% stenosis in one of the major coronary arteries, or thallium or technetium-99m sestamibi studies indicating a perfusion defect.
Chest Pain Assessment
The Chest Pain Quality Scale was used by cardiologists to classify chest pain into typical, atypical, or nonanginal (noncardiac) categories. Typical angina was defined as having pain that is substernal, primarily exertional, and rapidly relieved by rest or nitroglycerin. Atypical angina was defined as having at least two of the above typical chest pain characteristics. Nonanginal (or noncardiac) chest pain was defined as having one or none of the typical chest pain characteristics.
Follow-up Interview
Patients were contacted by phone an average±SD of 23±5.4 months after their initial visit to the ED for a 15- to 20-minute interview. The interview protocol consisted of 40 questions modified from Beitman et al.15 For the purpose of this report we examined questions that pertained to the following six categories: 1) chest pain (presence and frequency per period); 2) ED visits and hospitalizations; 3) patient`s perceived cause for symptoms; 4) disability caused by symptoms; 5) perceived overall health status; and 6) current use of psychotropic medication. The interviewer was a first-year graduate student in clinical psychology who was blind to patients' initial medical and psychiatric diagnoses as well as to the specific hypotheses of the study.
Statistical Analyses
To evaluate the relationship between the interview diagnosis (No PD, NFPD, or PD) and categorical variables (e.g., gender, comorbid Axis I disorders, and CAD), chi-square tests and Fisher's exact test were used. Continuous variables (e.g., mean group differences in age and scores on self-report questionnaire scores) were evaluated using ANOVAs. To examine change in self-report scores from baseline to follow-up, repeated measures ANOVAs were conducted. Level of significance was set at P<0.05.
|
RESULTS
|
Initial Emergency Visit (Baseline)
Structured Interview Diagnoses
Interrater reliability: Average interrater reliability among the three interviewers for PD diagnosis (presence or absence of the disorder confounding NFPD and PD) was high (kappa=0.81)16 based on the rating of 100 randomly selected taped interviews (23% of the total number of interviews).
Study groups (diagnoses): A total of 108 patients (25%) met DSM III-R criteria for PD (with or without agoraphobia). Of these 108 patients, 48 (44%) met Beitman et al.'s1 aforementioned criteria for NFPD. Hence, the three study groups were No PD (n=333), NFPD (n=48), and PD (n= 60).
Patients: There was a significant difference between the three groups in terms of age (F=9.9, df=434; P=0.0001). Post hoc tests using Scheffé's procedure indicated that patients without PD were significantly older (58.2±12.2 years) than patients with PD (51.4±11.7 years) and patients with NFPD (53.6±12.0 years). The two latter groups did not differ in terms of age. However, no gender difference was found among the three groups ( 2=3.4, df=2, P=0.180). Of patients without PD, 203 (61%) were men. The proportion of men in the NFPD and in the PD groups was respectively, 71% (34/48) and 53% (32/60).
Psychiatric comorbidity: We compared the proportion of patients with NFPD and PD with respect to most common comorbid Axis I diagnoses. Results are displayed in Table 1. In sum, a greater proportion of patients with PD had comorbid generalized anxiety disorder (GAD) and agoraphobia (PDA) than patients with NFPD. There was no difference between NFPD and PD in the proportion of patients with comorbid social phobia, posttraumatic stress disorder, obsessive-compulsive disorder, major depressive disorder, and dysthymic disorder.
Scores on self-report questionnaires: Table 2 contains the means and SD of the scores from the self-report questionnaires. The three groups differed significantly on all of the measures. Post-hoc analyses with Scheffé's procedure revealed the following specific differences between groups. On the BDI, patients in both the PD and NFPD groups had significantly higher scores than patients without PD. Yet, PD and NFPD patients did not differ on this measure. On both the State and Trait scales of the STAI, patients without PD had the lowest scores, patients with PD had the highest, and NFPD patients had scores intermediate between the two former groups. The profile of scores on the MIA (because the two subscales of the MIA were correlated at over 0.9, only results of the MIA-accompanied scale are presented) was similar to the STAI, with NFPD patients scoring intermediate between patients without PD and patients with PD. However, a different pattern of scores was observed on the BSQ and on the ACQ. NFPD patients had lower scores than PD patients, but their scores did not differ from patients without PD.
View this table:
[in this window]
[in a new window]
|
TABLE 2. Comparison of means (±SD) on self-report measures for patients without panic disorder (No PD), patients with non-fearful panic disorder (NFPD), and patients with panic disorder (PD)
|
Panic attack symptoms: Table 3 shows the frequency of panic symptoms experienced during the patients' most recent severe episode for patients with NFPD and patients with PD. The two groups differed significantly only on the three following cognitive symptoms of a panic attack: fear of going crazy or losing control, fear of dying, and depersonalization or derealization. NFPD patients reported significantly fewer cognitive symptoms compared with PD patients. Chest pain was not included in the analyses as 100% of patients in both groups reported chest pain (an inclusion criteria).
View this table:
[in this window]
[in a new window]
|
TABLE 3. Comparison of panic symptoms between patients with non-fearful panic disorder (NFPD) and panic disorder (PD)
|
Cardiac and chest pain diagnoses: Twenty-three patients with NFPD (48%) and 26 patients with PD (44%) had a documented history of CAD, a difference that was not significant (Fisher's exact test, P=0.25). A total of 168 No-PD patients (50%) had a documented history of CAD, which when compared with both PD groups was not significantly different (2=1.01, df=1, P= 0.315). However, patients did differ significantly with respect to chest pain diagnoses. At discharge, 75% (n=81) of panic patients (including NFPD patients) were diagnosed with noncardiac chest pain compared to only 52% (n=174) No-PD patients (2=17.45, df=1, P<0.0002). Therefore, although No-PD patients had equally prevalent histories of CAD relative to both groups of panic patients, No-PD patients did differ significantly from both groups of panic patients in terms of their chest pain diagnosis at discharge. Significantly more panic patients (both PD and NFPD) were discharged with a diagnosis of noncardiac chest pain relative to No-PD patients who received greater diagnoses of stable and unstable angina, or MI at discharge. A more detailed description of the cardiac, chest pain, and other medical diagnoses for the entire sample is presented elsewhere.8,17
Follow-up
Patients: A total of 301 patients completed the phone interview (71% of the initial ED sample) and 228 patients completed the self-report questionnaires (50% of the initial sample). The mean age of respondents was 56.8 years and 61% were male.
Assessment for Sampling Bias: Using baseline data obtained at the initial ED visit, we compared patients who participated in the follow-up phase (N=228) with those who did not (N=213) to evaluate whether subject attrition was biased or random. These comparisons revealed that participants and nonparticipants did not differ in terms of gender, age, initial scores on self-report questionnaires, or psychiatric and cardiac diagnoses. We therefore concluded that subject attrition was random.
Telephone Interview: Results of the follow-up interview are presented in Table 4. No chi-square statistics were reported when the expected cell size was below 5.
View this table:
[in this window]
[in a new window]
|
TABLE 4. Results of the follow-up telephone interview for patients with no panic disorder (No PD), non-fearful panic disorder (NFPD), and panic disorder (PD)
|
Chest Pain: A greater proportion of panic patients reported chest pain episodes in the previous month than patients without PD. A two-by-two comparison between NFPD and PD patients did not reveal any statistical difference. The means±SD number of chest pain episodes per month for each group was as follows: No PD=3.9±8.6; NFPD=5.0±8.8; and PD=6.6±10.5.
ED Visits and Hospitalizations: A greater proportion of PD patients reported visiting the ED and being hospitalized during the year previous to the follow-up interview than patients without PD. Two-by-two comparison between NFPD and PD patients did not reveal any difference between these two groups.
Work Disability: Although the small number of expected cases per cell did not allow for overall comparisons (chi-square analyses) among the three groups on the question: "Do your current symptoms render you unable to work?" two-by-two comparisons between NFPD and PD with Fisher exact test (two-sided) did not reveal any difference (P=0.43). A comparison between PD and No-PD groups was also nonsignificant (P=0.16) as was the comparison between NFPD and No PD (P=1.0).
Perceived Overall Health Status: A significantly greater proportion of panic patients perceived their general health to be poor compared to No-PD patients. There was no difference between NFPD and PD patients in terms of the perception of the quality of their general health.
Perceived Cause for Symptoms: It was not possible to conduct chi-square analyses in order to compare patients' perceived cause for their symptoms. However, visual inspection of the data suggests similar responses across groups. Interestingly, over half of patients in each group attributed an emotional or psychological cause to their symptoms.
Psychotropic Medication: It was not possible to conduct chi-square analyses comparing the current use of antidepressant medication between groups. However, analyses revealed that a significantly greater proportion of panic patients were treated with benzodiazepines than were No-PD patients. No difference was observed between NFPD and PD patients in terms of proportion of benzodiazepine use.
Scores on Self-Report Questionnaires: Repeated measures ANOVAs were conducted to examine change in self-report scores of depression, anxiety and panic-agoraphobia symptomatology from baseline to follow-up.
Depression: No time effect was observed on the BDI. Hence, at follow-up, depression scores did not significantly change from baseline for all three groups. In contrast to baseline observations, scores on the BDI for NFPD and PD groups differed at follow-up. This could possibly be attributed to a slight, yet nonsignificant, increase in BDI scores for the PD group coupled with a slight also nonsignificant decrease in scores for the NFPD at follow-up. Moreover, differences in baseline BDI scores between NFPD and No-PD patients were no longer present at follow-up. However, this finding may be explained by a decrease in statistical power associated with patient loss at follow-up.
State Anxiety: No time effect was observed on the State scale of the STAI with patient scores for all groups essentially unchanged from baseline at follow-up assessment. At follow-up, the pattern of differences among the three groups was the same as baseline. PD patients still presented the highest scores and No-PD patients the lowest. NFPD patients had scores intermediate between the two groups.
Trait Anxiety: A very different pattern of scores was observed on the Trait scale of the STAI. First, a significant time effect was observed with patients in all groups displaying an increase in trait anxiety scores. However, in contrast to what was observed at baseline, scores for the NFPD group were no longer different from those of the PD and No-PD groups at follow-up.
Panic-Agoraphobia Symptomatology: On the panic-agoraphobia measures no consistent pattern of scores was observed. First, a significant time effect was observed on the MIA with all groups displaying an increase on this measure of agoraphobic avoidance. At follow-up, scores for the NFPD patients did not differ from those of the other groups. Second, a significant time effect was also observed on the ACQ with all groups displaying an increase in scores on this measure of fear. As was the case with the MIA, patients with NFPD no longer differed from patients in the other two groups. Finally, no time effect was observed on the BSQ. Like those observed on the other panic-agoraphobia measures, patients with NFPD no longer differed from patients with PD and those without PD.
|
DISCUSSION
|
In this study we reexamined the controversial concept of NFPD first coined by Beitman et al. more than a decade ago. Similar to previous studies, we compared patients with NFPD and typical PD in terms of scores on self-report measures and panic symptoms. Extending previous research, we compared these groups with respect to current comorbid Axis I disorders. Furthermore, NFPD and PD patients were compared with No-PD patients. Finally, we examined the prognostic significance of NFPD by reinterviewing patients 2 years after their initial ED visit and examined changes in scores on self-report measures.
Beitman et al.'s1,2 reports suggested that besides fear, NFPD and PD patients were identical. In contrast, our results showed some noteworthy differences between these two groups. First, at their emergency visit, patients with NFPD and PD differed in terms of comorbid Axis-I disorders as well as on several self-report questionnaires. Specifically, NFPD patients had lower rates of comorbid GAD and Agoraphobia, lower scores on both the State and Trait scales of the STAI, and lower scores on all of the panic-agoraphobia measures (BSQ, ACQ, and MIA). The differences between these two groups of panic patients, however, need to be seen in the context of the following similarities: 1) patients did not differ in terms of age and gender; 2) patients reported almost identical panic symptoms, with the exception of the psychological symptoms of panic (depersonalization, fear of dying or losing control); and 3) patients also did not differ in terms of other comorbid Axis I diagnoses with the exception of GAD and Agoraphobia.
We propose the following explanations for the differences between our baseline findings and those of previous studies. First, the fact that previous studies did not find any significant difference between NFPD and PD groups could possibly be explained by a problem of insufficient statistical power. As previously mentioned, sample size of NFPD groups in previous studies ranged from 6 to 13 patients. Hence, one cannot exclude the possibility of a Type II error (i.e., concluding that no statistical difference between groups exists when in fact a difference does exist). Furthermore, the use of different self-report measures of anxiety than Beitman et al.1,2 may also account for the discrepant findings. Beitman et al. used the Zung Self-Rated Anxiety Scale,18 the Marks Matthews Fear Questionnaire,19 and the Anxiety and Phobic-Anxiety scales of the Brief Symptom Inventory20 as measures of panic-anxiety. Although these questionnaires are widely used measures of anxiety, they have not been developed to specifically study PD. Finally, the discrepant findings may also be attributed to differences in patient populations. Previous studies of NFPD were conducted on cardiology patients (patients referred for angiograms, cardiology outpatients, or patients hospitalized in the coronary care unit). It is unclear how ED and cardiology patients differ. Yet, it is possible that only more symptomatic patients are hospitalized in the coronary care unit or are referred to cardiologists for invasive cardiac testing. Hence the panic patients, including the NFPD patients in Beitman et al.'s studies, may have been more ill relative to patients in our ED study, who likely had a more recent onset of their panic symptoms (accounting for their presentation to an ED rather than to their cardiologist).
The fact that at baseline, patients with NFPD had lower scores than PD patients on most self-report measures could suggest that the former diagnosis may not be clinically significant. In fact, scores on several measures of panic-agoraphobia symptomatology for NFPD patients were not different from those No-PD patients. However, the No-PD group was comprised of chest pain patients primarily afflicted by CAD. It is therefore possible that these patients scored higher on the self-report scales because of their cardiac symptoms and associated distress. Furthermore, 2 years after their ED visit, patients with NFPD were still symptomatic and their condition had either not improved or had worsened as indicated by scores on all self-report measures—a finding similar to patients with PD. Hence, although apparently less distressed on their initial ED visit than PD patients, NFPD and PD have similar negative prognoses, suggesting that NFPD is a clinically significant diagnosis.
There is no question that a significant proportion of patients with PD in medical settings meet Beitman et al.'s criteria of NFPD. However, a salient question remains partially unanswered. Is NFPD simply a variant of PD or is it a distinct diagnostic entity altogether? Feighner et al.21 maintain that once a psychiatric phenomenon is described clinically, establishing its validity requires laboratory, longitudinal, treatment, and family studies. For the NFPD diagnoses, an important goal is to show these patients are similar to patients with classic PD. The results of our baseline comparison between patients with NFPD and PD only partially support this notion. In addition, our data reveal an atypical age distribution for No-PD and PD (including NFPD) patients, with No-PD patients being significantly older than both NFPD and PD patients. It is generally believed that PD begins in the early twenties, however, our findings are consistent with previous studies showing that PD patients seeking treatment for unexplained chest pain (such as those in the present study) are older, with mean age ranging from 45 to 54 years.1,3,22,23 The reason why the PD (including NFPD) patients are older is unclear. Beitman and Kushner24 obtained lifetime diagnoses of PD patients complaining of chest pain and found that the panic syndrome had a late onset for several patients. It was proposed that the older mean age of PD patients may also reflect the normal distribution of the disorder, with older patients more likely to consider heart disease as a cause. Still, the concept of late-onset PD discovered in medical settings requires further investigation. We have previously hypothesized that physicians may be more reluctant to consider a diagnosis of PD in older patients with relatively new symptoms if it is generally believed that PD typically begins in early adulthood.
The fact that NFPD had a chronic course similar to PD can serve as a longitudinal validation of the concept of NFPD as a variant of PD. Beitman et al.5 provided further support for the concept of NFPD through a family history study of PD in first-degree relatives of chest pain patients with either NFPD, PD, or No PD. As hypothesized, the same elevated rates of PD in first-degree relatives were found in NFPD and PD groups (17%) compared to the No-PD group (4.6%). Finally, a preliminary biological validation for NFPD comes from the study of three neurology patients with NFPD who underwent a sodium lactate challenge in a double-blind test.25 Results showed a similar panic response to the lactate challenge to that observed in PD patients. Hence, there is mounting validating evidence, albeit preliminary, that provides support for the notion that NFPD is a variant of PD rather than a different diagnostic entity altogether. Although we tried to locate treatment data to support the NFPD construct, we were unable to find any published reports. As mentioned previously, the NFPD construct is still relatively new, which may reflect the paucity of original research on the subject.
Nevertheless, alternative explanations or diagnoses need to be considered. First, medical testing was not standard in this study and in previous studies on NFPD. Although this probably reflected clinical reality, it is impossible to exclude the possibility that certain patients had an undiagnosed medical condition (e.g., gastroesophageal reflux disease) that would account for the panic-like syndrome. We were unable to discount all other possible medical explanations for chest pain. However, it is doubtful that one or several undiagnosed medical conditions would have accounted for the high prevalence of the panic-like syndrome reported in our and all previous studies. Second, perhaps what is referred to as NFPD is actually a somatoform disorder. It is also impossible to exclude this possibility as previous studies have not evaluated the presence of somatoform disorders through structured psychiatric interviews. Yet, one can question whether any of the somatoform disorders could better account for the similarities between PD and NFPD reported to date. Furthermore, according to the DSM-IV4 it is now possible to have both an anxiety and a somatoform diagnosis, so it is not a simple question of mutually exclusive diagnoses. Finally, we need to question whether NFPD patients really have panic attacks without fear. Although admitting to discomfort associated with their symptoms to meet criteria for NFPD, is discomfort in itself enough to rush to an ED? Perhaps different patients simply have different words for expressing similar emotions (e.g., like fear) associated with their panic attacks. Several patients may not have the words to express their emotions (e.g., alexythymia) and thus have alexythymic panic as described by Jones26 in a case report. Other patients may have noncognitive panic attacks. Rachman et al.27 suggest that these patients have fearful cognitions yet they are unable to detect, recall, or describe them. It is also possible that patients are aware of but do not want to admit their psychological symptoms, including fear, simply because of the negative stigma associated with mental disorders. The higher (although not significantly different) proportion of men in the NFPD group may be in part related to men's reluctance to admit fearful emotions relative to women. This possibility is consistent with gender stereotypes and socialization practices when it comes to the expression of emotions, particularly those representing anxiety or fear.28,29 Therefore, as we consider alternative diagnoses to NFPD in future research, we need to further investigate the question of alexythymic panic, noncognitive panic, and simply the denial of emotional symptoms as explanations of NFPD.
Study Limitations
The primary limitation of the present study concerns the validity of our PD and NFPD diagnoses in patients with a history of CAD. According to DSM-III-R criteria, in order to receive an anxiety or mood disorder diagnosis, organic factors must be ruled out as causes or maintenance factors of the disturbance and psychiatric symptoms must not be a direct cause of a medical condition. The naturalistic characteristics of this study limit us to the cardiac testing prescribed by the attending ED cardiologists. In the absence of more elaborate medical testing (e.g., angiography and nuclear medicine perfusion investigations) we cannot absolutely exclude that some patients diagnosed with PD or NFPD may have had an underlying CAD that could account for some of their symptoms. However, cardiologists deal with probability estimates, not absolutes. Although 44% of our PD patients (including NFPD) were determined to have a prior history of CAD, 75% of our PD and NFPD patients were discharged from the hospital with noncardiac diagnoses based on probability estimates that take into account medical examination and test results. However, the remaining 25% of PD and NFPD patients(27/108) were found to have either stable angina, unstable angina, or MI. To respect DSM-III-R exclusion criteria, by removing these patients from the sample, the PD-NFPD prevalence still remains remarkably high at 19.5% of the total sample. To illustrate, patients (in the total sample of 441 subjects) discharged with a diagnosis of "noncardiac chest pain" (n=255) had significantly higher scores on all psychological scales (table not presented) relative to patients discharged with a "cardiac chest pain" diagnosis (n=186, P<0.05). These results support the argument that it is not the cardiac diagnosis that explains the psychopathology in chest pain patients taken as a group. Thus, it is possible that the "cardiac chest pain" patients also found to have PD or NFPD truly had one of these disorders regardless of their medical condition.
An additional limitation of our study pertains to how our results generalize to an ED of a general hospital (as opposed to primarily cardiology), and to nonambulatory patients (i.e., those who arrive by ambulance rather than by themselves). This study was conducted at the Montreal Heart Institute, which specializes in cardiac care. Although anyone with cardiovascular complaints may consult the ED, many who do consult have history of CAD. Our decision to examine only ambulatory (walk-in) patients was based on the fact that most patients arriving at the ED by ambulance are in too precarious a medical condition to participate in a study. Thus, the PD and NFPD results reported here are not based on all ED chest pain patients and should be interpreted accordingly.
The relatively low participation rates of the follow-up phase of the study (71% of the original sample completed the phone interview and 50% completed the self-report questionnaires) are also a limitation of this study. In an attempt to control for sampling bias, we examined baseline data obtained during patients' initial ED visit for both follow-up participants and non-participants. Our comparisons did not reveal any differences with respect to age, gender, initial self-report scores, or both psychiatric and cardiac diagnoses. We therefore concluded that subject attrition was random and not because of any meaningful differences between participants and nonparticipants.
|
CONCLUSION
|
Despite the preliminary nature of the evidence, we suggest that NFPD be considered a variant of PD, at least until such time as future research disconfirms this notion. PD is one of the most common yet highly unrecognized psychiatric disorders in medical settings. Because one-third to almost one-half of these patients meet criteria for NFPD, it is important to inform physicians of the existence of this variant of PD which may account for the high rates of misdiagnoses and extensive costly medical testing on these patients. Although treatment research for NFPD is scarce, classic PD is highly treatable. If anything, NFPD is a less severe form of panic disorder that may also be treatable. The decision to consider the diagnosis and treatment of NFPD should be influenced by the finding that these patients do not appear to improve with time—their condition may even worsen.
The authors wish to thank Dr. Julie Turcotte for her advice on the manuscript.
|
REFERENCES
|
-
Beitman BD, Basha I, Flaker G, et al: Non-fearful panic disorder: panic attacks without fear. Behav Res Ther 1987; 25:487–492[CrossRef][Medline]
-
Beitman BD, Kushner M, Lamberti JW, et al: Panic disorder without fear in patients with angiographically normal coronary arteries. J Nerv Ment Dis 1990; 178:307–312[CrossRef][Medline]
-
Carter C, Maddock R, Zoglio M, et al: Panic disorder and chest pain: a study of cardiac stress scintigraphy patients. Am J Cardiol 1994; 74:296–298[CrossRef][Medline]
-
American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Washington, DC, American Psychiatric Association, 1994
-
Beitman BD, Thomas AM, Kushner MG: Panic disorder in the families of patients with normal coronary arteries and non-fear panic disorder. Behav Res Ther 1992; 30:403–406[CrossRef][Medline]
-
Wilson KG, Sandler LS, Asmundson GJ: Fearful and non-fearful panic attacks in a student population. Behav Res Ther 1993; 31:407–411[CrossRef][Medline]
-
Fleet RP, Marchand A, Dupuis G, et al: Comparing emergency department and psychiatric setting patients with panic disorder. Psychosomatics 1998; 39:512–518[Abstract/Free Full Text]
-
Fleet RP, Dupuis G, Marchand A, et al: Panic disorder in emergency department chest pain patients: prevalence, comorbidity, suicidal ideation, and physician recognition. Am J Med 1996; 101:371–380[CrossRef][Medline]
-
DiNardo PA, Barlow DH: Anxiety Disorders Interview Schedule-Revised (ADIS-R). Albany, NY, Phobia and Anxiety Disorders Clinic, 1988
-
Shear MK, Maser JD: Standardized assessment for panic disorder research: a conference report. Arch Gen Psychiatry 1994; 51:346–354[Abstract/Free Full Text]
-
Beck AT: Depression: clinical experimental and theoretical aspects. New York, Harper & Row, 1967
-
Speilberger CD: Manual for the State-Trait Anxiety Inventory (STAI form Y). Paolo Alto, CA, Consulting Psychologists Press, 1993
-
Chambless DL, Caputo GC, Bright P, et al: Assessment of fear in agoraphobics: the Body Sensations Questionnaire and the Agoraphobic Cognitions Questionnaire. J Consult Clin Psychol 1984; 52:1090–1097
-
Chambless DL, Caputo GC, Jasin SE, et al: The mobility inventory for agoraphobia. Behav Res Ther 1985; 23:35–44[CrossRef][Medline]
-
Beitman BD, Kushner MG, Basha I, et al: Follow-up status of patients with angiographically normal coronary arteries and panic disorder. JAMA 1991; 265:1545–1549
-
Landis JR, Koch GG: The measurement of observer agreement for categorical data. Biometrics 1977; 33:159–174[CrossRef][Medline]
-
Fleet RP, Dupuis G, Kaczorowski J, et al: Suicidal ideation in emergency department chest pain patients: panic disorder a risk factor. Am J Emerg Med 1997; 15:345–349[CrossRef][Medline]
-
Zung WK: A rating instrument for anxiety disorders. Psychosomatics 1971; 12:271–379
-
Marks IM, Mathews AM: Brief standard self-rating for phobic patients. Behav Res Ther 1979; 17:263–267[CrossRef][Medline]
-
Derogatis LR, Melisaratos N: The brief symptom inventory: an introductory report. Psychol Med 1983; 13:595–605[Medline]
-
Feighner JP, Robins E, Guze SB, et al: Diagnostic criteria for use in psychiatric research. Arch Gen Psychiatry 1972; 26:57–63[Abstract/Free Full Text]
-
Beitman BD, Murkerji V, Lamberti JW: Panic disorder in patients with chest pain and angiographically normal coronary arteries. Am J Cardiol 1989; 63:1399–1403
-
Carter C, Maddock R, Zoglio M, et al: Panic disorder and chest pain: A study of cardiac stress scintigraphy patients. Am J Cardiol 1994; 74:296–298
-
Beitman BD, Kushner M: Late-onset panic disorder: Evidence from a study of patients with chest pain and normal cardiac evaluations. Int J Psychatry Med 1991; 21:29–35
-
Russell J, Kushner M, Beitman B, et al: Non-fear panic disorder in neurological patients validated by lactate infusion. Am J Psychiatry 1991; 148:361–364[Abstract/Free Full Text]
-
Jones BA: Panic attacks with panic masked by alexithymia. Psychosomatics 1984; 25:858–859[Free Full Text]
-
Rachman SJ, Levitt K, Lopatka C: Panic: the links between cognitions and bodily symptoms—I. Behav Res Ther 1987; 25:411–423[CrossRef][Medline]
-
Brody LR: Gender differences in emotional development: a review of theories and research. J Pers 1985; 53:102–149[CrossRef]
-
Mirowsky J, Ross CE: Sex differences in distress: real or artifact? Am Sociol Rev 1995; 60:449–468
This article has been cited by other articles:
|
|
|
|
 
C. B. Bringager, K. Gauer, H. Arnesen, S. Friis, and T. Dammen
Nonfearful Panic Disorder in Chest-Pain Patients: Status After Nine-Year Follow-Up
Psychosomatics,
September 1, 2008;
49(5):
426 - 437.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C. B. Bringager, T. Dammen, and S. Friis
Nonfearful Panic Disorder in Chest Pain Patients
Psychosomatics,
February 1, 2004;
45(1):
69 - 79.
[Abstract]
[Full Text]
[PDF]
|
|
|
Get information about faster international access.
Privacy Policy
Copyright © 2000
Academy of Psychosomatic Medicine.
All rights reserved.
Home
| Search
| Current Issue
| Past Issues
| Subscribe
| All APPI Journals
| Help
| Contact Us
|
TOP
ABSTRACT
INTRODUCTION
