INT'L. J. PSYCHIATRY IN MEDICINE, Vol. 22(3) 197-203,1992
RESPONSE TO 35 % C02 IN PATIENTS WITH CHEST PAIN AND ANGIOGRAPHICALLY NORMAL CORONARY ARTERIES
BERNARD D. BEITMAN, M.D. University of California, lrvine MARY BETH LOGUE, M.A. University of Missouri, Columbia ANN MUlR THOMAS, B.A. University of California, lrvine KIM BARTELS, M.A. University of Missouri, Columbia
ABSTRACT
Objectives: Several interview studies have suggested that panic disorder
(PD) exists in patients with angiographically normal coronary arteries (NCA). Interview studies require corroboration by other studies in order to validate them. The purpose of this study is to test whether response to the inhalation of 35% C02 reliably discriminates between PD and non-panic disorder patients in this population. Method: Three groups were studied: six with NCA and PD, five with NCA and no PD, and ten in the control group. All subjects breathed room air, then 35% CO;? in a single-blind fashion. Each completed the Acute Panic Inventory (API) before and during the procedure. Results: The NCA-panic group scored significantly higher than the other two groups on the Acute Panic Inventory from baseline to post- inhalation. Conclusion: Despite several methodological limitations including a relatively small number people in each cells, 35% C02 was shown to trigger more intense responses in panic patients, thus helping to validate the interview findings. (/nt%J. Psychiatry in Medicine 22: 197-203, 1992)
Key Words: panic disorder, C02 challenge test, coronary artery disease, chest pain, normal coronary arteries, diagnostic validation 197
8 1992, Baywood Publishing Co., Inc.
doi: 10.2190/BC4G-WUA0-KN5X-GL63 http://baywood.com
198 / BEITMAN ETAL.
Two studies have established that more than one-third of patients with chest pain and angiographically normal coronary arteries (NCA) fit DSM-I11 criteria for panic disorder [l,21, and a further study showed that more than one-third of such patients fit ICD-9 criteria for anxiety neurosis [3]. Interview studies require additional data to validate their findings. As suggested by Feigner and colleagues, psychiatrists may validate interview findings in several ways, including structured interviews, family studies, follow-up studies, placebo controlled medication trials, and challenge tests [4]. Such studies are crucial in helping to convince skeptical audiences in other medical disciplines of the diagnostic validity of panic disorder in medical populations. The present study reports the results from a pilot study of the effects of 35% carbon dioxide (CO2) inhalation in NCA patients. Since 35% C02 has been shown to trigger panic attacks in psychiatric patients with panic disorder [5, 61, we hypothesized that responses to this challenge would reliably discriminate between those NCA patients who meet criteria for panic disorder and those who do not. METHOD Subjects
The subjects included eleven persons who had both chest pain and a recent normal or near-normal (30%or less stenosis of any major vessel) cardiac catheterization, six of whom had a further interview-based diagnosis of panic disorder (NCA-Panic group)' and five of whom did not (NCA-No Panic group); and ten persons who had no chest pain, history of cardiac catheterization, or panic disorder (Comparison group). Subjects in the NCA groups were recruited via their cardiologists within six months of their catheterizations, and the comparison subjects were recruited via newspaper advertisements. There were no significant differences among the groups in age and sex distribution. Procedure
Before the C02 procedure, all subjects were interviewed using the SCID-UP-R to assess for history of panic disorder, major depression, and substance abuse [7]. They also completed the Zung Self-rating Anxiety Scale [S], Chronic Panic Inventory [6], and Brief Symptom Inventory [9]. These inventories were given in order to better assess baseline anxiety, panic, and other psychiatric symptomatology, respectively. Five of the six NCA-Panic subjects met criteria for panic disorder at the time of the SCID interview. The remaining subject in this group was experiencing panic attacks less frequently, and had met all criteria for panic disorder in the recent past.
CO2 RESPONSE IN CHEST PAIN/PANIC PATIENTS / 199
Subjects were then given a consent form describing the COZprocedure, including information about potential physical symptoms and anxiety reactions. Three sets of inhalations were then performed: 1)practice with mask disconnected from gases; 2) room air; and 3) 35% COZ.For each inhalation, the following procedure was employed: 1) subject completed the Acute Panic Inventory (API), a twentyeight-item checklist of physicaVpsychologica1symptoms [6]; 2) subject took two breaths with the mask on, exhaling for five seconds, inhaling for five seconds, then holding for five seconds; 3) subject exhaled, breathing normally; 4) heart and respiration rates were measured; and 5) subject again completed the API. Ten minutes were allotted between sets, and subjects were blind to the ordering of gases inhaled, although the order did not vary across subjects. The API was completed again ten minutes after the final (COZ)inhalation. Data Analysis Two change scores were used as dependent variables: API after COzminus API after room air (COz - air API), and API after COZminus baseline API (COz baseline API). The second measure is commonly reported in other studies of this type [6], but the first takes into account changes in anxiety produced by the inhalation procedure itself, regardless of gas content (e.g., placebo response), and can be considered a more rigorous test of COZ effects. Because of the high variability in these two change scores and the baseline API, analyses of variance were run on their rank-transformation [101 using group (NCA-Panic, NCA-No Panic, or Comparison) as the independent variable. Group differences in the other questionnaires were tested using analysis of variance on the untransformed data. Tukey’s pairwise comparisonsbetween groups were done with alpha set at .05.
RESULTS All statistical test results are displayed in Table 1. As can be seen, the groups did not differ significantly in history of major depression or substance abuse. However, the NCA-Panic group scored significantly higher on the Chronic Panic Inventory [F(2,20) = 4 . 7 2 , ~= .02], and approached significantly higher scores on the Zung Self-rating Anxiety Scale [F(2,20) = 3.57, p = .05] and baseline API [F(2,20) = 2.87, p .08]. The two NCA groups approached a significant difference from the Comparison group on the BSI total score [F(2,20) = 3 . 2 0 , ~= .05]. The NCA-Panic group scored marginally higher than the other two groups on the COz-airAPI change score [F(2,20) = 2 . 8 0 , = ~ .09], and significantlyhigher on the COz-baseline API change score [F(2,20) = 4.07, p = .03]. These results indicate that the NCA-Panic group was more likely to respond to COz challenge with panic symptoms than the other two groups, although statistical significance was reached only on the more methodologically suspect change score.
200 / BEITMAN ET AL.
Table 1. Initial Anxiety Ratings and Panic Symptom Level in Response to CO2 Inhalation, and Group Characteristics
Variable
NCA-No NCA-Panic Panic Group Group (N = 6) (N = 5)
Baseline API'
7.50
Comparison Group ( N = 10)
3.40
F
p
1.70
2.87
.08*
Con-air API'
(6.06)
(4.45)
(3.02)
14.17 (10.34)
2.80 (0.84)
4.30 (7.30)
20.178
Con-baseline API' .03*
2.60ab
2.80
.Og*
3.00b
4.07
(19.36)
(1.95)
(6.32)
BSI (total score)
49.67 (33.00)
49.75 (44.81)
19.40 (8.98)
3.20
.OF
Chronic Panic Inventory
25.1p (21.77)
15.6Pb (I 3.28)
2.25' (2.71)
4.72
.02**
Zung Anxiety Scale
43.1? (13.14)
33.OPb (8.60)
30.226 (6.22)
3.57
.05*
46.00 (6.81)
42.40 (6.07)
44.90 (4.20)
0.62
.55
2M/4F
3M/2F
5m15f
0.82
.66
History of major depression
50%
60%
30%
1.40
(3)
(3)
(3)
History of substance abuse (%lW2
33%
50%
10%
(2)
(2)
(11
Sex2
.49 2.76 .25
Note: Except where noted, values are means; standard deviations are listed below each mean in parentheses. 'Indicates dependent variable was rank-transformed before analysis. 21ndicatesChi-square rather than f test. 4Bb,bDifferentsuperscripts indicate a significant difference between groups. *p < .10 "p < .05
CO2 RESPONSE IN CHEST PAIN/PANIC PATIENTS / 201
Two Comparison subjects had a marked symptom response to 35% COz; this is to be expected given that about 15 percent of persons without panic disorder respond in this way [5]. Two NCA-Panic subjects did not respond to COZwith increased panic symptoms; this finding also agrees with previous indications that about 35 percent of psychiatric panic disorder patients do not experience marked increase in anxiety under 35 percent COZchallenge [5]. Note that one of these non-responders had a recent history of panic disorder, but did not meet the diagnostic criteria at the time of the study. DISCUSSION
This pilot study provides preliminary support for the idea that patients with angiographicallynormal coronary arteries and panic disorder respond to 35% C 0 2 with a statistically significant increase in panic symptoms, as compared to patients with angiographicallynormal coronary arteries but without panic disorder, as well as to a comparison sample with neither chest pain nor panic disorder. This is the first study to use 35% COZchallenges to support interviews showing that panic disorder exists and is prevalent in patients with angiographically normal coronary arteries [I, 21. The findings of this study are somewhat compromised by several methodological limitations. The most obvious of these is the small number of subjects in the study, which limits its statistical power. Furthermore, because of equipment limitations, we were not able to randomize the order of the COZand room air, so the greater panic response seen after COZ inhalation may have been due to increasing anxiety as the experiment progressed, rather than any effect of the C02 itself. Finally, the person who administered the gases (KMB) was aware of their contents, and she may have given subjects subtle cues for more panic symptoms when she was administering COz gas, compared to when she was administering room air. A further objection to our suggestion that COZ is a valid marker for panic disorder in persons with angiographically normal coronary arteries and chest pain is that subjects in the NCA-Panic group had higher anxiety levels at baseline than either of the other subject groups. Their COZresponses might thus be a function of general anxiety level rather than panic disorder. However, a recent study comparing COz response in panic disorder patients versus those with other anxiety disorders found that only panic disorder patients responded significantly to COZ [ l l ] ; thus, COZresponse appears to be a specific marker of panic disorder. Our findings, although preliminary, fit an emerging pattern of evidence that supports the validity of panic disorder as a diagnosis for some patients with angiographically normal coronary arteries. This includes evidence that patients with angiographicallynormal coronary arteries and panic disorder have a stronger family history of panic disorder and major depression [121, and are less occupationally, physically, and socially functional over time [131 than patients with
202 / BEITMAN ET AL.
angiographically normal coronary arteries who do not have panic disorder. Further, we have preliminary evidence that these patients respond favorably to standard anti-panic medications [14]. Strong efforts must be made to inform cardiologists and primary care physicians of this possible explanation for cardiological symptoms in persons without objective evidence of heart disease. Significant patient relief may result, because panic disorder is an often unrecognized but highly treatable disorder. REFERENCES 1. B. D. Beitman, V. Mukerji, J. W. Lamberti, L. Schmid, L. DeRosear, M. Kushner, G. Flaker, and I. Basha, Panic Disorder in Patients with Chest Pain and Angiographically Normal Coronary Arteries, American Journal of Cardiology, 63:1399-1403, 1989. 2. W. Katon, M. L. Hall, J. Russo, L. Cormier, M. Hollifield, P. P. Vitaliano, and B. D. Beitman, Chest Pain: The Relationship of Psychiatric Illness to Coronary Arteriography Results, American Journal of Medicine, 84: 1-9,1988. 3. C. Bass and C. Wade, Chest Pain with Normal Coronary Arteries: A Comparative Study of Psychiatric and Social Morbidity, Psychological Medicine, 1451-61, 1984. 4. J. P. Feighner, E. Robins, S. B. Guze, R. A. Woodruff, Jr., G. Winokur, and R. Munoz, Diagnostic Criteria for Use in Psychiatric Research, Archives of General Psychiatry, 265743,1972. 5 . E. J. L. Griez, H. Lousberg, M. A. van den Hout, and G. M. van der Molen, C02 Vulnerability in Panic Disorder, Psychiatry Research, 20237-95,1986. 6. M. R. Fyer, J. Uy, J. Martinez, R. Goetz, D. F. Klein, A. Fyer, M. R. Liebowitz, and J. Gorman, C02 Challenge of Patients with Panic Disorder, American Journal of Psychiatry, 144:1080-1082,1987. 7. R. L. Spitzer and J. B. W. Williams, Structured Clinical Interview for DSM-III-RUpjohn Version, Biometrics Research Department, New York State Psychiatric Institute, New York, 1988. 8. W. K. Zung, A Rating Instrument for Anxiety Disorders, Psychosomatics, 22:271-279, 1971. 9. L. R. Derogatis and N. Melisaratos, The Brief Symptom Inventory: An Introductory Report, Psychological Medicine, 13595-605,1983. 10. w. J. Conover and R. L. Iman, Rank Transformations as a Bridge between Parametric and Nonparametric Statistics, American Statistician, 35124-129,1981. 11. E. Griez, J. Zandbergen, H. Pols, and C. de Loof, Response to 35%C02 as a Marker of Panic in Severe Anxiety, American Journal of Psychiatry, 142796797,1990. 12. B. D. Beitman, V. Mukeji, M. Kushner, A. M. Thomas, J. Russell, and M. B. Logue, Validating Studies for Panic Disorder in Patients with Normal Coronary Arteries, Medical Clinics ofNorth America, in press. 13. B. D. Beitman, M. G. Kushner, I. Basha, J. Lamberti, V. Mukerji, and K. Bartels, Follow-up Status of Patients with Angiographically Normal Coronary Arteries and Panic Disorder, Journal of fhe American Medical Association, 265:1545-1549, 1991.
CO2 RESPONSE IN CHEST PAIN/PANIC PATIENTS / 203
14. B. D. Beitman, I. M. Basha, L. H. Trombka, M. A. Jayaratna, B. Russell, G. Flaker, and S. Anderson, Pharmacotherapeutic Treatment of Panic Disorder in Patients Presenting with Chest Pain, Journal of Family Practice, 28177-180,1989.
Direct reprint requests to: Bernard Beitman, M.D. Psychiatry Clinic University of Missouri-Columbia Hospital and Clinics One Hospital Drive Columbia, MO 65212
RESPONSE TO 35 % C02 IN PATIENTS WITH CHEST PAIN AND ANGIOGRAPHICALLY NORMAL CORONARY ARTERIES
BERNARD D. BEITMAN, M.D. University of California, lrvine MARY BETH LOGUE, M.A. University of Missouri, Columbia ANN MUlR THOMAS, B.A. University of California, lrvine KIM BARTELS, M.A. University of Missouri, Columbia
ABSTRACT
Objectives: Several interview studies have suggested that panic disorder
(PD) exists in patients with angiographically normal coronary arteries (NCA). Interview studies require corroboration by other studies in order to validate them. The purpose of this study is to test whether response to the inhalation of 35% C02 reliably discriminates between PD and non-panic disorder patients in this population. Method: Three groups were studied: six with NCA and PD, five with NCA and no PD, and ten in the control group. All subjects breathed room air, then 35% CO;? in a single-blind fashion. Each completed the Acute Panic Inventory (API) before and during the procedure. Results: The NCA-panic group scored significantly higher than the other two groups on the Acute Panic Inventory from baseline to post- inhalation. Conclusion: Despite several methodological limitations including a relatively small number people in each cells, 35% C02 was shown to trigger more intense responses in panic patients, thus helping to validate the interview findings. (/nt%J. Psychiatry in Medicine 22: 197-203, 1992)
Key Words: panic disorder, C02 challenge test, coronary artery disease, chest pain, normal coronary arteries, diagnostic validation 197
8 1992, Baywood Publishing Co., Inc.
doi: 10.2190/BC4G-WUA0-KN5X-GL63 http://baywood.com
198 / BEITMAN ETAL.
Two studies have established that more than one-third of patients with chest pain and angiographically normal coronary arteries (NCA) fit DSM-I11 criteria for panic disorder [l,21, and a further study showed that more than one-third of such patients fit ICD-9 criteria for anxiety neurosis [3]. Interview studies require additional data to validate their findings. As suggested by Feigner and colleagues, psychiatrists may validate interview findings in several ways, including structured interviews, family studies, follow-up studies, placebo controlled medication trials, and challenge tests [4]. Such studies are crucial in helping to convince skeptical audiences in other medical disciplines of the diagnostic validity of panic disorder in medical populations. The present study reports the results from a pilot study of the effects of 35% carbon dioxide (CO2) inhalation in NCA patients. Since 35% C02 has been shown to trigger panic attacks in psychiatric patients with panic disorder [5, 61, we hypothesized that responses to this challenge would reliably discriminate between those NCA patients who meet criteria for panic disorder and those who do not. METHOD Subjects
The subjects included eleven persons who had both chest pain and a recent normal or near-normal (30%or less stenosis of any major vessel) cardiac catheterization, six of whom had a further interview-based diagnosis of panic disorder (NCA-Panic group)' and five of whom did not (NCA-No Panic group); and ten persons who had no chest pain, history of cardiac catheterization, or panic disorder (Comparison group). Subjects in the NCA groups were recruited via their cardiologists within six months of their catheterizations, and the comparison subjects were recruited via newspaper advertisements. There were no significant differences among the groups in age and sex distribution. Procedure
Before the C02 procedure, all subjects were interviewed using the SCID-UP-R to assess for history of panic disorder, major depression, and substance abuse [7]. They also completed the Zung Self-rating Anxiety Scale [S], Chronic Panic Inventory [6], and Brief Symptom Inventory [9]. These inventories were given in order to better assess baseline anxiety, panic, and other psychiatric symptomatology, respectively. Five of the six NCA-Panic subjects met criteria for panic disorder at the time of the SCID interview. The remaining subject in this group was experiencing panic attacks less frequently, and had met all criteria for panic disorder in the recent past.
CO2 RESPONSE IN CHEST PAIN/PANIC PATIENTS / 199
Subjects were then given a consent form describing the COZprocedure, including information about potential physical symptoms and anxiety reactions. Three sets of inhalations were then performed: 1)practice with mask disconnected from gases; 2) room air; and 3) 35% COZ.For each inhalation, the following procedure was employed: 1) subject completed the Acute Panic Inventory (API), a twentyeight-item checklist of physicaVpsychologica1symptoms [6]; 2) subject took two breaths with the mask on, exhaling for five seconds, inhaling for five seconds, then holding for five seconds; 3) subject exhaled, breathing normally; 4) heart and respiration rates were measured; and 5) subject again completed the API. Ten minutes were allotted between sets, and subjects were blind to the ordering of gases inhaled, although the order did not vary across subjects. The API was completed again ten minutes after the final (COZ)inhalation. Data Analysis Two change scores were used as dependent variables: API after COzminus API after room air (COz - air API), and API after COZminus baseline API (COz baseline API). The second measure is commonly reported in other studies of this type [6], but the first takes into account changes in anxiety produced by the inhalation procedure itself, regardless of gas content (e.g., placebo response), and can be considered a more rigorous test of COZ effects. Because of the high variability in these two change scores and the baseline API, analyses of variance were run on their rank-transformation [101 using group (NCA-Panic, NCA-No Panic, or Comparison) as the independent variable. Group differences in the other questionnaires were tested using analysis of variance on the untransformed data. Tukey’s pairwise comparisonsbetween groups were done with alpha set at .05.
RESULTS All statistical test results are displayed in Table 1. As can be seen, the groups did not differ significantly in history of major depression or substance abuse. However, the NCA-Panic group scored significantly higher on the Chronic Panic Inventory [F(2,20) = 4 . 7 2 , ~= .02], and approached significantly higher scores on the Zung Self-rating Anxiety Scale [F(2,20) = 3.57, p = .05] and baseline API [F(2,20) = 2.87, p .08]. The two NCA groups approached a significant difference from the Comparison group on the BSI total score [F(2,20) = 3 . 2 0 , ~= .05]. The NCA-Panic group scored marginally higher than the other two groups on the COz-airAPI change score [F(2,20) = 2 . 8 0 , = ~ .09], and significantlyhigher on the COz-baseline API change score [F(2,20) = 4.07, p = .03]. These results indicate that the NCA-Panic group was more likely to respond to COz challenge with panic symptoms than the other two groups, although statistical significance was reached only on the more methodologically suspect change score.
200 / BEITMAN ET AL.
Table 1. Initial Anxiety Ratings and Panic Symptom Level in Response to CO2 Inhalation, and Group Characteristics
Variable
NCA-No NCA-Panic Panic Group Group (N = 6) (N = 5)
Baseline API'
7.50
Comparison Group ( N = 10)
3.40
F
p
1.70
2.87
.08*
Con-air API'
(6.06)
(4.45)
(3.02)
14.17 (10.34)
2.80 (0.84)
4.30 (7.30)
20.178
Con-baseline API' .03*
2.60ab
2.80
.Og*
3.00b
4.07
(19.36)
(1.95)
(6.32)
BSI (total score)
49.67 (33.00)
49.75 (44.81)
19.40 (8.98)
3.20
.OF
Chronic Panic Inventory
25.1p (21.77)
15.6Pb (I 3.28)
2.25' (2.71)
4.72
.02**
Zung Anxiety Scale
43.1? (13.14)
33.OPb (8.60)
30.226 (6.22)
3.57
.05*
46.00 (6.81)
42.40 (6.07)
44.90 (4.20)
0.62
.55
2M/4F
3M/2F
5m15f
0.82
.66
History of major depression
50%
60%
30%
1.40
(3)
(3)
(3)
History of substance abuse (%lW2
33%
50%
10%
(2)
(2)
(11
Sex2
.49 2.76 .25
Note: Except where noted, values are means; standard deviations are listed below each mean in parentheses. 'Indicates dependent variable was rank-transformed before analysis. 21ndicatesChi-square rather than f test. 4Bb,bDifferentsuperscripts indicate a significant difference between groups. *p < .10 "p < .05
CO2 RESPONSE IN CHEST PAIN/PANIC PATIENTS / 201
Two Comparison subjects had a marked symptom response to 35% COz; this is to be expected given that about 15 percent of persons without panic disorder respond in this way [5]. Two NCA-Panic subjects did not respond to COZwith increased panic symptoms; this finding also agrees with previous indications that about 35 percent of psychiatric panic disorder patients do not experience marked increase in anxiety under 35 percent COZchallenge [5]. Note that one of these non-responders had a recent history of panic disorder, but did not meet the diagnostic criteria at the time of the study. DISCUSSION
This pilot study provides preliminary support for the idea that patients with angiographicallynormal coronary arteries and panic disorder respond to 35% C 0 2 with a statistically significant increase in panic symptoms, as compared to patients with angiographicallynormal coronary arteries but without panic disorder, as well as to a comparison sample with neither chest pain nor panic disorder. This is the first study to use 35% COZchallenges to support interviews showing that panic disorder exists and is prevalent in patients with angiographically normal coronary arteries [I, 21. The findings of this study are somewhat compromised by several methodological limitations. The most obvious of these is the small number of subjects in the study, which limits its statistical power. Furthermore, because of equipment limitations, we were not able to randomize the order of the COZand room air, so the greater panic response seen after COZ inhalation may have been due to increasing anxiety as the experiment progressed, rather than any effect of the C02 itself. Finally, the person who administered the gases (KMB) was aware of their contents, and she may have given subjects subtle cues for more panic symptoms when she was administering COz gas, compared to when she was administering room air. A further objection to our suggestion that COZ is a valid marker for panic disorder in persons with angiographically normal coronary arteries and chest pain is that subjects in the NCA-Panic group had higher anxiety levels at baseline than either of the other subject groups. Their COZresponses might thus be a function of general anxiety level rather than panic disorder. However, a recent study comparing COz response in panic disorder patients versus those with other anxiety disorders found that only panic disorder patients responded significantly to COZ [ l l ] ; thus, COZresponse appears to be a specific marker of panic disorder. Our findings, although preliminary, fit an emerging pattern of evidence that supports the validity of panic disorder as a diagnosis for some patients with angiographically normal coronary arteries. This includes evidence that patients with angiographicallynormal coronary arteries and panic disorder have a stronger family history of panic disorder and major depression [121, and are less occupationally, physically, and socially functional over time [131 than patients with
202 / BEITMAN ET AL.
angiographically normal coronary arteries who do not have panic disorder. Further, we have preliminary evidence that these patients respond favorably to standard anti-panic medications [14]. Strong efforts must be made to inform cardiologists and primary care physicians of this possible explanation for cardiological symptoms in persons without objective evidence of heart disease. Significant patient relief may result, because panic disorder is an often unrecognized but highly treatable disorder. REFERENCES 1. B. D. Beitman, V. Mukerji, J. W. Lamberti, L. Schmid, L. DeRosear, M. Kushner, G. Flaker, and I. Basha, Panic Disorder in Patients with Chest Pain and Angiographically Normal Coronary Arteries, American Journal of Cardiology, 63:1399-1403, 1989. 2. W. Katon, M. L. Hall, J. Russo, L. Cormier, M. Hollifield, P. P. Vitaliano, and B. D. Beitman, Chest Pain: The Relationship of Psychiatric Illness to Coronary Arteriography Results, American Journal of Medicine, 84: 1-9,1988. 3. C. Bass and C. Wade, Chest Pain with Normal Coronary Arteries: A Comparative Study of Psychiatric and Social Morbidity, Psychological Medicine, 1451-61, 1984. 4. J. P. Feighner, E. Robins, S. B. Guze, R. A. Woodruff, Jr., G. Winokur, and R. Munoz, Diagnostic Criteria for Use in Psychiatric Research, Archives of General Psychiatry, 265743,1972. 5 . E. J. L. Griez, H. Lousberg, M. A. van den Hout, and G. M. van der Molen, C02 Vulnerability in Panic Disorder, Psychiatry Research, 20237-95,1986. 6. M. R. Fyer, J. Uy, J. Martinez, R. Goetz, D. F. Klein, A. Fyer, M. R. Liebowitz, and J. Gorman, C02 Challenge of Patients with Panic Disorder, American Journal of Psychiatry, 144:1080-1082,1987. 7. R. L. Spitzer and J. B. W. Williams, Structured Clinical Interview for DSM-III-RUpjohn Version, Biometrics Research Department, New York State Psychiatric Institute, New York, 1988. 8. W. K. Zung, A Rating Instrument for Anxiety Disorders, Psychosomatics, 22:271-279, 1971. 9. L. R. Derogatis and N. Melisaratos, The Brief Symptom Inventory: An Introductory Report, Psychological Medicine, 13595-605,1983. 10. w. J. Conover and R. L. Iman, Rank Transformations as a Bridge between Parametric and Nonparametric Statistics, American Statistician, 35124-129,1981. 11. E. Griez, J. Zandbergen, H. Pols, and C. de Loof, Response to 35%C02 as a Marker of Panic in Severe Anxiety, American Journal of Psychiatry, 142796797,1990. 12. B. D. Beitman, V. Mukeji, M. Kushner, A. M. Thomas, J. Russell, and M. B. Logue, Validating Studies for Panic Disorder in Patients with Normal Coronary Arteries, Medical Clinics ofNorth America, in press. 13. B. D. Beitman, M. G. Kushner, I. Basha, J. Lamberti, V. Mukerji, and K. Bartels, Follow-up Status of Patients with Angiographically Normal Coronary Arteries and Panic Disorder, Journal of fhe American Medical Association, 265:1545-1549, 1991.
CO2 RESPONSE IN CHEST PAIN/PANIC PATIENTS / 203
14. B. D. Beitman, I. M. Basha, L. H. Trombka, M. A. Jayaratna, B. Russell, G. Flaker, and S. Anderson, Pharmacotherapeutic Treatment of Panic Disorder in Patients Presenting with Chest Pain, Journal of Family Practice, 28177-180,1989.
Direct reprint requests to: Bernard Beitman, M.D. Psychiatry Clinic University of Missouri-Columbia Hospital and Clinics One Hospital Drive Columbia, MO 65212
