Hum Psychopharmacol Clin Exp 2001; 16: 189±192. D01:10.1002/hup.234
Lack of association between joint hyperlaxity and, I: panic disorder, and II: reactivity to carbon dioxide in healthy volunteers Jonathan Benjamin1*, Itzhak Z Ben-Zion1, Pinhas Dannon2, Shaul Schreiber2, Gal Meiri1, Andre Ofek3 and Alex Palatnik4 1
Psychiatry Department, Barzilai Medical Centre, Ashkelon, Israel Anxiety Disorders Clinic, Division of Psychiatry, Chaim Sheba Medical Center, Tel Hashomer, Faculty of Medicine, University of Tel Aviv, Israel 3 Faculty of Health Sciences, Ben Gurion University of the Negev, Beer-sheba, Israel 4 Ministry of Health Mental Health Center, Faculty of Health Sciences, Ben Gurion University of the Negev, Beer-sheba, Israel 2
This study attempted to replicate previous reports of associations between panic disorder and joint hyperlaxity. The authors also examined possible associations between reacitivity to carbon dioxide (CO2), a model for panic vulnerability, and hyperlaxity in healthy volunteers. One hundred and one patients with DSM-IV panic disorder and 39 healthy volunteers were assessed for hyperlaxity by Beighton's criteria. Healthy volunteers also received two vital capacity inhalations of CO2. Thirteen (13%) patients had ®ve or more hyperlax joints. This rate did not differ from that in the healthy volunteers. Anxiety in healthy volunteers, as measured by the NIMH self-rating scale, DSM-IV panic symptom scores, and 100 mm visual analog scales of anxiety, increased after CO2 from a mean of 1.8 to 2.8 (not signi®cant); from 0.5 to 4 ( p < 0.001) and from 8.7 to 11.6 mm ( p < 0.1), respectively. There were no associations between responses to CO2 and hyperlaxity. Copyright # 2001 John Wiley & Sons, Ltd. key words Ð anxiety; hypermobility, joint; joint instability
INTRODUCTION Bulbena and colleagues have reported a `peculiar, suggestive association' (Bulbena et al., 1993) between joint hypermobility1 and panic disorder. They found the association both in the form of an elevated prevalence of panic disorder in patients with joint hyperlaxity compared to rheumatological controls (Bulbena et al., 1988, 1993) and in the form of an elevated prevalence of joint hyperlaxity in patients with panic disorder (68%) compared to psychiatric controls (10%) (Martin-Santos et al., 1998). They have also found associations between panic disorder and *Correspondence to: Prof. J. Benjamin, Psychiatry Department, Barzilai Medical Centre, Ashkelon, Israel. Tel: +972-8-674-5203. Fax: +972-8-674-5255. E-mail: [email protected]
1 ``Hypermobility'' is the customary term. We have preferred the more accurate ``hyperlaxity'' in this paper. Copyright # 2001 John Wiley & Sons, Ltd.
mitral valve prolapse, and between mitral valve prolapse and joint hyperlaxity, but the strongest association, with odd ratios as high as 20:1 (Bulbena et al., 1993), was between panic disorder and joint hyperlaxity itself. This ®nding, if replicable elsewhere, suggests renewed searches for a novel etiological or pathogenetic mechanism of panic disorder, perhaps involving connective tissues. Here we report that we did not ®nd an elevated prevalence of joint hyperlaxity in 101 Israeli patients with panic disorder compared to the reported prevalence in the general population, or compared to the prevalence in 39 healthy controls. In addition, reactivity to carbon dioxide (CO2) panicogenesis in symptom-free ®rst-degree relatives of patients with panic disorder is higher than in normal controls (Perna et al., 1995), suggesting that reactivity to CO2 may serve as a marker for an underlying, hidden vulnerability to panic disorder. We therefore Received 1 April 2000 Accepted 20 June 2000
j. benjamin ET AL.
also investigated possible associations between reactivity to CO2 and joint hyperlaxity in young healthy volunteers. Parametric and non-parametric tests failed to detect any association between reactivity to CO2 and joint hyperlaxity. METHODS Study 1 One hundred and one consecutive patients ful®lling DSM-IV criteria for panic disorder, with or without agoraphobia, but with no other psychiatric diagnosis, were recruited at three anxiety disorder clinics in Israel. Joint hyperlaxity was assessed according to Beighton's criteria (Beighton et al., 1973, 1989); for non-parametric analyses a subject with ®ve or more (out of nine) hyperlax joints (Martin-Santos et al., 1998) was considered a case. On 2 testing days, at one center, 20 of the patients with panic disorder, together with 20 psychiatric outpatient controls, were assessed by an examiner who was blind regarding the diagnosis. Ten of these patients were assessed independently by two examiners, and inter-rater reliability was assessed.
by the institutional review board. Subjects took two vital capacity inhalations of 35% CO2 within 10 s. Because previous studies of CO2 anxiogenesis have not found anxiety after inhalations of room air, and because in our experience it is impossible to blind subjects to the identity of the gases, we did not employ a room air control. Immediately before and after the CO2 inhalations volunteers completed the NIMH self-rating scale of physical and mental symptoms (Murphy et al., 1989), a checklist of the DSMIV symptoms of a panic attack, each rated from 0 to 3 (`panic symptoms scale', PSS), and seven 100 mm visual analog scales (VASs) of anxiety (Benjamin et al., 1996). A panic attack was de®ned as a score of ``moderate'' or higher on at least four DSM-IV symptoms, one of which had to be a mental symptom (Benjamin et al., 1996). Tests of association with joint hyperlaxity used the difference (after CO2 minus before CO2) for each of the three measures. Joint hyperlaxity was assessed according to Beighton's criteria (Beighton et al., 1973, 1989) independent of the CO2 inhalations, and the physician responsible for one assessment was unaware of the result on the other. RESULTS
Study 2 Thirty-nine medical students were recruited. They were screened with the overview section of the SCID questionnaire (Spitzer et al., 1990) and a standard medical interview and examination to exclude psychiatric and medical illness, drug or medication use. All gave written, informed consent after the procedure had been explained, and the study was approved
Study 1 The mean ( standard deviation) age of the patients with panic disorder was 39.3 (11) years. Sixty-®ve (65%) were female. Table 1 shows the frequency of hyperlax joints in patients and controls. Thirteen (13%) out of 101 panic disorder patients had ®ve or more hyperlax joints. This ®gure is within the
Table 1. Frequency of number of hyperlax joints, by Beighton's criteria (range 0±9), in 101 patients with panic disorder and 39 healthy controls # of affected joints 0 1 2 3 4 5 6 7 8 9
Cumulative frequency (percent)a
38 5 26 9 10 4 1 1 3 4
17 5 4 6 1 2 3 0 0 1
101 (100) 63 (63) 58 (58) 32 (32) 23 (23) 13 (13) 9 (9) 8 (8) 7 (7) 4 (4)
39 (100) 22 (56) 17 (44) 13 (33) 7 (18) 6 (15) 4 (10) 1 (3) 1 (3) 1 (3)
The ®gures in these two columns refer to numbers of subjects with n or more hyperlax joints, e.g., there are 13 patients and six healthy controls with ®ve or more hyperlax joints.
Copyright # 2001 John Wiley & Sons, Ltd.
Hum Psychopharmacol Clin Exp 2001; 16: 189±192.
joint hyperlaxity and anxiety
Table 2. Mean ( SD) scores on three measures of anxiety before and after two vital capacity inhalations of carbon dioxide in 39 healthy volunteers Measure
t-test (38 df)
NIMH anxiety Panic symptom score Visual analogue scales (100 mm)
1.8 (3) 0.5 (1) 8.7 (11)
2.8 (5) 4.0 (7) 11.6 (14)
1.5 3.6 1.7
NS < 0.001 < 0.10
12±15% range for the general population cited by Bulbena and colleagues (Martin-Santos et al., 1998) and reported by Gedalia et al. (1993) in Israel, and did not differ from the frequency in the healthy volunteers (six (15%) out of 39, X2 0.24, df 1, not signi®cant). None of the 20 patients with panic disorder who were assessed by an examiner blind to diagnosis had hyperlaxity by the above criteria. Inter-rater reliability for hyperlaxity in 10 patients examined by two examiners was 100% (there were disagreements about a single joint in two instances, but in neither instance was the hyperlax `caseness' of the patient affected). Study 2 The mean ( standard deviation) age of the healthy controls was 23.4 (3) years. Fourteen (36%) were female. Two subjects panicked in response to CO2; one of these had joint hyperlaxity. If we relax our criteria for a panic attack to those of DSM-IV (any four symptoms on the PSS, not necessarily including a fear symptom) ®ve subjects panicked, and two of these had joint hyperlaxity. This frequency of hyperlaxity did not differ from the frequency of hyperlaxity in those subjects who did not have a panic attack (®ve out of 33); X2 with Yates' correction 0.47, df 1, not signi®cant; Fisher's exact test p 0.22. Table 2 shows mean scores on the three quantitative measures of anxiety before and after CO2. There were no significant correlations between number of hyperlax joints and changes in any measure of anxiety after CO2 in healthy volunteers. For the NIMH anxiety subscale r 0.06, not signi®cant; for the PSS r 0.08, not signi®cant; for the VAS r 0.09, not signi®cant. The correlation matrix for the NIMH anxiety subscale and Beighton's score is shown in Figure 1; the virtually ¯at line illustrates the absence of interaction between the two measures. Analysis of covariance with age and sex as covariates found no signi®cant differences in measures of anxiety between subjects with and without hyperlaxity in ®ve or more joints. For the NIMH anxiety subscale F ratio 0.1, df 1,35, not signi®cant; for the PSS F ratio 1.8, df 1,35, Copyright # 2001 John Wiley & Sons, Ltd.
Figure 1. Lack of association between number of hyperlax joints, on the x-axis, and positive or negative change in NIMH anxiety score after carbon dioxide, on the y-axis, in 39 healthy volunteers; Pearson's r 0.06, not signi®cant; y 0.11x 0.79
not signi®cant; for the VAS F ratio 0.01, df 1,35, not signi®cant.
DISCUSSION Finding a positive association between reactivity to CO2 and joint hyperlaxity in healthy volunteers would have been highly suggestive convergent evidence consistent with the association in patients reported by Bulbena and colleagues. However, failure to ®nd such an association in healthy volunteers says little, for we lack ®rm data about the frequency of a possible `panic diathesis' in the general population. The reports of associations between panic disorder and joint hyperlaxity by Bulbena and colleagues were not replicated in Israeli patients with panic disorder. This suggests that the association found in Barcelona may be due to a genetic founder effect in that region (chance association of panic disorder and joint hyperlaxity in a founder) or due to a negative founder effect (absence of one of these phenotypes in the presence of the other) in a Jewish Israeli founder. Unless the association is replicated elsewhere, efforts to Hum Psychopharmacol Clin Exp 2001; 16: 189±192.
j. benjamin ET AL.
ascertain underlying cause-and-effect mechanisms may be premature. REFERENCES Beighton P, Grahame R, Bird H. 1989. Hypermobility of Joints (2nd edn). Springer-Verlag: London. Beighton P, Solomon L, Soskolne C. 1973. Articular mobility in an African population. Ann Rheum Dis 32: 413±418. Benjamin J, Greenberg B, Murphy D. 1996. Daily administration of m-chlorophenyl-piperazine to healthy human volunteers rapidly attenuates many of its behavioral, hormonal, cardiovascular and temperature effects. Psychopharmacology 127: 140±149. Bulbena A, Dure J, Porta M et al. 1993. Anxiety disorders in the joint hypermobility syndrome. Psychiatry Res 46: 59±68.
Copyright # 2001 John Wiley & Sons, Ltd.
Bulbena A, Duro J, Mateo A, Porta M, Vallejo J. 1988. Anxiety disorders in the joint hypermobility syndrome. Lancet II: 694. Gedalia A, Press J, Klein M, Buskila D. 1993. Joint hypermobility and ®bromyalgia in schoolchildren. Ann Rheum Dis 52: 494±496. Martin-Santos R, Bulbena A, Porta M, Gago J, Molina L, Duro J. 1998. Association between joint hypermobility syndrome and panic disorder. Am J Psychiatry 155: 1578±83. Murphy D, Mueller E, Hill J, Tolliver T, Jacobsen F. 1989. Comparative anxiogenic, neuroendocrine, and other physiologic effects of m-chlorophenylpiperazine given intravenously or orally to healthy volunteers. Psychopharmacology 98: 275±282. Perna G, Cocchi S, Bertani A, Arancio C, Bellodi L. 1995. Sensitivity to 35% CO2 in healthy ®rst-degree relatives of patients with panic disorder. Am J Psychiatry 152: 623±5. Spitzer R, Williams J, Gibbon M, First M. 1990. Structured Clinical Interview for DSM-III-R (SCID, version 1.0). American Psychiatric Press: Washington, DC.
Hum Psychopharmacol Clin Exp 2001; 16: 189±192.