- Email: [email protected]
Association Between Joint Hypermobility Syndrome and Panic Disorder: A Case–Control Study
Association Between Joint Hypermobility Syndrome and Panic Disorder: A Case–Control Study Javier Garcı´a Campayo, M.D., Elena Asso, M.D. Marta Alda, M.D., Eva Maria Andres, Ph.D. Natalia Sobradiel, Ph.D.
Background: A significant association between joint hypermobility syndrome (JHS) and panic disorder was observed in a sample of rheumatology outpatients. Objective: The aim of this study was to assess whether JHS is more frequent in panic-disorder than in control subjects. Method: The authors conducted a case– control study comparing 55 untreated patients with panic disorder and three matched-control groups: psychiatric patients, fibromyalgia patients, and healthy persons. Results: JHS was more frequent among panic-disorder than among psychiatric patients, the healthy group, or the fibromyalgia group. In the panic-disorder group, there was a significant correlation between severity of JHS and anxiety. Conclusion: The strong association between JHS and panic disorder points to a genetic association. There is also a possibility that JHS and mitral valve prolapse, another condition frequently associated with panic disorder, share a common pathophysiological mechanism. (Psychosomatics 2010; 51:55– 61)
J
oint hypermobility syndrome (JHS) is a highly hereditary clinical condition characterized by increased distensibility of joints in passive movements and hypermobility in active movements.1 Its prevalence in the general population appears to range between 10% and 15%, and it is more common in women than in men.1 Although JHS often goes unnoticed, affected individuals may suffer from repeated injuries of the musculoskeletal system,2 so its early diagnosis is recommended to avoid these complications. A significant association between JHS and panic disorder was observed in a sample of rheumatology outpatients.3,4 There is also a report on the high association of panic disorder and/or agoraphobia with JHS (the odds ratio [OR] compared with psychiatric control subjects was 18.6, and the OR compared with medical control subjects was 14.7).5 Also, another study relates JHS to both state and trait anxiety (the former with less intensity) in the general population.6 The magnitude of the JHS and panic disorder association, uncommon in medical research, points to a causal connection between these two conditions. Grataco´s et al.7 Psychosomatics 51:1, January-February 2010
identified an interstitial duplication of the human chromosome 15q24 –26 (named DUP25), which was significantly associated with panic/agoraphobia/social phobia/joint laxity in families, and with panic disorder in nonfamilial cases. Mosaicism, different forms of DUP25 within the same family, and absence of segregation of 15q24 –26 markers with DUP25, and the associated psychiatric phenotypes, suggest a non-Mendelian mechanism of diseasecausing mutation. These authors7 proposed that DUP25, which is present in 7% of the control subjects, is a susceptibility factor for a clinical phenotype that includes Received August 27, 2007; revised October 31, 2007; accepted November 26, 2007. From the Dept. of Psychiatry, Miguel Servet University Hospital and University of Zaragoza, Spain; Casablanca Health Centre, Zaragoza, Spain; Statistical Methods Department, Faculty of Mathematics, Zaragoza, Spain; Grupo Aragone´s de Investigacio´n en Atencio´n Primaria; Red de Actividades Preventivas y Prevencio´n de la Salud (REDIAPP G03/170); and the Instituto Aragone´s de Ciencias de la Salud. Send correspondence and reprint requests to Javier Garcia Campayo, Dept. of Psychiatry, Miguel Servet University Hospital, Avda Isabel La Cato´lica 1, 50.009 Zaragoza, Spain. e-mail: [email protected] © 2010 The Academy of Psychosomatic Medicine
http://psy.psychiatryonline.org
55
Joint Hypermobility and Panic Disorder panic and phobic disorders and joint laxity. However, this genetic finding was not confirmed by two other research groups, who achieved different results.8,9 To date, most of the research papers published on the association between JHS and panic disorders/agoraphobia3–7 have been developed by the same team of researchers. To our knowledge, three replication studies have been carried out, with inconclusive results: two10,11 failed to confirm this association; the first, probably owing to the unbelievably high JHS percentage (52.6%) found in the control group; and the second, probably resulting from methodological limitations,11 such as the kind of instruments used. The third study, the only one to confirm the association between panic disorder and JHS,12 has only been published in Turkish. The present study was designed to replicate the research of the Barcelona group and, more specifically, to answer the following questions: 1) Is JHS more common in patients with panic disorder than among appropriate comparison subjects (psychiatric controls, medical controls, and healthy people)? and 2) Does severity of panic/anxiety symptoms correlate with JHS severity in these groups? METHOD A Case–Control Study Setting The study was carried out in the Torrero, Casablanca, and Arrabal primary-care health centers, in the city of Zaragoza, Spain. Patients Four groups of patients were included: a Case group of patients diagnosed with panic disorder, and three other Control groups (psychiatric, medical, and healthy) that were selected on the basis of the two previous papers published on the subject5,6 and matched one-to-one by sex, age (plus-or-minus 2 years), and ethnic group, to the group of panic-disorder patients. The following is a description of each sample group: Patients diagnosed with panic disorder: All consecutive, newly diagnosed, and untreated patients attending the aforementioned primary-care health centers over an 18month period (from January 2005 to June 2006) who fulfilled inclusion criteria were included. We applied the following criteria: 1) DSM–IV criteria for panic disorder (with or without agoraphobia); 2) age 18 – 65 years; 3) ability to understand and read Spanish. Exclusion criteria: 1) Any other DSM–IV Axis I psychiatric disorder, such as psychosis, drug abuse, cognitive disorder; 2) severe medical disorders that hamper physical (including joint) examination. The psychiatric control group: These patients were 56
http://psy.psychiatryonline.org
recruited from the same primary-care health centers and included consecutive, newly diagnosed psychiatric patients who did not suffer from DSM–IV panic disorder and who had never met criteria for any anxiety disorder. This selection criterion was also used by the original authors.5 Inclusion criteria: 1) any DSM–IV psychiatric diagnosis except those described in the exclusion section; 2) age: 18 – 65 years; and 3) untreated patients, or patients able to discontinue the treatment at least 2 weeks before the JHS assessment. Exclusion criteria: 1) lifetime diagnosis of any anxiety disorder; 2) any psychiatric diagnosis that, from the point of view of the clinician, would not allow the study to be completed (e.g., severe psychosis or personality disorder, cognitive deterioration, drug abuse); and 3) severe medical disorders that hamper physical (including joint) examination. Many of these patients were getting pharmacological treatment and were not able to discontinue them before the assessment. The medical control (fibromyalgia) group: The original authors5 recruited this second control group from four medical outpatient clinics (dermatology, gastroenterology, general surgery, and internal medicine) at the same hospital. We preferred to include in this group only patients with a specific rheumatologic disorder, fibromyalgia, which has been reported to show a high association with JHS. In the Spanish population, this association has been calculated to be 27.3%.13 All consecutive, newly diagnosed patients with fibromyalgia attending the aforementioned primary-care health centers were invited to participate in the study. Inclusion criteria: 1) diagnosis of primary fibromyalgia by a rheumatologist according to the American College of Rheumatology criteria;14 and 2) age: 18 – 65. Exclusion criteria: 1) any psychiatric diagnosis that, from the point of view of the clinician, would not allow the study to be completed (e.g., severe psychosis or personality disorder, cognitive deterioration, or drug abuse); and 2) severe medical disorders that hamper physical (including joint) examination. Every effort was made to recruit only untreated patients; however, most of these patients were taking various kinds of analgesics and/or antidepressants. They were asked to discontinue the treatment at least 2 weeks before the JHS assessment. The healthy-control group: These were persons accompanying patients with panic disorder, or other psychiatric disorders, who were visiting the previously mentioned primary-care health centers and who, according to the Standardized Polyvalent Psychiatric Interview (the SPPI), did not suffer from any psychiatric disorders. A sample group of this type was also assessed by the original authors6 to evaluate JHS’s association with anxiety in Psychosomatics 51:1, January-February 2010
Campayo et al. nonclinical populations. Inclusion criteria: 1) no DSM–IV psychiatric diagnosis; and 2) age: 18 – 65. Exclusion criteria: 1) severe medical disorders that would hamper physical (including joint) examination. Sample: Sample size (N⫽55 for each of the four groups) was calculated by standard methods for ␣ at 5% (two-tailed),  at 80%, and a difference of odds ratios (OR) between the panic-disorder group and control group of ⬎3.15 In total, a sample of 220 subjects was recruited. All control subjects were matched to the panic-disorder group patients by sex, age (within 2 years), and ethnic group. Nonetheless, given the predominance of women among fibromyalgia patients, it was impossible to match this group by gender. To obtain the expected sample, a total of 58 panic-disorder patients were interviewed, with 3 refusing to participate in the study. For the three control groups, it was necessary to interview 60 psychiatric patient (5 refused), 56 fibromyalgia patients (only 1 refused), and 59 control subjects (4 refused). To replace participants who had refused during recruitment, the first subject attending the health center who fulfilled the inclusion criteria with the same sex, age, and ethnicity of the individual who had refused was invited to participate.
original research, is more specific for the assessment of panic and agoraphobia symptoms. State–Trait Anxiety Scale (STAI), Spanish version24 We used this scale for the assessment of anxiety in healthy subjects, in order to replicate the previous study of the group on nonclinical individuals in which this questionnaire was used. It seems more appropriate than the PAS for this specific population, and it yields information about state- and trait-anxiety. This questionnaire was also used with psychiatric patients because panic disorder had been specifically excluded from this group. Hospital Anxiety and Depression Scale (HADS), Anxiety Subscale The HADS25 is a self-report scale that screens for the presence of depression and anxiety in patients with organic disorders. It comprises 14 items that are rated on a 4-point Likert-type scale, and it is appropriate for use in community and hospital settings. Two subscales, HADS– Dep and HADS–Anx, independently assess depression and anxiety. The HADS was validated in a Spanish population.26 HADS was selected for use in the present study because it is considered one of the best questionnaires for assessing depression and anxiety in patients with physical illness.27
Instruments Standardized Polyvalent Psychiatric Interview (SPPI) The SPPI is a psychiatric interview developed by our group for the multiaxial assessment of psychiatric morbidity in outpatient psychiatric clinics and primary-care settings.16 It permits the use of various diagnostic criteria, including DSM–IV and ICD–10. Our group has wide research experience with this interview.17–21 Panic and Agoraphobia Scale (PAS)22 The PAS is the first-ever scale for assessing the severity of panic disorder with or without agoraphobia. Compatible with both DSM–IV and ICD–10 classifications, and available in both self-rated and observer-rated versions, the PAS was specially developed for monitoring the efficacy of both drug and psychotherapy treatments. It has been translated and validated in Spanish,23 has excellent psychometric properties,22,23 and is quick to use: the observer-rated version that we used can be completed in 5 to 10 minutes. We used this scale for the assessment of panic symptoms in the panic-disorder group. This questionnaire is considered more suitable than the scale used by the original authors,5 the Hamilton Rating Scale for Anxiety, because the PAS, which was not available for the authors at the time of the Psychosomatics 51:1, January-February 2010
Assessment of Joint Hypermobility Syndrome (JHS) To assess joint hypermobility syndrome (JHS), we used Beighton’s criteria.1 The scores range from 0 to 9 and depend on the degree of movement. Subjects who scored ⱖ5 on the 9-point scale were considered to have JHS. The reliability and validity of the assessment criteria for JHS have been previously described.28 Procedure Patients were recruited from our research department by two researchers (NS, EMA). Two experienced psychiatrists (JGC, MA) administered the SPPI psychiatric interview to establish the DSM–IV psychiatric disorder and rated the psychological questionnaires used in this study (the PAS, STAI, and HADS). In a later encounter, a family doctor (EA), unaware of either psychiatric diagnosis or anxiety scores, assessed JHS by use of Beighton’s criteria. Patients were informed at the beginning of the study that they were not to give any unsolicited information, so as to maintain the blinding. After the interviews with patients and examiners (JGC, MA, EA), the two research assistants (NS, EMA) confirmed that blinding had not been broken. In http://psy.psychiatryonline.org
57
Joint Hypermobility and Panic Disorder
TABLE 1.
Sociodemographic and Clinical Characteristics of the Sample, Mean (SD)
Sex, female, N (%) Age, years Ethnicity, white European, N (%) PAS HADS STAI-trait STAI-state Total N: JHS Beighton’s criteria
Panic Disorder (Nⴝ55)
Psychiatric Controls (Nⴝ55)
Fibromyalgia (Nⴝ55)
Healthy Controls (Nⴝ55)
Difference, p
45 (81.8%) 38.1 (10.5) 55 (100%) 22.5 (7.8) 6.5 (3.0) 20.5 (9.8) 21.7 (8.8) 5.4 (2.3)
45 (81.8%) 39.4 (10.8) 55 (100%) 6.7 (2.3) 5.8 (2.8) 19.4 (9.6) 20.5 (9.8) 2.4 (2.3)
47 (85.4%) 40.1 (10.9) 55 (100%) 7.8 (2.6) 6.2 (3.1) 17.8 (8.2) 18.6 (7.9) 2.7 (2.6)
45 (81.8%) 38.2 (10.4) 55 (100%) 5.6 (1.7) 2.3 (0.8) 11.9 (6.4) 12.1 (7.5) 2.4 (2.3)
NSa NSb NSa ⬍0.001c ⬍0.05c ⬍0.01c ⬍0.01c ⬍0.001c
Values are mean (standard deviation [SD]), unless otherwise indicated. PAS: Panic and Agoraphobia Scale; HADS: Hospital Anxiety and Depression Scale. a chi-square. b **Student t-test, paired samples. c ***Mann-Whitney U test.
order to adequately replicate previous studies, the family doctor was standardized with the members of the Barcelona, Spain, team, obtaining an interrater reliability (kappa) of 0.85– 0.90 for the different items of the Beighton’s criteria. The research was approved by the Ethical Review Board of the regional health authority, and informed consent from the patients was obtained after the procedures had been fully explained. Statistical Analysis In contingency tables, Pearson’s chi-square test for homogeneity or independence was applied to assess the relationship between two qualitative or categorical variables. In 2⫻2 chi-squares, the Yates-corrected chi-square value is given. Where 20% or more of the cells had expected counts of ⬍5, Fisher’s exact test was used.29 Student’s t-test and one-factor analysis of variance were used to determine the relationship between a categorical variable with two or three levels, respectively, and a normally distributed quantitative variable; the Mann-Whitney U test was used to analyze quantitative variables that did not have normal distribution.30 The normality of the variables was contrasted with a Kolmogorov-Smirnov test. Spearman rho () was used to determine correlation between different variables because distributions were not normal. The logit estimator of the prevalence odds ratio (OR) was estimated with precision-based confidence intervals (CI).29 Data management and statistical analysis were carried out using the SPSS 14 Statistical Package. RESULTS The mean age of the 220 subjects of the study was 41.3 years (standard deviation [SD]: 11.0); 82.7% (182/220) 58
http://psy.psychiatryonline.org
were women. As shown in Table 1, Case and Control groups did not significantly differ by age, sex, or ethnicity because of the matching process, despite a slight predominance of women in our fibromyalgia group. It was impossible for us to recruit a matching number of men with fibromyalgia (female predominance in this disorder has been calculated at 7/1 to 9/1).31 All of the patients were white Europeans. Of the anxiety cases, 70.9% (39/55) were diagnosed as having panic disorder with agoraphobia, and the remaining 29.1% (16/55) of panic disorder without agoraphobia. The distribution of case severity according to SPPI was as follows: 50% (28/55) light, 30% (17/55) moderate, and 18.1% (10/55) severe; 20 of 55 (36.3%) were also diagnosed with major depression or dysthymia. In the psychiatric control group, diagnostic distribution was as follows: major depressive disorder (36/ 55; 65.4%), dysthymia (9/55; 16.3%), schizophrenia (8/ 55; 14.5%), and somatoform disorders (6/55; 10.9%), with some of the patients having more than one psychiatric diagnosis. Fibromyalgia patients were also diagnosed with depression in 32.7% of cases (18/55), and 4 of these (7.2%) were diagnosed with panic disorder. Panic and Joint Hypermobility Syndrome (JHS) Table 1 summarizes the mean PAS score (range: 0 –54), the mean STAI-trait score (range: 0 – 60), the mean STAI-state score (range: 0 – 60), and the mean HADS-Anx score (range: 0 –21) for all the groups. It also shows the mean Beighton’s criteria for JHS scores (range: 0 –9) for the four different groups. As shown, there are significant differences between Beighton’s criteria scores for the panic-disorder group, as compared with healthy-control subPsychosomatics 51:1, January-February 2010
Psychosomatics 51:1, January-February 2010
2.0–14.1 5.4 — ⬍0.001 0.215 11.2 1.5 12/40 (30) 2/15 (13.3) 6.1–72.1 21 — ⬍0.001 0.085 27.5 2.9 4/40 (10) 1/15 (6.6) 1.0 1.0 28/40 (70) 6/15 (40)
N⫽55 for all groups. OR: odds ratio; CI: confidence interval. a Reference group. b With Yates correction (df: 1).
5.1–51.8 ⬍0.001 16.3 0.085 — 24.9 2.9
1.0 1.0 4/10 (40) 30/45 (66.6)
5/40 (12.5) 1/15 (6.6)
2.1–12.7 — 5.2 0.444 ⬍0.001 0.5 12.5 1/8 (12.5) 13/47 (27.6) 6.1–68.0 — 20.5 0.302 ⬍0.001 1.0 29.5 1/10 (10) 4/45 (8.8) 5.2–48.9 0.302 — ⬍0.001 16 1.0 26.9
4.8–36.2 ⬍0.001 13.2 28.6 1.0 34 (61.8)
Total Gender Men Women Age ⬍45 years ⱖ45 years
6 (10.9)
N (%) OR N (%)
1/10 (10) 5/45 (11.1)
95% CI
2.0–10.7 4.7
OR p
⬍0.001 13.3 14 (25.4) 5.5–47.1 16.1 ⬍0.001 31.1
N (%) 95% CI p
OR
5 (9.0)
2 N (%) p
OR
95% CI
Fibromyalgia Control Group Psychiatric Control Group
2 2b
Healthy-Control Group Panic-Disorder Groupa
Group
TABLE 2.
jects (Mann-Whitney U⫽622.5; Z ⫽ –5.3; p ⬍0.001), psychiatric group (Mann-Whitney U⫽391; Z ⫽ – 6.7; p ⬍0.001), and patients with fibromyalgia (Mann-Whitney U⫽631.5; Z ⫽ –5.3; p ⬍0.001). As is summarized in Table 2, JHS according to Beighton’s criteria was significantly more frequent among patients with panic disorder (34/55; 61.8%) than among psychiatric-control patients (6/55; 10.9%); Yates-corrected 2: 2.6; df⫽1; p ⬍0.001), with an OR of 13.2, and the 95% CI: 4.8 –36.2. JHS is also much more prevalent in the panic-disorder group than in the healthy-control group (5/55; 9.0%); Yates-corrected 2: 33.8; df⫽1; p ⬍0.001), with OR 20.6, and 95% CI: 6.5– 65.4. It is also more frequent in the panic-disorder group than in the medicalcontrol group with fibromyalgia (14/55; 25.4%); Yatescorrected 2: 13.3; df⫽1; p ⬍0.001; OR: 4.7 (95% CI: 2.0 –10.7). Stratified analysis by sex and age (cutoff point: 45 years) showed highly significant differences in women and in the sample under age 45, but failed to find these group differences for persons age 45⫹ (p⫽0.08) and for men. There was a significant correlation, in patients with panic disorder, between the number of Beighton’s criteria and PAS (Spearman ⫽0.520; p ⬍0.001), but not with STAI-state (Spearman ⫽0.234; p⫽0.411), STAI-trait (Spearman ⫽0.298; p⫽0.376) or HADS (Spearman ⫽0.098; p⫽0.706). For the other groups of patients, there was no correlation between Beighton’s criteria and any of the different questionnaires on anxiety (PAS, HADS, STAI-state, or STAI-trait). Among patients with panic disorder, we compared the groups with and without JHS. We found that the subgroup with JHS was younger (mean age: 35.2 years; SD: 10.7) than patients with panic disorder but without JHS (mean age: 41.7 years; SD: 9.8; t ⫽ –2.2; df: 53; p ⬍0.05). PAS scores were also higher in patients with both panic disorder and JHS (mean: 19.8; SD: 4.3) than in the panicdisorder group without JHS (mean: 13.7; SD: 6.0; MannWhitney U⫽137; Z ⫽ –3.8; p ⬍0.001). There was no difference in the proportion of women in the hypermobility group (30/34; 88%), as compared with the non-hypermobility subjects (15/21; 71.4%; Yates-corrected 2: 1.4; df: 1; p⫽0.226). For panic symptoms measured with the SPPI interview, the only statistical differences between Case and Control groups were on the items “trembling or shaking” (30/34; 88.2% versus 10/21; 47.6%; OR: 8.2; 95% CI: 2.1–31.8) and “chills or hot flashes” (31/34; 91.1% versus 12/21; 57.1%; OR: 7.7; 95% CI: 1.7–33.5). However, total
Prevalence of Joint Hypermobility Syndrome (JHS) in Primary-Care Patients With Panic Disorder, Healthy-Control Subjects, Psychiatric Control Subjects, and Fibromyalgia
Campayo et al.
http://psy.psychiatryonline.org
59
Joint Hypermobility and Panic Disorder number of panic symptoms, severity of panic symptoms measured with SPPI, and age at onset of the disorder were not significantly different between groups. There was also no difference in agoraphobia scores. DISCUSSION This is the first study, to our knowledge, that replicates the data of Bulbena et al.3–7 We have used the same design and have standardized the JHS diagnosis according to this group, obtaining adequate kappa values. Some of the strengths of our study are the following: First, sample recruitment was carried out in primary-care settings, thus avoiding selection bias of specialized care. Second, we used a psychiatric interview to confirm all the psychiatric diagnoses (or absence of psychiatric disorder). And, third, we utilized a more sophisticated assessment of panic severity symptoms, the PAS questionnaire. In our study, the prevalence of JHS according to Beighton’s criteria in subjects with panic disorder was 61.8%, a figure quite similar to the 67.7% described by Bulbena’s group,5 and significantly higher than the 10.9% observed in our healthy-control subjects (OR⫽13.2; 95% CI: 4.8 –36.2), and the 9% observed in our psychiatriccontrol group (OR: 16.1; 95% CI: 5.5– 47.1). The figures found in the healthy-control group and the psychiatric group are similar to those described by Bulbena et al.5 and similar to the 10%–15% described for the general population.33 It is remarkable that JHS is even more frequent in panic disorder (OR: 4.7; 95% CI: 2.0 –10.7) than among the medical control group with fibromyalgia (25.4%), a disorder that has also been associated with JHS.13 The figures we have found in our sample are similar to the 27.3% detected in other studies in Spanish patients with fibromyalgia.13 When we carried out a stratified analysis of JHS prevalence by sex and age, we found highly significant differences for women and the younger age-group, but failed to find these differences for persons age 45⫹ and for men. The reason for this finding could be the limitation resulting from the small sample size in each of the four groups. The power calculation for this stratification analysis was 0.72, slightly lower than the 0.80 usually used. These data are similar to those found by Bulbena et al.,5 who described the highest odds ratio for JHS (more than double) in women and younger people in both the panic-disorder and psychiatric-control groups. We have also found that, in patients with panic disorder, there is a significant correlation between the number of Beighton’s criteria and PAS scoring, but not with the 60
http://psy.psychiatryonline.org
other tests of anxiety used (STAI-trait, STAI-state, HADS). As far as we are aware, this is the first time that this correlation has been described in scientific literature, and it reinforces a possible genetic association between panic disorder and JHS, as has been suggested.7 However, we did not find any correlation between the number of Beighton’s criteria and STAI-state, STAI-trait, or HADS scoring in healthy individuals, psychiatric patients, or patients with fibromyalgia. This surely implies that JHS only correlates with symptoms of panic disorder (measured with the PAS), but not with nonspecific anxiety symptoms measured with the other questionnaires. The sociodemographic and clinical characteristics of the JHS panic-disorder subgroup are somewhat nonspecific. Apart from a predominance of younger people, higher scoring in PAS, and certain clinical symptoms, these patients are strikingly similar to the remaining patients with panic. The implications of these findings are the following: 1) From a genetic perspective, the strong association described between JHS and panic disorder points to a genetic association. The original authors7 proposed DUP25 as the possible genetic factor related to both disorders. Despite the fact that other, later studies8,9 failed to confirm this genetic finding, it seems that there is a need to continue to search for the genetic basis of this association. 2) From a neurobiological perspective, this finding also reinforces the finding that JHS and mitral valve prolapse, another disorder frequently associated with panic disorder, share a common pathophysiological mechanism that some authors have thought could be a collagen disorder.33,34 Perhaps a possible brain fear-circuitry could exist that may be common to these different conditions. 3) From a clinical point of view, the frequent underdiagnosis of both disorders make it advisable to screen for JHS whenever a doctor diagnoses anxiety, and vice versa. Also, based on their family aggregation, it would be also recommended to screen close relatives so as to rule out both disorders. In summary, we confirmed the strong association of JHS and panic disorder and found that there was no association with healthy subjects, psychiatric controls, or medical controls. There is also a correlation between the severity of panic symptoms and the severity of JHS in patients with panic disorder. The etiology of the association of these two syndromes is still unclear. However, JHS seems to be a constitutional characteristic of many patients with panic disorder.
The authors report no financial conflicts of interest. Psychosomatics 51:1, January-February 2010
Campayo et al.
References 1. Beighton PH, Grahame R, Bird H: Hypermobility of Joints, 2nd Edition. London, UK, Springer-Verlag, 1989 2. Lewkonia RM: The biology and clinical consequences of articular hypermobility. J Rheumatol 1993; 20:220 –222 3. Bulbena A, Duro JC, Porta M, et al: Anxiety disorders in the joint hypermobility syndrome. Lancet 1988; 2:694 4. Bulbena A, Duro´ JC, Porta M, et al: Anxiety disorder in the joint hypermobility syndrome. Psychiatry Res 1993; 43:59 – 68 5. Martı´n-Santos R, Bulbena A, Porta M, et al: Association between joint hypermobility syndrome and panic disorder. Am J Psychiatry 1998; 155:1578 –1583 6. Bulbena A, Agullo´ A, Pailhez G, et al: Is joint hypermobility related to anxiety in a non-clinical population also? Psychosomatics 2004; 45:432– 437 7. Gratacos M, Nadal M, Martin-Santos R, et al: A polymorphic genomic duplication on human chromosome 15 is a susceptibility factor for panic and phobic disorders. Cell 2001; 106:367–379 8. Tabiner M, Youings S, Dennis N, et al: Failure to find DUP25 in patients with anxiety disorders, in control individuals, or in previously reported positive control cell lines. Am J Hum Genet 2003; 72:535–538 9. Weiland Y, Kraus J, Speicher MR: A multicolor fish assay does not detect DUP25 in control individuals or in reported positive control cells. Am J Hum Genet 2003; 72:1349 –1352 10. Gulpek D, Bayraktar E, Akbay SP, et al: Joint hypermobility syndrome and mitral valve prolapse in panic disorder. Prog Neuropsychopharmacol Biol Psychiatry 2004; 28:969 –973 11. Benjamin J, Ben-Zion IZ, Dannon P, et al: A: lack of association between joint hyperlaxity and I: panic disorder, and II: reactivity to carbon dioxide in healthy volunteers. Hum Psychopharmacol 2001; 16:189 –192 ¨ , et al: Hipermobilite sendromu olan 12. Gu¨lsu¨n M, Doruk A, Uzun O ¨ rnekleminde anksiyete bozukluklari. Yeni klinik disi erkek O Symposium J 2006; 44:165–168 13. Acasuso-Dı´az M, Collantes-Estevez E: Joint hypermobility in patients with fibromyalgia. Arthritis Care Res 1998; 11:39 – 42 14. Wolfe F, Smythe HA, Yunus MB, et al: The American College of Rheumatology 1990 Criteria for the Classification of Fibromyalgia. Report of the Multicenter Criteria Committee. Arthritis Rheum 1990; 33:160 –172 15. Schlesslman JJ, Stolley PD: Case–Control Studies: Design, Conduct, Analysis. New York, Oxford University Press, 1982 16. Lobo A, Campos R, Pe´rez-Echeverrı´a MJ, et al: A new interview for the multiaxial assessment of psychiatric morbidity in medical settings. Psychol Med 1993; 23:505–510 17. Lobo A, Garcı´a-Campayo J, Campos R, et al: Somatisation in
Psychosomatics 51:1, January-February 2010
primary care in Spain, I: estimates of prevalence and clinical characteristics. Br J Psychiatry 1996; 168:344 –348 18. Garcı´a-Campayo J, Campos R, Marcos G, et al: Somatisation in primary care in Spain, II: differences between somatisers and psychologisers. Br J Psychiatry 1996; 168:348 –353 19. Garcı´a-Campayo J, Sanz-Carrillo C, Pe´rez-Echeverrı´a MJ, et al: Screening of somatization disorder: validation of the Spanish version of the Othmer and De Souza test. Acta Psychiatr Scand 1996; 94:411– 415 20. Garcı´a-Campayo J, Larrubia J, Lobo A, et al: Attribution in somatisers: stability and relationship to outcome at 1-year follow-up. Acta Psychiatr Scand 1997; 95:433– 438 21. Garcı´a-Campayo J, Lobo A, Pe´rez-Echeverrı´a MJ, et al: Three forms of somatization presenting in primary care settings in Spain. J Nerv Ment Dis 1998; 186:554 –560 22. Bandelow B: Panic and Agoraphobia Scale (Spanish version): Test Manual. Toronto, Canada, Hogrefe & Huber, 1999 23. Garcı´a Campayo J: Avances en Psicometrı´a. Madrid, Spain, IM&C, 2006 24. Spielberger CD, Gorsuch RL, Lushene RE: STAI, Cuestionario de Ansiedad Estado/Rasgo. TEA, 2001 25. Zigmond AS, Snaith RP: The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand 1983; 67:361–370 26. Tejero A, Guimera EM, Farre JM, et al: Uso clı´nico del HAD (Hospital Anxiety and Depression Scale) en poblacion psiquiatrica: un estudio de su sensibilidad, fiabilidad, y validez. Rev Dep Psiquiatr Fac Med Barc 1986; 13:233–238 27. American Psychiatric Association: Handbook of Psychiatric Measures. Washington, DC, American Psychiatric Association, 2000 28. Bulbena A, Duro´ JC, Porta M, et al: Clinical assessment of hypermobility of joints: assembling criteria. J Rheumatol 1992; 19:115–122 29. Armitage P, Berry G: Statistical Methods in Medical Research, 3rd Edition. Oxford, UK, Blackwell Scientific, 1994 30. Siegel S, Castellan NJ Jr: Nonparametric Statistics for the Behavioral Sciences, 2nd Edition. New York, McGraw-Hill, 1988 31. Neumann L, Buskila D: Epidemiology of fibromyalgia. Curr Pain Headache Rep 2003; 7:362–368 32. Beighton PH, Solomon L, Soskolne CL: Articular mobility in an African population. Ann Rheum Dis 1973; 32:413– 418 33. Child A: Joint hypermobility syndrome: inherited disorder of collagen synthesis. J Rheumatol 1986; 13:239 –243 34. Tamura K, Fakuda Y, Yamanka N, et al: Structural abnormalities of connective tissue in floppy mitral valve (abstract). J Am Coll Cardiol 1994; 23:32A
http://psy.psychiatryonline.org
61
Background: A significant association between joint hypermobility syndrome (JHS) and panic disorder was observed in a sample of rheumatology outpatients. Objective: The aim of this study was to assess whether JHS is more frequent in panic-disorder than in control subjects. Method: The authors conducted a case– control study comparing 55 untreated patients with panic disorder and three matched-control groups: psychiatric patients, fibromyalgia patients, and healthy persons. Results: JHS was more frequent among panic-disorder than among psychiatric patients, the healthy group, or the fibromyalgia group. In the panic-disorder group, there was a significant correlation between severity of JHS and anxiety. Conclusion: The strong association between JHS and panic disorder points to a genetic association. There is also a possibility that JHS and mitral valve prolapse, another condition frequently associated with panic disorder, share a common pathophysiological mechanism. (Psychosomatics 2010; 51:55– 61)
J
oint hypermobility syndrome (JHS) is a highly hereditary clinical condition characterized by increased distensibility of joints in passive movements and hypermobility in active movements.1 Its prevalence in the general population appears to range between 10% and 15%, and it is more common in women than in men.1 Although JHS often goes unnoticed, affected individuals may suffer from repeated injuries of the musculoskeletal system,2 so its early diagnosis is recommended to avoid these complications. A significant association between JHS and panic disorder was observed in a sample of rheumatology outpatients.3,4 There is also a report on the high association of panic disorder and/or agoraphobia with JHS (the odds ratio [OR] compared with psychiatric control subjects was 18.6, and the OR compared with medical control subjects was 14.7).5 Also, another study relates JHS to both state and trait anxiety (the former with less intensity) in the general population.6 The magnitude of the JHS and panic disorder association, uncommon in medical research, points to a causal connection between these two conditions. Grataco´s et al.7 Psychosomatics 51:1, January-February 2010
identified an interstitial duplication of the human chromosome 15q24 –26 (named DUP25), which was significantly associated with panic/agoraphobia/social phobia/joint laxity in families, and with panic disorder in nonfamilial cases. Mosaicism, different forms of DUP25 within the same family, and absence of segregation of 15q24 –26 markers with DUP25, and the associated psychiatric phenotypes, suggest a non-Mendelian mechanism of diseasecausing mutation. These authors7 proposed that DUP25, which is present in 7% of the control subjects, is a susceptibility factor for a clinical phenotype that includes Received August 27, 2007; revised October 31, 2007; accepted November 26, 2007. From the Dept. of Psychiatry, Miguel Servet University Hospital and University of Zaragoza, Spain; Casablanca Health Centre, Zaragoza, Spain; Statistical Methods Department, Faculty of Mathematics, Zaragoza, Spain; Grupo Aragone´s de Investigacio´n en Atencio´n Primaria; Red de Actividades Preventivas y Prevencio´n de la Salud (REDIAPP G03/170); and the Instituto Aragone´s de Ciencias de la Salud. Send correspondence and reprint requests to Javier Garcia Campayo, Dept. of Psychiatry, Miguel Servet University Hospital, Avda Isabel La Cato´lica 1, 50.009 Zaragoza, Spain. e-mail: [email protected] © 2010 The Academy of Psychosomatic Medicine
http://psy.psychiatryonline.org
55
Joint Hypermobility and Panic Disorder panic and phobic disorders and joint laxity. However, this genetic finding was not confirmed by two other research groups, who achieved different results.8,9 To date, most of the research papers published on the association between JHS and panic disorders/agoraphobia3–7 have been developed by the same team of researchers. To our knowledge, three replication studies have been carried out, with inconclusive results: two10,11 failed to confirm this association; the first, probably owing to the unbelievably high JHS percentage (52.6%) found in the control group; and the second, probably resulting from methodological limitations,11 such as the kind of instruments used. The third study, the only one to confirm the association between panic disorder and JHS,12 has only been published in Turkish. The present study was designed to replicate the research of the Barcelona group and, more specifically, to answer the following questions: 1) Is JHS more common in patients with panic disorder than among appropriate comparison subjects (psychiatric controls, medical controls, and healthy people)? and 2) Does severity of panic/anxiety symptoms correlate with JHS severity in these groups? METHOD A Case–Control Study Setting The study was carried out in the Torrero, Casablanca, and Arrabal primary-care health centers, in the city of Zaragoza, Spain. Patients Four groups of patients were included: a Case group of patients diagnosed with panic disorder, and three other Control groups (psychiatric, medical, and healthy) that were selected on the basis of the two previous papers published on the subject5,6 and matched one-to-one by sex, age (plus-or-minus 2 years), and ethnic group, to the group of panic-disorder patients. The following is a description of each sample group: Patients diagnosed with panic disorder: All consecutive, newly diagnosed, and untreated patients attending the aforementioned primary-care health centers over an 18month period (from January 2005 to June 2006) who fulfilled inclusion criteria were included. We applied the following criteria: 1) DSM–IV criteria for panic disorder (with or without agoraphobia); 2) age 18 – 65 years; 3) ability to understand and read Spanish. Exclusion criteria: 1) Any other DSM–IV Axis I psychiatric disorder, such as psychosis, drug abuse, cognitive disorder; 2) severe medical disorders that hamper physical (including joint) examination. The psychiatric control group: These patients were 56
http://psy.psychiatryonline.org
recruited from the same primary-care health centers and included consecutive, newly diagnosed psychiatric patients who did not suffer from DSM–IV panic disorder and who had never met criteria for any anxiety disorder. This selection criterion was also used by the original authors.5 Inclusion criteria: 1) any DSM–IV psychiatric diagnosis except those described in the exclusion section; 2) age: 18 – 65 years; and 3) untreated patients, or patients able to discontinue the treatment at least 2 weeks before the JHS assessment. Exclusion criteria: 1) lifetime diagnosis of any anxiety disorder; 2) any psychiatric diagnosis that, from the point of view of the clinician, would not allow the study to be completed (e.g., severe psychosis or personality disorder, cognitive deterioration, drug abuse); and 3) severe medical disorders that hamper physical (including joint) examination. Many of these patients were getting pharmacological treatment and were not able to discontinue them before the assessment. The medical control (fibromyalgia) group: The original authors5 recruited this second control group from four medical outpatient clinics (dermatology, gastroenterology, general surgery, and internal medicine) at the same hospital. We preferred to include in this group only patients with a specific rheumatologic disorder, fibromyalgia, which has been reported to show a high association with JHS. In the Spanish population, this association has been calculated to be 27.3%.13 All consecutive, newly diagnosed patients with fibromyalgia attending the aforementioned primary-care health centers were invited to participate in the study. Inclusion criteria: 1) diagnosis of primary fibromyalgia by a rheumatologist according to the American College of Rheumatology criteria;14 and 2) age: 18 – 65. Exclusion criteria: 1) any psychiatric diagnosis that, from the point of view of the clinician, would not allow the study to be completed (e.g., severe psychosis or personality disorder, cognitive deterioration, or drug abuse); and 2) severe medical disorders that hamper physical (including joint) examination. Every effort was made to recruit only untreated patients; however, most of these patients were taking various kinds of analgesics and/or antidepressants. They were asked to discontinue the treatment at least 2 weeks before the JHS assessment. The healthy-control group: These were persons accompanying patients with panic disorder, or other psychiatric disorders, who were visiting the previously mentioned primary-care health centers and who, according to the Standardized Polyvalent Psychiatric Interview (the SPPI), did not suffer from any psychiatric disorders. A sample group of this type was also assessed by the original authors6 to evaluate JHS’s association with anxiety in Psychosomatics 51:1, January-February 2010
Campayo et al. nonclinical populations. Inclusion criteria: 1) no DSM–IV psychiatric diagnosis; and 2) age: 18 – 65. Exclusion criteria: 1) severe medical disorders that would hamper physical (including joint) examination. Sample: Sample size (N⫽55 for each of the four groups) was calculated by standard methods for ␣ at 5% (two-tailed),  at 80%, and a difference of odds ratios (OR) between the panic-disorder group and control group of ⬎3.15 In total, a sample of 220 subjects was recruited. All control subjects were matched to the panic-disorder group patients by sex, age (within 2 years), and ethnic group. Nonetheless, given the predominance of women among fibromyalgia patients, it was impossible to match this group by gender. To obtain the expected sample, a total of 58 panic-disorder patients were interviewed, with 3 refusing to participate in the study. For the three control groups, it was necessary to interview 60 psychiatric patient (5 refused), 56 fibromyalgia patients (only 1 refused), and 59 control subjects (4 refused). To replace participants who had refused during recruitment, the first subject attending the health center who fulfilled the inclusion criteria with the same sex, age, and ethnicity of the individual who had refused was invited to participate.
original research, is more specific for the assessment of panic and agoraphobia symptoms. State–Trait Anxiety Scale (STAI), Spanish version24 We used this scale for the assessment of anxiety in healthy subjects, in order to replicate the previous study of the group on nonclinical individuals in which this questionnaire was used. It seems more appropriate than the PAS for this specific population, and it yields information about state- and trait-anxiety. This questionnaire was also used with psychiatric patients because panic disorder had been specifically excluded from this group. Hospital Anxiety and Depression Scale (HADS), Anxiety Subscale The HADS25 is a self-report scale that screens for the presence of depression and anxiety in patients with organic disorders. It comprises 14 items that are rated on a 4-point Likert-type scale, and it is appropriate for use in community and hospital settings. Two subscales, HADS– Dep and HADS–Anx, independently assess depression and anxiety. The HADS was validated in a Spanish population.26 HADS was selected for use in the present study because it is considered one of the best questionnaires for assessing depression and anxiety in patients with physical illness.27
Instruments Standardized Polyvalent Psychiatric Interview (SPPI) The SPPI is a psychiatric interview developed by our group for the multiaxial assessment of psychiatric morbidity in outpatient psychiatric clinics and primary-care settings.16 It permits the use of various diagnostic criteria, including DSM–IV and ICD–10. Our group has wide research experience with this interview.17–21 Panic and Agoraphobia Scale (PAS)22 The PAS is the first-ever scale for assessing the severity of panic disorder with or without agoraphobia. Compatible with both DSM–IV and ICD–10 classifications, and available in both self-rated and observer-rated versions, the PAS was specially developed for monitoring the efficacy of both drug and psychotherapy treatments. It has been translated and validated in Spanish,23 has excellent psychometric properties,22,23 and is quick to use: the observer-rated version that we used can be completed in 5 to 10 minutes. We used this scale for the assessment of panic symptoms in the panic-disorder group. This questionnaire is considered more suitable than the scale used by the original authors,5 the Hamilton Rating Scale for Anxiety, because the PAS, which was not available for the authors at the time of the Psychosomatics 51:1, January-February 2010
Assessment of Joint Hypermobility Syndrome (JHS) To assess joint hypermobility syndrome (JHS), we used Beighton’s criteria.1 The scores range from 0 to 9 and depend on the degree of movement. Subjects who scored ⱖ5 on the 9-point scale were considered to have JHS. The reliability and validity of the assessment criteria for JHS have been previously described.28 Procedure Patients were recruited from our research department by two researchers (NS, EMA). Two experienced psychiatrists (JGC, MA) administered the SPPI psychiatric interview to establish the DSM–IV psychiatric disorder and rated the psychological questionnaires used in this study (the PAS, STAI, and HADS). In a later encounter, a family doctor (EA), unaware of either psychiatric diagnosis or anxiety scores, assessed JHS by use of Beighton’s criteria. Patients were informed at the beginning of the study that they were not to give any unsolicited information, so as to maintain the blinding. After the interviews with patients and examiners (JGC, MA, EA), the two research assistants (NS, EMA) confirmed that blinding had not been broken. In http://psy.psychiatryonline.org
57
Joint Hypermobility and Panic Disorder
TABLE 1.
Sociodemographic and Clinical Characteristics of the Sample, Mean (SD)
Sex, female, N (%) Age, years Ethnicity, white European, N (%) PAS HADS STAI-trait STAI-state Total N: JHS Beighton’s criteria
Panic Disorder (Nⴝ55)
Psychiatric Controls (Nⴝ55)
Fibromyalgia (Nⴝ55)
Healthy Controls (Nⴝ55)
Difference, p
45 (81.8%) 38.1 (10.5) 55 (100%) 22.5 (7.8) 6.5 (3.0) 20.5 (9.8) 21.7 (8.8) 5.4 (2.3)
45 (81.8%) 39.4 (10.8) 55 (100%) 6.7 (2.3) 5.8 (2.8) 19.4 (9.6) 20.5 (9.8) 2.4 (2.3)
47 (85.4%) 40.1 (10.9) 55 (100%) 7.8 (2.6) 6.2 (3.1) 17.8 (8.2) 18.6 (7.9) 2.7 (2.6)
45 (81.8%) 38.2 (10.4) 55 (100%) 5.6 (1.7) 2.3 (0.8) 11.9 (6.4) 12.1 (7.5) 2.4 (2.3)
NSa NSb NSa ⬍0.001c ⬍0.05c ⬍0.01c ⬍0.01c ⬍0.001c
Values are mean (standard deviation [SD]), unless otherwise indicated. PAS: Panic and Agoraphobia Scale; HADS: Hospital Anxiety and Depression Scale. a chi-square. b **Student t-test, paired samples. c ***Mann-Whitney U test.
order to adequately replicate previous studies, the family doctor was standardized with the members of the Barcelona, Spain, team, obtaining an interrater reliability (kappa) of 0.85– 0.90 for the different items of the Beighton’s criteria. The research was approved by the Ethical Review Board of the regional health authority, and informed consent from the patients was obtained after the procedures had been fully explained. Statistical Analysis In contingency tables, Pearson’s chi-square test for homogeneity or independence was applied to assess the relationship between two qualitative or categorical variables. In 2⫻2 chi-squares, the Yates-corrected chi-square value is given. Where 20% or more of the cells had expected counts of ⬍5, Fisher’s exact test was used.29 Student’s t-test and one-factor analysis of variance were used to determine the relationship between a categorical variable with two or three levels, respectively, and a normally distributed quantitative variable; the Mann-Whitney U test was used to analyze quantitative variables that did not have normal distribution.30 The normality of the variables was contrasted with a Kolmogorov-Smirnov test. Spearman rho () was used to determine correlation between different variables because distributions were not normal. The logit estimator of the prevalence odds ratio (OR) was estimated with precision-based confidence intervals (CI).29 Data management and statistical analysis were carried out using the SPSS 14 Statistical Package. RESULTS The mean age of the 220 subjects of the study was 41.3 years (standard deviation [SD]: 11.0); 82.7% (182/220) 58
http://psy.psychiatryonline.org
were women. As shown in Table 1, Case and Control groups did not significantly differ by age, sex, or ethnicity because of the matching process, despite a slight predominance of women in our fibromyalgia group. It was impossible for us to recruit a matching number of men with fibromyalgia (female predominance in this disorder has been calculated at 7/1 to 9/1).31 All of the patients were white Europeans. Of the anxiety cases, 70.9% (39/55) were diagnosed as having panic disorder with agoraphobia, and the remaining 29.1% (16/55) of panic disorder without agoraphobia. The distribution of case severity according to SPPI was as follows: 50% (28/55) light, 30% (17/55) moderate, and 18.1% (10/55) severe; 20 of 55 (36.3%) were also diagnosed with major depression or dysthymia. In the psychiatric control group, diagnostic distribution was as follows: major depressive disorder (36/ 55; 65.4%), dysthymia (9/55; 16.3%), schizophrenia (8/ 55; 14.5%), and somatoform disorders (6/55; 10.9%), with some of the patients having more than one psychiatric diagnosis. Fibromyalgia patients were also diagnosed with depression in 32.7% of cases (18/55), and 4 of these (7.2%) were diagnosed with panic disorder. Panic and Joint Hypermobility Syndrome (JHS) Table 1 summarizes the mean PAS score (range: 0 –54), the mean STAI-trait score (range: 0 – 60), the mean STAI-state score (range: 0 – 60), and the mean HADS-Anx score (range: 0 –21) for all the groups. It also shows the mean Beighton’s criteria for JHS scores (range: 0 –9) for the four different groups. As shown, there are significant differences between Beighton’s criteria scores for the panic-disorder group, as compared with healthy-control subPsychosomatics 51:1, January-February 2010
Psychosomatics 51:1, January-February 2010
2.0–14.1 5.4 — ⬍0.001 0.215 11.2 1.5 12/40 (30) 2/15 (13.3) 6.1–72.1 21 — ⬍0.001 0.085 27.5 2.9 4/40 (10) 1/15 (6.6) 1.0 1.0 28/40 (70) 6/15 (40)
N⫽55 for all groups. OR: odds ratio; CI: confidence interval. a Reference group. b With Yates correction (df: 1).
5.1–51.8 ⬍0.001 16.3 0.085 — 24.9 2.9
1.0 1.0 4/10 (40) 30/45 (66.6)
5/40 (12.5) 1/15 (6.6)
2.1–12.7 — 5.2 0.444 ⬍0.001 0.5 12.5 1/8 (12.5) 13/47 (27.6) 6.1–68.0 — 20.5 0.302 ⬍0.001 1.0 29.5 1/10 (10) 4/45 (8.8) 5.2–48.9 0.302 — ⬍0.001 16 1.0 26.9
4.8–36.2 ⬍0.001 13.2 28.6 1.0 34 (61.8)
Total Gender Men Women Age ⬍45 years ⱖ45 years
6 (10.9)
N (%) OR N (%)
1/10 (10) 5/45 (11.1)
95% CI
2.0–10.7 4.7
OR p
⬍0.001 13.3 14 (25.4) 5.5–47.1 16.1 ⬍0.001 31.1
N (%) 95% CI p
OR
5 (9.0)
2 N (%) p
OR
95% CI
Fibromyalgia Control Group Psychiatric Control Group
2 2b
Healthy-Control Group Panic-Disorder Groupa
Group
TABLE 2.
jects (Mann-Whitney U⫽622.5; Z ⫽ –5.3; p ⬍0.001), psychiatric group (Mann-Whitney U⫽391; Z ⫽ – 6.7; p ⬍0.001), and patients with fibromyalgia (Mann-Whitney U⫽631.5; Z ⫽ –5.3; p ⬍0.001). As is summarized in Table 2, JHS according to Beighton’s criteria was significantly more frequent among patients with panic disorder (34/55; 61.8%) than among psychiatric-control patients (6/55; 10.9%); Yates-corrected 2: 2.6; df⫽1; p ⬍0.001), with an OR of 13.2, and the 95% CI: 4.8 –36.2. JHS is also much more prevalent in the panic-disorder group than in the healthy-control group (5/55; 9.0%); Yates-corrected 2: 33.8; df⫽1; p ⬍0.001), with OR 20.6, and 95% CI: 6.5– 65.4. It is also more frequent in the panic-disorder group than in the medicalcontrol group with fibromyalgia (14/55; 25.4%); Yatescorrected 2: 13.3; df⫽1; p ⬍0.001; OR: 4.7 (95% CI: 2.0 –10.7). Stratified analysis by sex and age (cutoff point: 45 years) showed highly significant differences in women and in the sample under age 45, but failed to find these group differences for persons age 45⫹ (p⫽0.08) and for men. There was a significant correlation, in patients with panic disorder, between the number of Beighton’s criteria and PAS (Spearman ⫽0.520; p ⬍0.001), but not with STAI-state (Spearman ⫽0.234; p⫽0.411), STAI-trait (Spearman ⫽0.298; p⫽0.376) or HADS (Spearman ⫽0.098; p⫽0.706). For the other groups of patients, there was no correlation between Beighton’s criteria and any of the different questionnaires on anxiety (PAS, HADS, STAI-state, or STAI-trait). Among patients with panic disorder, we compared the groups with and without JHS. We found that the subgroup with JHS was younger (mean age: 35.2 years; SD: 10.7) than patients with panic disorder but without JHS (mean age: 41.7 years; SD: 9.8; t ⫽ –2.2; df: 53; p ⬍0.05). PAS scores were also higher in patients with both panic disorder and JHS (mean: 19.8; SD: 4.3) than in the panicdisorder group without JHS (mean: 13.7; SD: 6.0; MannWhitney U⫽137; Z ⫽ –3.8; p ⬍0.001). There was no difference in the proportion of women in the hypermobility group (30/34; 88%), as compared with the non-hypermobility subjects (15/21; 71.4%; Yates-corrected 2: 1.4; df: 1; p⫽0.226). For panic symptoms measured with the SPPI interview, the only statistical differences between Case and Control groups were on the items “trembling or shaking” (30/34; 88.2% versus 10/21; 47.6%; OR: 8.2; 95% CI: 2.1–31.8) and “chills or hot flashes” (31/34; 91.1% versus 12/21; 57.1%; OR: 7.7; 95% CI: 1.7–33.5). However, total
Prevalence of Joint Hypermobility Syndrome (JHS) in Primary-Care Patients With Panic Disorder, Healthy-Control Subjects, Psychiatric Control Subjects, and Fibromyalgia
Campayo et al.
http://psy.psychiatryonline.org
59
Joint Hypermobility and Panic Disorder number of panic symptoms, severity of panic symptoms measured with SPPI, and age at onset of the disorder were not significantly different between groups. There was also no difference in agoraphobia scores. DISCUSSION This is the first study, to our knowledge, that replicates the data of Bulbena et al.3–7 We have used the same design and have standardized the JHS diagnosis according to this group, obtaining adequate kappa values. Some of the strengths of our study are the following: First, sample recruitment was carried out in primary-care settings, thus avoiding selection bias of specialized care. Second, we used a psychiatric interview to confirm all the psychiatric diagnoses (or absence of psychiatric disorder). And, third, we utilized a more sophisticated assessment of panic severity symptoms, the PAS questionnaire. In our study, the prevalence of JHS according to Beighton’s criteria in subjects with panic disorder was 61.8%, a figure quite similar to the 67.7% described by Bulbena’s group,5 and significantly higher than the 10.9% observed in our healthy-control subjects (OR⫽13.2; 95% CI: 4.8 –36.2), and the 9% observed in our psychiatriccontrol group (OR: 16.1; 95% CI: 5.5– 47.1). The figures found in the healthy-control group and the psychiatric group are similar to those described by Bulbena et al.5 and similar to the 10%–15% described for the general population.33 It is remarkable that JHS is even more frequent in panic disorder (OR: 4.7; 95% CI: 2.0 –10.7) than among the medical control group with fibromyalgia (25.4%), a disorder that has also been associated with JHS.13 The figures we have found in our sample are similar to the 27.3% detected in other studies in Spanish patients with fibromyalgia.13 When we carried out a stratified analysis of JHS prevalence by sex and age, we found highly significant differences for women and the younger age-group, but failed to find these differences for persons age 45⫹ and for men. The reason for this finding could be the limitation resulting from the small sample size in each of the four groups. The power calculation for this stratification analysis was 0.72, slightly lower than the 0.80 usually used. These data are similar to those found by Bulbena et al.,5 who described the highest odds ratio for JHS (more than double) in women and younger people in both the panic-disorder and psychiatric-control groups. We have also found that, in patients with panic disorder, there is a significant correlation between the number of Beighton’s criteria and PAS scoring, but not with the 60
http://psy.psychiatryonline.org
other tests of anxiety used (STAI-trait, STAI-state, HADS). As far as we are aware, this is the first time that this correlation has been described in scientific literature, and it reinforces a possible genetic association between panic disorder and JHS, as has been suggested.7 However, we did not find any correlation between the number of Beighton’s criteria and STAI-state, STAI-trait, or HADS scoring in healthy individuals, psychiatric patients, or patients with fibromyalgia. This surely implies that JHS only correlates with symptoms of panic disorder (measured with the PAS), but not with nonspecific anxiety symptoms measured with the other questionnaires. The sociodemographic and clinical characteristics of the JHS panic-disorder subgroup are somewhat nonspecific. Apart from a predominance of younger people, higher scoring in PAS, and certain clinical symptoms, these patients are strikingly similar to the remaining patients with panic. The implications of these findings are the following: 1) From a genetic perspective, the strong association described between JHS and panic disorder points to a genetic association. The original authors7 proposed DUP25 as the possible genetic factor related to both disorders. Despite the fact that other, later studies8,9 failed to confirm this genetic finding, it seems that there is a need to continue to search for the genetic basis of this association. 2) From a neurobiological perspective, this finding also reinforces the finding that JHS and mitral valve prolapse, another disorder frequently associated with panic disorder, share a common pathophysiological mechanism that some authors have thought could be a collagen disorder.33,34 Perhaps a possible brain fear-circuitry could exist that may be common to these different conditions. 3) From a clinical point of view, the frequent underdiagnosis of both disorders make it advisable to screen for JHS whenever a doctor diagnoses anxiety, and vice versa. Also, based on their family aggregation, it would be also recommended to screen close relatives so as to rule out both disorders. In summary, we confirmed the strong association of JHS and panic disorder and found that there was no association with healthy subjects, psychiatric controls, or medical controls. There is also a correlation between the severity of panic symptoms and the severity of JHS in patients with panic disorder. The etiology of the association of these two syndromes is still unclear. However, JHS seems to be a constitutional characteristic of many patients with panic disorder.
The authors report no financial conflicts of interest. Psychosomatics 51:1, January-February 2010
Campayo et al.
References 1. Beighton PH, Grahame R, Bird H: Hypermobility of Joints, 2nd Edition. London, UK, Springer-Verlag, 1989 2. Lewkonia RM: The biology and clinical consequences of articular hypermobility. J Rheumatol 1993; 20:220 –222 3. Bulbena A, Duro JC, Porta M, et al: Anxiety disorders in the joint hypermobility syndrome. Lancet 1988; 2:694 4. Bulbena A, Duro´ JC, Porta M, et al: Anxiety disorder in the joint hypermobility syndrome. Psychiatry Res 1993; 43:59 – 68 5. Martı´n-Santos R, Bulbena A, Porta M, et al: Association between joint hypermobility syndrome and panic disorder. Am J Psychiatry 1998; 155:1578 –1583 6. Bulbena A, Agullo´ A, Pailhez G, et al: Is joint hypermobility related to anxiety in a non-clinical population also? Psychosomatics 2004; 45:432– 437 7. Gratacos M, Nadal M, Martin-Santos R, et al: A polymorphic genomic duplication on human chromosome 15 is a susceptibility factor for panic and phobic disorders. Cell 2001; 106:367–379 8. Tabiner M, Youings S, Dennis N, et al: Failure to find DUP25 in patients with anxiety disorders, in control individuals, or in previously reported positive control cell lines. Am J Hum Genet 2003; 72:535–538 9. Weiland Y, Kraus J, Speicher MR: A multicolor fish assay does not detect DUP25 in control individuals or in reported positive control cells. Am J Hum Genet 2003; 72:1349 –1352 10. Gulpek D, Bayraktar E, Akbay SP, et al: Joint hypermobility syndrome and mitral valve prolapse in panic disorder. Prog Neuropsychopharmacol Biol Psychiatry 2004; 28:969 –973 11. Benjamin J, Ben-Zion IZ, Dannon P, et al: A: lack of association between joint hyperlaxity and I: panic disorder, and II: reactivity to carbon dioxide in healthy volunteers. Hum Psychopharmacol 2001; 16:189 –192 ¨ , et al: Hipermobilite sendromu olan 12. Gu¨lsu¨n M, Doruk A, Uzun O ¨ rnekleminde anksiyete bozukluklari. Yeni klinik disi erkek O Symposium J 2006; 44:165–168 13. Acasuso-Dı´az M, Collantes-Estevez E: Joint hypermobility in patients with fibromyalgia. Arthritis Care Res 1998; 11:39 – 42 14. Wolfe F, Smythe HA, Yunus MB, et al: The American College of Rheumatology 1990 Criteria for the Classification of Fibromyalgia. Report of the Multicenter Criteria Committee. Arthritis Rheum 1990; 33:160 –172 15. Schlesslman JJ, Stolley PD: Case–Control Studies: Design, Conduct, Analysis. New York, Oxford University Press, 1982 16. Lobo A, Campos R, Pe´rez-Echeverrı´a MJ, et al: A new interview for the multiaxial assessment of psychiatric morbidity in medical settings. Psychol Med 1993; 23:505–510 17. Lobo A, Garcı´a-Campayo J, Campos R, et al: Somatisation in
Psychosomatics 51:1, January-February 2010
primary care in Spain, I: estimates of prevalence and clinical characteristics. Br J Psychiatry 1996; 168:344 –348 18. Garcı´a-Campayo J, Campos R, Marcos G, et al: Somatisation in primary care in Spain, II: differences between somatisers and psychologisers. Br J Psychiatry 1996; 168:348 –353 19. Garcı´a-Campayo J, Sanz-Carrillo C, Pe´rez-Echeverrı´a MJ, et al: Screening of somatization disorder: validation of the Spanish version of the Othmer and De Souza test. Acta Psychiatr Scand 1996; 94:411– 415 20. Garcı´a-Campayo J, Larrubia J, Lobo A, et al: Attribution in somatisers: stability and relationship to outcome at 1-year follow-up. Acta Psychiatr Scand 1997; 95:433– 438 21. Garcı´a-Campayo J, Lobo A, Pe´rez-Echeverrı´a MJ, et al: Three forms of somatization presenting in primary care settings in Spain. J Nerv Ment Dis 1998; 186:554 –560 22. Bandelow B: Panic and Agoraphobia Scale (Spanish version): Test Manual. Toronto, Canada, Hogrefe & Huber, 1999 23. Garcı´a Campayo J: Avances en Psicometrı´a. Madrid, Spain, IM&C, 2006 24. Spielberger CD, Gorsuch RL, Lushene RE: STAI, Cuestionario de Ansiedad Estado/Rasgo. TEA, 2001 25. Zigmond AS, Snaith RP: The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand 1983; 67:361–370 26. Tejero A, Guimera EM, Farre JM, et al: Uso clı´nico del HAD (Hospital Anxiety and Depression Scale) en poblacion psiquiatrica: un estudio de su sensibilidad, fiabilidad, y validez. Rev Dep Psiquiatr Fac Med Barc 1986; 13:233–238 27. American Psychiatric Association: Handbook of Psychiatric Measures. Washington, DC, American Psychiatric Association, 2000 28. Bulbena A, Duro´ JC, Porta M, et al: Clinical assessment of hypermobility of joints: assembling criteria. J Rheumatol 1992; 19:115–122 29. Armitage P, Berry G: Statistical Methods in Medical Research, 3rd Edition. Oxford, UK, Blackwell Scientific, 1994 30. Siegel S, Castellan NJ Jr: Nonparametric Statistics for the Behavioral Sciences, 2nd Edition. New York, McGraw-Hill, 1988 31. Neumann L, Buskila D: Epidemiology of fibromyalgia. Curr Pain Headache Rep 2003; 7:362–368 32. Beighton PH, Solomon L, Soskolne CL: Articular mobility in an African population. Ann Rheum Dis 1973; 32:413– 418 33. Child A: Joint hypermobility syndrome: inherited disorder of collagen synthesis. J Rheumatol 1986; 13:239 –243 34. Tamura K, Fakuda Y, Yamanka N, et al: Structural abnormalities of connective tissue in floppy mitral valve (abstract). J Am Coll Cardiol 1994; 23:32A
http://psy.psychiatryonline.org
61