Characterization of Migraine in Children and Adolescents With Generalized Joint Hypermobility

Abstract

Background and Objectives

The prevalence of generalized joint hypermobility (GJH) is 5–65% in children and adolescents. The hypothesis of this study was to see whether there is an association between headache characteristics and GJH in children and adolescents with migraine.

Methods

We performed a primary retrospective case-control analysis of an established database of patients with headache aged 5–17 years.

Results

We included 5435 participants. Approximately 31.6% of participants (1,719/5,435) were diagnosed with GJH (Beighton score ≥ 6). Nausea (73.1% vs 67.5%, χ² with 1 degree of freedom = 17.0, p < 0.0001), phonophobia (87.3% vs 78.8%, χ² with 1 degree of freedom = 18.0, p < 0.0001), and the PedMIDAS score (48.2 ± 52.5, 95% CI 45.7–50.6 vs 41.6 ± 51.2, 95% CI 40.0–43.3, effect size = 0.13, p < 0.0001) were noted to be more severe in participants with GJH than those without GJH.

Discussion

Youths with GJH and migraine were noted to have more severe migraine characteristics.

Introduction

Hypermobile Ehlers-Danlos syndrome (hEDS) and hypermobility spectrum disorder (HSD) are heterogeneous groups of connective tissue disorders. Previously, hEDS and HSD were clinically indistinguishable from the joint hypermobility syndrome.¹ In 2017, revised criteria were introduced for hEDS, and patients not meeting the criteria for hEDS but with joint hypermobility symptoms were diagnosed as those with HSD.² With HSD, prepubertal children and adolescents with the Beighton score of ≥6 (Table 1) without any musculoskeletal involvement were diagnosed with asymptomatic generalized joint hypermobility (GJH).² The Beighton score is a valid measure for GJH in children,³ based on the analysis of the major joints.

Table 1.

The 9-Point Beighton Hypermobility Score Based on Information From Beighton et al.¹

The range of headache disorders seen in adults with GJH can range from migraine,³ the most prevalent headache disorder, to cervicogenic headaches,⁴ a rare form. Past studies indicate migraine is prevalent in 75% of adult women with GJH.⁵ In addition, the frequency of headaches and a disability associated with migraine is significantly higher in these patients relative to those without GJH.⁴ Although these symptoms could be present in children and adolescents with headache disorders, the evaluation and diagnosis are hindered by a lack of familiarity with pediatric GJH-related symptomatology and utilization of the Beighton scale in clinical practice.

The primary purpose of this study was to characterize the relationship between headache disorders, specifically migraine subtypes, and GJH in children and adolescents. The primary objective was to determine the prevalence of GJH in patients with migraine and compare the frequency and severity of migraine in participants with GJH with that in children and adolescents with migraine but no evidence of GJH. The secondary objective was to determine the prevalence of muscular tenderness, temporomandibular joint disorders (TMD), and comorbid psychiatric conditions in youth with GJH compared with that in those without. We hypothesized that the participants experiencing migraine would have a higher prevalence of GJH.

Methods

Design and Settings

This was a primary retrospective cohort analysis of an established electronic database that serves as a clinical and research headache registry⁵ of patients seen in a tertiary care headache program. Those who presented for an initial headache clinic evaluation between 2015 and2020 were included. As approved by the institutional review board, participants or their guardians provided written informed consent to have their deidentified clinical data recorded into the registry, including characteristics and treatment of their headaches. We used the STROBE case-control reporting guidelines for this study.⁶ We did not find any missing data and information in this retrospective study.

Standard Protocol Approvals, Registrations, and Patient Consents

This study uses the clinical and research repository. This repository was approved by the Institutional Review Board at Cincinnati Children's Hospital Medical Center; patients' families provided informed consent to have patient information stored in the Headache Center data repository.

Participants

Participants aged 5 to 18 years who presented to the outpatient headache clinic for evaluation of their headaches from 2015 to 2020 were included in the study. Participants were included if they had a diagnosis of migraine and were excluded if they had any secondary causes of headaches except for medication overuse headaches developed in the setting of established diagnosis of migraine. The diagnosis of headache disorders was made using the International Classification of Headache Disorders (ICHD) criteria.⁷ As part of the regular clinical examination evaluation, all participants were assessed for hypermobility using the Beighton score with GJH defined as a Beighton score ≥6.² Clinical equipment such as a protractor was only used by the clinicians when applicable.

Data Collected

The following data were collected for every participant included in the study: sex, race, age at onset of headaches, primary ICHD headache diagnosis (i.e., migraine without aura or migraine with aura, chronic migraine), headache frequency, headache duration, associated symptoms, psychiatric comorbidity, the PedMIDAS score,⁸ clinical examination findings such as the Beighton score, and muscular tightness (Table 2). The temporomandibular joint was examined for range of mouth opening, joint clicking, or crepitation felt on palpation, and any muscle or joint sensitivity and pain.

Table 2.

Demographics and Migraine Characteristics in the 2 Groups

Statistical Analysis

Two groups, migraine with and without GJH, were compared using the χ² test for categorical variables. For continuous-type quantitative variables, the 2 groups were compared using the 2-sample t test (which agreed with the Wilcoxon rank sum test). A p value less than 0.05 was considered statistically significant in all hypothesis tests. Multiple testing was examined using the false discovery rate (FDR) with a family-wise error rate of 0.05. In all cases where unadjusted p values were less than 0.05, they remained so after the FDR adjustment. As a result, the raw unadjusted p values are reported in the Tables. All statistical analyses were conducted using SAS ® version 9.4 (SAS Institute Inc., Cary, NC).

Data Availability

Limited anonymized data not published within this article will be made available on an appropriate and valid request to the authors.

Results

Ten thousand five hundred participants who were evaluated in the headache clinic from 2015 to 2020 were examined for eligibility. Five thousand four hundred thirty-five participants fulfilled the inclusion criteria. Of them, 67.5% of participants were female individuals (3,671/5,435), and 32.5% of participants were male individuals (1,764/5,435). Approximately 31.6% of participants (1,719/5,435) were diagnosed with GJH, and 68.4% of participants (3,716/5,435) were included in the comparison group. All participants with GJH had upper extremity joint flexibility (100%; 95% CI 99.7%–100%), and lower extremity joint flexibility was noted in 87.8% (95% CI 86.1%–89.3%) of participants. The mean Beighton score was 2.2 ± 3.7 (95% CI 2.1–2.4). Demographics and migraine characteristics of both groups are summarized in Table 2.

Headache Characteristics

Migraine without aura was the most common diagnosis in both groups (95.2% of participants with GJH and 91.1% of participants without GJH, n = 5435, χ² with 1 degree of freedom = 28.1, p < 0.001) as opposed to migraine with aura (19.7% of participants with GJH and 23.4% of participants without GJH, n = 5435, χ² with 1 degree of freedom = 9.1, p = 0.002) and chronic migraine (55.9% of participants with GJH and 51.7% of participants without GJH, n = 5435, χ² with 1 degree of freedom = 8.4, p = 0.004). Using the guidance of ICHD recommendations, participants could have multiple migraine subtypes, thus making the total greater than 100%. Concerning associated headache symptoms such as vomiting were noted to be similar in both groups (migraine with GJH and migraine without GJH); however, nausea (73.1% vs 67.5%, n = 5435, χ² with 1 degree of freedom = 17.0, p < 0.0001) and phonophobia (83.7% vs 78.8%, n = 5435, χ² with 1 degree of freedom = 18.0, p < 0.0001) were noted to be more common and severe in participants with GJH compared with those in participants without GJH. Secondary headache disorders such as medication overuse headache was noted to be more common in participants without GJH than in participants with GJH (19.7% vs 16.2%, n = 5435, χ² with 1 degree of freedom = 9.5, p = 0.002). Headache duration, headache frequency, and pain intensity were noted to be similar in both groups. The PedMIDAS score was higher in participants with GJH than that in participants without GJH (48.2 ± 52.5, 95% CI 45.7–50.6 vs 41.6 ± 51.2, 95% CI 40.0–43.3, effect size = 0.13, p < 0.0001). Yet, both averages are in the moderate grade or clinical range.⁸

Physical Examination

Participants with GJH were noted to have nonpainful temporomandibular joint clicking on examination more frequently when compared with participants without GJH (24.4% vs 10.2%, n = 5435, χ² with 1 degree of freedom = 187.6, p < 0.0001) (Table 3).

Table 3.

Clinical Physical Examination Findings in the 2 Groups

Psychiatric Comorbidity

We found that children and adolescents with GJH self-reported psychiatric comorbid conditions such as anxiety (20% vs 16.8%, n = 5316, χ² with 1 degree of freedom = 7.9, p = 0.005) more frequently when compared with participants without GJH (Table 4).

Table 4.

Psychiatry Comorbidities in the 2 Groups

Subanalyses were conducted by splitting the sample into 2 age groups based on the general puberty age of 13 years⁹ (young group younger than 13 years and older group older than 13 years). Concerning the results based on the combined sample, the significance of the association between GJH and headache frequency per month was driven by the young group (without GJH had a mean value = 13.4, 95% CI 13.0–13.8, with GJH had a mean value = 14.5, 95% CI 13.8–15.1, p = 0.0086, effect size = 0.11). The significance of the PedMIDAS score was driven by the older group (without GJH had a mean value = 46.6, 95% CI 44.1–49.2, with GJH had a mean value = 57.7, 95% CI 53.8–61.5, p = < 0.0001, effect size = 0.20). Furthermore, a logistic regression model was fitted with GJH status, and age groups were included as predictors of the probability of having chronic migraine. The interaction between GJH status and age group was examined but was removed because of a lack of statistical significance. Those with GJH had increased odds of experiencing chronic migraine (OR = 1.19, 95% CI 1.05–1.34, p = 0.0046). Similarly, older patient age group had increased odds (OR = 2.68, 95% CI 2.40–2.99, p < 0.0001) of experiencing chronic migraine. Last, the GJH status was replaced by the raw Beighton scale treated as an ordinal variable. The age group effect is approximately the same as in the previous model. For each unit increase in the Beighton scale, the odds of experiencing chronic migraine increased by 2.5% (OR = 1.025, 95% CI 1.008–1.043, p = 0.0035). Thus, the OR of those with GJH compared with those without GJH, 1.19, translates to an approximate 7-point difference in the Beighton scale (1.025⁷ = 1.19).

Discussion

The principal finding of our study was that children and adolescents with migraine who experience GJH have more disabling headache characteristics than those without GJH. Specific questionnaires such as the PedMIDAS, which provides subjective evaluation of the impact of headache on daily life, showed very high scores in patients with GJH. These children and adolescents also self-reported psychiatric comorbid conditions such as anxiety and depression more frequently. On clinical examination, the flexibility of temporomandibular joints was twice as common compared with that in participants with migraine without GJH.

The reported prevalence of GJH varies in the literature from 5% to 65%.^3,4,10 This variation may be due to differences in populations studied, examination methods, and the Beighton score cutoff.

The etiology of GJH is unknown. Genetic heterogeneity is likely the cause, and it is most likely inherited in an autosomal dominant fashion.¹¹ Penetrance is believed to be 100%, although expressivity is variable.¹¹ Expression of this clinical disorder is influenced by sex, with female individuals being the most affected.¹² There are currently no genetic or biochemical tests available to diagnose GJH.

We observed that the prevalence of GJH in children and adolescents affected with migraine is approximately 31.6%. GJH although, an autosomal dominant genetic condition, shows remarkable sex influence.⁴ Following the past study conducted in adults with GJH,¹² we found that female individuals were more predominantly affected by this condition. Increased joint laxity in female individuals is believed to be due to hormonal influences.¹⁰

Per a prior case-control study conducted in 33 adult patients with GJH affected by headache,¹³ we observed that chronic migraine was more predominant in participants with GJH than in controls. However, in contrast to the adult studies,^4,13 the characteristics of migraine headaches, including duration, frequency, pain intensity, and age at onset, were noted to be similar in both groups. Of interest, associated symptoms of migraine, such as nausea, photophobia, and phonophobia, were noted to be more assertive in patients with GJH. We also found that participants with higher Beighton scores had a higher probability of chronic migraines. Participants with GJH also had higher subjective disability on the PedMIDAS questionnaire as opposed to controls.

TMD have been identified as a significant cause of orofacial pain and a recognized cause of headache in the ICHD-3.⁷ TMD can be defined as the clinical conditions affecting the temporomandibular joint and associated structures. Signs and symptoms of TMD can vary from clicking joint sounds to restricted painful mouth opening and asymmetric mandibular movements.^14,15 There is a well-established association between joint hypermobility and TMD.^16-21 In one study 64% of adults with TMD were noted to have GJH.²¹ Another study showed that joint clicking is a sound of disk displacement of the temporomandibular joint in adult hypermobile individuals.¹⁶ Most data suggest that the prevalence of TMD increases with age.^14,22 Although most of the research has been performed in adults with GJH, there are few studies conducted in adolescents.^23-25 The reported prevalence of TMD in children and adolescents can range from 11.9% to 32.5%.^15,22 This variation can be explained by the examination methods and diagnostic criteria.^23-26

Our study sample observed that 24.4% of patients with GJH had TMJ clicking sounds compared with 10.2% of those without GJH. None of the other signs and symptoms of TMD were noted in our study. TMJ click sound or crepitation is the most reported sign of TMD in children and adolescents, given the degenerative changes of the joint increase with age.^14,22 Early diagnosis will help train these patients to tighten the jaw muscles by avoiding overuse of the joints and wearing occlusive splints if needed.

Recent studies have shown a statistical association between anxiety and joint hypermobility in pediatric patients.²⁷ We found that both anxiety and depression were more frequently reported in patients with GJH when compared with those without GJH. The pathogenesis of psychiatric comorbidity in this patient population remains unknown but could relate to other pain disorders and sleep disorders.^28,29

Genetic association between migraine and GJH remains unknown. Tenascin-X (TNXB) haploinsufficiency has been reported in a small subset (5%–10%) of patients with GJH.¹¹ This glycoprotein is involved in the regulation of collagen fibrillogenesis.³⁰ More recent findings suggest a role of other genes such as SLC39A13, which encodes for zinc transporter,³¹ a peptidyl-prolyl cis-trans isomerase (FKBP14),³² and genes involved in the biosynthesis of glycosaminoglycans (B4GALT7, B3GALT6, CHST14, and DSE)^31,32 in the pathogenesis of joint hypermobility and Ehlers-Danlos–like conditions. These genetic changes are known to cause endothelial dysfunction in blood vessels, which can lead to lower vascular resistance, which in turn could lead to increased blood flow. In addition, alterations in specific brain regions such as the brainstem, caudate nucleus, thalamus, anterior cingulate, inferolateral prefrontal cortex, inferotemporal cortex, amygdala, and insula have been noted in patients with GJH.^31,32

This is a retrospective analysis, and the outcome assessment relied on accurate recordkeeping. All participants were seen in a specialized tertiary center headache program. This has the advantage of a very well-characterized group of patients while being complicated by the higher odds of these patients being more disabled from their headaches that resulted in their evaluation in the first place. Comorbid medical disorders such as anxiety and depression were self-reported by the patients and could be subject to error.

The strength of this study is that it was conducted in a large sample size, and each migraine diagnosis was defined by ICHD diagnostic criteria. The diagnosis was made by neurologists, UCNS-Board Certified in Headache Medicine.

In conclusion, GJH is a common finding in pediatric patients with migraine. Children and adolescents with migraine seen in the clinic should be assessed for GJH using the Beighton scale because these patients can have more disabling migraine characteristics than those without GJH.

TAKE-HOME POINTS

• → Generalized joint hypermobility (GJH) is a common finding among children and adolescents with migraine.

• → Children and adolescents with migraine who experience GJH were noted to have higher migraine-related disability and more severe migraine characteristics when compared with children and adolescents with migraine without GJH.

• → Temporomandibular joint flexibility was noted to be twice as common in participants with GJH.

Appendix. Authors

Study Funding

The authors report no targeted funding.

Disclosure

M. Samaha Kabbouche acted as a consultant for Theranica, AMGEN, Teva, and Lilly; was a PI/CoPI for Biohaven, and was a coinvestigator for the NIH. J. Kacperski has received research funding for participation in ongoing studies sponsored by Amgen, Eli-Lilly, Teva, Impax, Biohaven, Currax, Lundbeck, Upsher-Smith, and Curelator. All consulting and research funds for this work are paid directly to Cincinnati Children's Hospital Medical Center. This work does not affect the current presentation. Dr. Kacperski declares that there is no other potential conflict of interest. S. Powers provided scientific consulting for Theranica; received funding to his institution from the NIH and Patient-Centered Outcomes Research Institute; is an Associate Editor for Headache; is a Member of the Board of Directors for the American Headache Society; and provided grant reviews for the NIH and the Cystic Fibrosis Foundation. A.D. Hershey has received funds directly to his institution for advising and implementation of studies from Amgen, AbbVie, Lilly, Teva, Biohaven, Theranica, and Supernus and grant support from the NIH/NINDS and NICHDS. The other authors report no relevant disclosures.