Engelbert, R H. H. PhD, PCS, PT; Kooijmans, F T. C. MA, PT, PCS; van Riet, A M. H. MSc, PT, PCS; Feitsma, T M. MSc, PT, PCS; Uiterwaal, C S. P. M. MD, PhD; Helders, P J. M. MSc, PhD, PCS, PT
Generalized or “common” joint hypermobility has been described extensively in the literature.1–4 Joint hypermobility may result from ligamentous laxity, which is inherent and determined by the genes that encode for collagen, elastin, fibrillin, and tenascin. Excessive range of joint motion is present and may be detected by simple screening techniques. Most persons with hypermobility have no ill effects and enjoy a symptom-free life; in fact, it is an advantage in certain fields (ballet dancers, gymnasts, musicians like violinists and pianists).5,6 However, when generalized hypermobility becomes symptomatic, benign joint hypermobility syndrome (JHS) is said to exist, provided that the subjects do not show signs of any rheumatic, neurological, skeletal, or metabolic diseases.1,4,7–9 The signs include musculoskeletal symptoms with arthralgia in more than two joints for a period exceeding 12 weeks, exercise-induced pain and exercise intolerance. The clinical features of JHS are variable in terms of their nature, severity, and occurrence. Pain is the overriding symptom that may result from the effects of trauma, either acute or more chronic overuse. The term benign is used because of the favorable prognosis in comparison with other, more serious, connective tissue disorders associated with hypermobility such as Ehlers-Danlos, Marfan syndrome, and osteogenesis imperfecta. Criteria for JHS have recently been established for adults, but similar criteria have not yet been validated for children.10
Epidemiological studies have illustrated that hypermobility is seen in up to 10% of individuals in the white population and up to 25% in African and Asian populations. In the Netherlands, idiopathic generalized hypermobility of the joints is reported to be present in about 14% of children between four and 17 years of age.3,10–12 The incidence varies, depending on the criteria used to define the condition, and is reported in about 3.3% of females and 0.6% of males.4 The syndrome is considered usually to be an autosomal dominant inherited disorder.13,14
Many children have hypermobile joints, but only a small percentage of those present complaints. Complaints may occur at any age, and in addition to pain arising from ligamentous and other soft tissues, subluxation, fatigue, autonomic dysfunction, and skin abnormalities such as striae, hyperextensibility and papyraceous scar formation have been reported.15
Clinically manifest symptoms in otherwise healthy children with generalized joint hypermobility are accompanied by an increase in the laxity of other collagen-containing body tissues including bone, vessels, and skin. Thus, generalized joint hypermobility with musculoskeletal symptoms does not seem to be restricted to joint tissue only. In symptomatic hypermobile children, a decrease was also reported in bone density and blood pressure with an increase in skin extensibility when compared to children who are asymptomatic but hypermobile.16
Fatigue is also reported frequently, and increased joint hypermobility has been found in adolescents with chronic fatigue syndrome.17
Generalized joint hypermobility may affect motor development. Some authors stated that in early childhood a delay in motor development is present with catch up in most children before the age of two years, while others stated that gross and fine motor development remained significantly delayed in children who exhibited joint hypermobility and motor delay in infancy.18,19
Multiple explanations for the motor delay in infancy exist. As early as birth, congenital benign hypotonia or floppy infant syndrome and hypermobility of the joints have been associated with each other.20 In older children, diminished proprioception has been reported to be linked to joint hypermobility.21 The characteristic age of prevalence, between three and six years, is not the age of most rapid growth in childhood, but may represent a critical time of change in physical demands and activities, coincident with important changes in body morphology, such as muscle power, balance, and ligamentous support.20 Underlying hypermobility may contribute, therefore, to the appearance of symptoms. As children become older, recurrent injuries related to sporting activities may increase. It is not uncommon to obtain a history of a child who has been proficient at sport, gymnastics, or ballet who has had to give up participation because of musculoskeletal symptoms or problems at the time of increased demands of training and frequency of competition. Still other children report being clumsy in early childhood and having difficulties with participation in physical or sporting activities.20
The presence of generalized hypermobility can be measured by many scoring systems. The first system designed to measure the presence of generalized joint hypermobility was proposed by Carter and Wilkinson22 and modified by Beighton et al23 and is used as the gold standard all over the world. In their original reports, no psychometrical data were mentioned, but validity of the Beighton scoring system has been described recently.24
The Bulbena score was introduced to obtain detailed information about more joints by passive movement.25 Gender related cutoff points were established. In women, a score between 4 and 10 and in men a score between 5 and 10 were the criteria used to establish the presence of generalized joint hypermobility. An intrinsic difficulty in measuring range of passive movement is that the observed range depends on the force applied to the moving part. Until now, a set of criteria for assessing hypermobility has only been validated for adults.23,25 In general practice, the Beighton and Bulbena criteria have been used beginning at six years of age.26
The purpose of this study was to use the Bulbena criteria to establish the presence and location of joint hypermobility and examine its relationship with delay in motor development. Based on clinical experience, we hypothesized that children with generalized joint hypermobility would present a delay in motor development. We also hypothesized that children with generalized joint hypermobility would walk independently at a later age, compared to children without generalized joint hypermobility.
This study was based on a retrospective chart review, consisting of all patients admitted to our facility between 1994 and 2001. Our hospital is a national center for diagnosis and treatment of collagen diseases. Children with generalized joint hypermobility with or without musculoskeletal complaints (arthralgia, fatigue, delay in motor development) have been referred to the department of pediatric physical therapy to contribute to the diagnostic process by documenting the presence of joint hypermobility and musculoskeletal complaints and assessing motor development. The chart review was performed by experienced pediatric physical therapists (F.T.C.K., A.M.H. van R., and T.M.F.) Most children were seen by the first author, a senior pediatric physical therapist, specialized in collagen diseases and symptomatic generalized joint hypermobility.
The medical records of 200 children younger than 2.5 years of age and older than four years of age included documented joint hypermobility. A motor assessment had been performed on 72 children with generalized joint hypermobility. Sixteen were younger than 2.5 years of age, and 56 were four years of age or older. Thirty-three of 72 (48%) were referred to our department by the department of pediatric orthopedics, 23 of 72 (34%) by the department of pediatrics, and 12 of 72 (18%) by others. Indications for referral were joint hypermobility associated with musculoskeletal complaints (arthralgia, fatigue, and delay in motor development) in 24 of 72 (35%), while 48 of 72 (65%) were referred for various other reasons that did not include generalized joint hypermobility.
Inclusion and Exclusion Criteria
All children with documented generalized joint hypermobility and an assessment of motor development were included. Children with generalized joint hypermobility due to collagen diseases (Ehlers-Danlos syndrome, Marfan syndrome, and osteogenesis imperfecta) or children in which a diagnosis was not established were excluded. Children with documented generalized joint hypermobility without an assessment of motor development were excluded.
No instrument for assessing motor development for children between 2.5 and four years of age had, at the time of the study, been standardized for the Dutch population. Therefore, children in this age range were excluded from the study.
Seventy-two children were included in the study. Their mean age was 5.4 years (SD 2.5, range 1.3 to 11.6 years). Boys comprised 58.8% of the sample.
Tests and Measures
Hypermobility According to Bulbena.
The presence of generalized hypermobility of the joints was assessed using the scoring system of Bulbena et al.25 Based on passive maneuvers the presence of hypermobility is scored in nine joints (thumb, little finger, elbow, shoulder, hip, knee, patella, ankle, and first metatarsophalangeal joint) and the presence of ecchymoses is recorded (Table 1). Generalized hypermobility of the joints is present when a score ≥5 is obtained in females and ≥4 in males. Bulbena et al reported high concurrent validity (>0.85) with the Beighton score, and high reliability of repeated measurements (κ interrater values >0.8). Examination of internal consistency reliability showed that the study set of variables was more homogeneous (Cronbach α coefficient = 0.933) than the hypermobility scores of Carter (Cronbach α coefficient = 0.898), Beighton (Cronbach α coefficient = 0.836) and Rotes (Cronbach α coefficient = 0.893).25
An intrarater reliability study of the use of the Bulbena scoring system in our hospital (R.H.H.E.) was performed (Spearman rank correlation coefficient = 0.92).
Bayley Scales of Infant Development.
In children between one and 2.5 years of age, motor development was measured using the Bayley Scales of Infant Development (BSID). This instrument consisted of items related to gross and fine motor activities and was translated and standardized for the Dutch population.27,28 The BSID is a validated assessment tool used to discriminate between normal, suspect, and delayed motor development. The BSID has construct validity of 0.89 and concurrent validity with the Peabody Motor scales (gross motor function: 0.78–0.96 and fine motor skills: 0.2–0.57). Intrarater reliability varies from 0.53 to 0.91 and interrater has been reported to range from 0.88 to 0.99.27,28
A standard score <1 SD below the mean has been defined as normal motor development. A standard score between 1 and 2 SD below the mean was defined as suspect for a delay in motor development. A standard score >2 SD below the mean has been defined as a severe delay in motor development.
Movement Assessment Battery for Children.
In children between four and 12 years of age, the Movement Assessment Battery for Children (MABC) was used with high test-retest reliability (0.75–0.95) and concurrent validity with the Bruininks-Oseretsky test of Motor Proficiency (0.6–0.9)29 This test consists of items measuring manual dexterity, ball skills, and dynamic and static balance. Item scores in these categories can be transformed into a percentile score. A percentile score more than the 15th percentile (P15) has been defined as normal motor development. A percentile score between the fifth percentile (P5) and P15 has been defined as at risk of motor delay. A percentile score lower than P5 has been defined as a severe delay in motor performance.29
At the time of the first physical examination, the age (months) of first independent walking was provided by the parents and recorded. When the child did not walk at first examination due to young age, data on first walking were not recorded.
Central estimators of all relevant variables were calculated as means (SD). The data of the Bulbena score were presented as medians (50th percentile [P50]) and range. The association between the presence of a motor delay (yes/no) and the presence of generalized joint hypermobility was analyzed using univariate regression analysis. Afterward multivariate logistic regression was used to study the association between the amount of joint hypermobility and presence of a delay in motor development based on corrected age. These procedures were performed for children between one and 2.5 years of age as well as for children older than four years of age. Results are presented as odds ratios (ORs) and 95% confidence intervals (95% CIs).
The association between the age of independent walking for the whole population and presence of joint hypermobility was analyzed using linear regression. The age of independent walking was used as the dependent variable, whereas the Bulbena score and age were used as predictors. Associations were expressed as linear regression coefficients with their corresponding 95% CIs.
Data were analyzed using SPSS, version 11.5 (SPSS Inc., Chicago, IL).
Descriptive data of motor development and Bulbena score for all the children are presented in Table 2. In nine of 16 children between one and 2.5 years, a delay (>−2 SD) in motor development was found. No significant association was found between the presence of a delay in motor development and the amount of hypermobility according to Bulbena (OR: 0.4, 95% CI: 0.07–2.1).
In children between four and 12 years of age (n = 56), the median score of motor development assessed with the MABC was at P15 (range 0.5–38). Fourteen of 56 children (25%) had a severe delay in motor development (<P5), while 12 of 56 children (21%) were at risk (P5–P15) and 30 of 56 (54%) were age appropriate (>P15).
No significant association between the presence of a delay in motor development and Bulbena score was found (OR: 1.3, 95% CI: 0.8–2.1; Figure 1). The mean age of reaching the motor milestone of independent walking, which was anamnestically reported by the parents, was 18.0 months (SD: 4.9, range: 11–32) recorded in 35 of 72 children (49%). Univariate regression analysis revealed that the age of independent walking was not significantly associated with the presence of generalized joint hypermobility (Bulbena score) (linear regression coefficient: 0.3, 95% CI: −1.5 to 2.1).
In this retrospective study, we found no significant association between joint hypermobility and a delay in motor development or the age of independent walking.
To fully appreciate the results, some aspects of the study need to be addressed. Retrospective study designs have intrinsic weaknesses. Only associations can be studied, and data for possible confounding variables are lacking. In a cross-sectional or prospective study design, causal associations can be studied and corrected for the possibility of residual confounding. Of the 200 children with generalized joint hypermobility found in the chart review, only in 72 had an assessment of motor development been performed. This selection bias might influence the outcome of the study since we do not know why a motor assessment was not performed in the other children.
Comparing the results of this study with the literature is difficult because of the different research methods and instruments used to measure joint hypermobility and motor development. We found a delay in motor development in 56% of children younger than three years of age, whereas Jaffe et al 18 reported a delay in about 30% in a group of 715 children between eight and 14 months of age. Davidovitch et al 30 reported, as we also found in our study, no relationship between generalized joint hypermobility and gross and fine motor performance in children between five and seven years of age.
Although no significant association was found between the localization and amount of hypermobility and motor development, we found that the median score of motor development items in hypermobile children between four and 12 years of age was at P15, whereas in healthy children, normal motor development is defined as between P15 and the 100th percentile. This might indicate that the motor development of these children is lower than in the general population, but not associated with joint hypermobility. Other factors might influence developmental outcome. Decreased muscle strength has been reported,31 whereas in a cross-sectional study, no significant differences in muscle strength between symptomatic hypermobile children and controls were found.16 The presence of arthralgia in the lower extremities and exercise-induced fatigue and intolerance, which are frequently seen in these children,4 might contribute to age-appropriate but lower scores in motor development. This might suggest that the children are able to perform these motor skills when tested, but in clinical practice, we have observed that they do not perform these motor skills in daily activities. In prospective research, investigation of daily activities should be measured and arthralgia and exercise-induced fatigue could be examined as possible confounders.
In younger children, a tendency for delay in motor development was found. The milestone of independent walking was reached at the mean age of 18.0 months, whereas in the healthy population, this milestone is reached at the mean age of 14.1 months (range: 10.7–17.6).27 This might imply that children with generalized hypermobility have a tendency toward delay in attainment of upright standing and walking. In our practice, we frequently observe axial hypotonia in young children with generalized joint hypermobility. We believe this is primarily related to a collagen disorder rather than being related to a problem in the central nervous system.
Remarkably, 35% of the children in our study were referred with knowledge of the presence of generalized joint hypermobility, whereas in the remaining children the presence of generalized joint hypermobility was reported after visitation to our department. These children were referred because of delay in motor development, or discrepancies in achieving motor milestones. As stated by Grahame,32,33 joint hypermobility is easy to recognize if one looks for it, but equally easy to overlook if one does not.
Hypermobile children are frequently referred to pediatric physical therapists, who may play a role in diagnostics. Assessment of the amount of joint mobility and skin laxity through clinical maneuvers or with a vacuum tissue compliance meter16,24 and description of musculoskeletal complaints may facilitate differential diagnostics. In this way, hypermobility syndrome can be differentiated from collagen diseases as Ehlers-Danlos syndrome, Marfan syndrome, and osteogenesis imperfecta.
Children with generalized hypermobility often report being clumsy in early childhood and having difficulties in any participation in physical or sporting activities. As children progress through primary school with increased demands on handwriting skills, children who are hypermobile may develop upper limb problems. Not infrequently, pain and fatigue in the hand, wrist, or lower arm are reported in these children, particularly by teachers.20
As children become older, recurrent injuries related to sporting activities increase. It is not uncommon to obtain a history of a child who has been proficient at sport, gymnastics, or ballet who has had to give up participation because of musculoskeletal symptoms or problems increasing at the time of increased demands of training and frequency of competition. It is essential that parents and children understand the presence of hypermobility-related musculoskeletal complaints (eg, benign nocturnal leg pains or “growing pains”) may be due to minor injury or repetitive strain to musculotendinous or ligamentous structures, possibly related to unusual or excessive exercise.20
Other chronic musculoskeletal pain syndromes have been linked to hypermobility in childhood. The co-occurrence of fibromyalgia or a fibromyalgia-type pain pattern with generalized joint hypermobility has been reported in schoolchildren.34 It is perhaps understandable that chronic underlying arthralgia and musculoskeletal symptoms can be associated with sleep disturbance and the development of more generalized pain disorders in the vulnerable adolescent.20
The successful management of patients with symptomatic joint hypermobility includes early recognition of joint laxity and knowledge of the symptom complex that the child may present at different ages. Education of the parents as to the nature of the condition and defusing anxiety about less benign conditions is important. Similarly, advice to schoolteachers and sport coaches may be critical in improving symptoms and allowing more gradual rehabilitation and return to full activities or activities that are less likely to lead to recurrent joint injury. Appropriate intervention with physiotherapy and occupational therapy can be important for the optimal management of problems. The intervention may also include psychological support or counselling if the child has developed a chronic pain syndrome complex.20 General management strategies are described in the literature.35
In nearly one third of children with generalized joint hypermobility, a severe delay in motor development was found without a significant association between the amount of generalized joint hypermobility and the delay in motor development. Pediatric physical therapists can play an essential role in differential diagnostics, tailored interventions, and explanation to children with symptomatic joint hypermobility and to their parents.
1. Grahame R. Joint hypermobility and genetic collagen disorders: are they related? Arch Dis Child. 1999;80:188–191.
2. Dalgleish R. The human type I collagen mutation database. Nucleic Acids Res. 1997;25:181–187.
3. Wordsworth P, Ogilvie D, Smith R, et al. Joint mobility with particular reference to racial variation and inherited connective tissue disorders. Br J Rheumatol. 1987;26:9–12.
4. Grahame R. The hypermobility syndrome, review article. Ann Rheum Dis 1990;49:190–200.
5. Klemp P, Stevens JE, Isaacs S. Joint hypermobility study in ballet dancers. J Rheumatol. 1984;11:692–696.
6. Larsson LG, Baum J, Muldolkar GS, et al. Benefits and disadvantages of joint hypermobility among musicians. N Engl J Med. 1993;329:1079–1082.
7. Child AH. Joint hypermobility syndrome: inherited disorder of collagen synthesis. J Rheumatol. 1986;13:239–243.
8. Biro F, Gewanter H, Baum J. The hypermobility syndrome. Pediatrics. 1983;72:701–706.
9. Jessee E, Owen D, Sagar K. The benign hypermobile syndrome. Arthritis Rheum. 1980;23:1053–1056.
10. Birrell FN, Adebajo AO, Hazleman BL, et al. High prevalence of joint laxity in West Africans. Br J Rheumatol. 1994;33:56–59.
11. Rikken-Bultman DG, Wellink L, Van Dongen PW. Hypermobility in two Dutch school populations. Eur J Obstet Gynaecol Reprod Biol. 1997;77:189–192.
12. Larsson LG, Baum J, Muldolkar GS. Hypermobility: features and differential incidence between the sexes. Arthritis Rheum. 1987;30:1426–1430.
13. Horan F, Beighton P. Recessive inheritance of generalized joint hypermobility. Rheumatol Rehabil. 1973;12:47–49.
14. Grahame R, Edwards JC, Pitcher D, et al. A clinical and echocardiographic study of patients with the hypermobility syndrome. Ann Rheum Dis. 1981;40:541–546.
15. Grahame R. Pain, distress and joint hyperlaxity. Joint Bone Spine. 2000;67:157–163.
16. Engelbert RHH, Bank RA, Beemer FA, et al. Pediatric generalised joint hypermobility with musculoskeletal complaints; a localized or systemic disorder ? Pediatrics. 2003;111:e248–e254.
17. Barron DF, Cohen BA, Geraghty MT, et al. Joint hypermobility is more common in children with chronic fatigue syndrome than in healthy controls. J Pediatr. 2002;141:421–425.
18. Jaffe M, Tirosh E, Cohen A, et al. Joint mobility and motor development. Arch Dis Child. 1988;63:158–161.
19. Tirosh E, Jaffe M, Marmur R, et al. Prognosis of motor development and joint hypermobility. Arch Dis Child. 1991;66:931–933.
20. Murray KJ, Woo P. Benign joint hypermobility in childhood. Rheumatology. 2001;40:489–491.
21. Hall MG, Ferrell WR, Sturrock RD, et al. The effect of the hypermobility syndrome on knee joint proprioception. Br J Rheumatol. 1995;34:121–5.
22. Carter C, Wilkinson J. Persistent joint laxity and congenital dislocation of the hip. J Bone Joint Surg Br. 1964;46:40–45.
23. Beighton PH, Solomon L, Soskolne CL. Articular mobility in an African population. Ann Rheum Dis. 1973;32:413–418.
24. van der Giessen LJ, Liekens D, Rutgers KJ, et al. Validation of Beighton score and prevalence of connective tissue signs in 773 Dutch children. J Rheumatol. 2001;28:2726–2730.
25. Bulbena A, Duro JC, Porta M, et al. Clinical assessment of hypermobility of the joints: assembling criteria. J Rheumatol. 1992;19:115–121.
26. Adib N, Davies K, Grahame R, et al. Joint hypermobility syndrome in childhood. A not so benign multisystem disorder? Rheumatology (Oxford). 2005;44:744–750.
27. Bayley N. Bayley Scales of Infant Development. New York: The Psychological Corporation; 1996.
28. van der Meulen B.F., Smrkovsky M. Bayley ontwikkelingsschalen. Lisse, The Netherlands: Swets & Zeitlinge; 1983.
29. Henderson SE, Sugden DA. Movement Assessment Battery for Children. Kent, UK: Psychological Cooperation; 1992.
30. Davidovitch M, Tirosh E, Tal Y. The relationship between joint hypermobility and neurodevelopmental attributes in elementary school children. J Child Neurol. 1994;9:417–419.
31. Maillard S, Murray KJ. Hypermobility syndrome in children. In: Keer G, ed. Hypermobility Syndrome, Recognition and Management for Physiotherapists. Edinburgh: Butterworth/Heinemann; 2003.
32. Grahame R, Bird H. British consultant rheumatologists’ perceptions about the hypermobility syndrome: a national survey. Rheumatology (Oxford). 2001;40:559–562.
33. Grahame R. Time to take hypermobility seriously (in adults and children) [Editorial]. Rheumatology. 2001;40:485–486.
34. Gedalia A, Garcia CO, Molina JF, et al. Fibromyalgia syndrome: experience in a pediatric rheumatology clinic. Clin Exp Rheumatol. 2000;18:415–419.
35. Middleditch A. Management of the hypermobile adolescent. In: Keer G, ed. Hypermobility Syndrome, Recognition and Management for Physiotherapists. Edinburgh: Butterworth/Heinemann; 2003.
© 2005 Lippincott Williams & Wilkins, Inc.