Joint hypermobility in children with idiopathic scoliosis: SOSORT award 2011 winner

The case-control study included 70 subjects aged 9-18 years (mean 13.2 ± 2.2) with idiopathic scoliosis. The criteria for inclusion to the study group were the following: diagnosis of idiopathic scoliosis on antero-posterior radiogram in accordance with SRS criteria (Cobb angle > 10° with rotation), age range 9-18 years, absence of systemic diseases related to JHM (Ehlers-Danlos, Down, Marfan, Larsen) and participation consent. The control group consisted of 58 subjects (mean age 12.6 ± 2.1 years) randomly chosen according to the inclusion criteria: age range 9-18 years, less than 5° of angle of trunk rotation as measured with Bunnell scoliometer, absence of systemic diseases and participation consent. Approval of the local ethical committee was obtained. The basic data for both groups is presented in Table 1.

Additionally, the comparability between the study group and the control group was checked separately for the girls (n = 92) and boys (n = 36). The girls from the study group were significantly taller than the girls from the control group (p = 0.02), however, the groups were comparable in respect to age (p = 0.06), weight (p = 0.2) and BMI (p = 0.4). Boys from the study group were comparable to boys from the control group in respect to age (p = 0.1), height (p = 0.1), weight (p = 0.3) and BMI (p = 0.8).

Within the study group, the Cobb angle range was 10° to 53°. In 34 cases, a single-curve thoracic scoliosis was present, while in the remaining 36 cases - a double-curve thoracic and lumbar scoliosis. The average number of vertebrae forming single-curve scoliosis was 6.5 vertebrae (range 4-10). The value of apical vertebra rotation (AVR) was determined using the Cobb method [22]. 47 subjects were treated exclusively with physiotherapy, while the remaining 23 subjects received both physiotherapy and treatment with Cheneau brace.

A nine-degree Beighton scale was used to determine the occurrence of generalized joint hypermobility [23] (Figure 1, 2, 3, 4, 5). In order to assess the range of joint mobility, a set of goniometers was used (Baseline, USA).

Additionally, a five-part questionnaire by Hakim and Graham was used [24] (Table 2). Obtaining 4 or more points on Beighton scale and simultaneously at least 2 points in the ques-tionnaire by Hakim and Graham was adopted as a criterion for JHM diagnosis.

The frequency of occurrence of the JHM was compared between the study and control groups. The same analysis was also performed separately for girls and boys. JHM occurrence was compared between mild (Cobb 10°-24°), moderate or severe (Cobb ≥25°) idiopathic scoliosis. Correlation between the Cobb angle and the number of points on the Beighton scale was checked. The relation of JHM occurrence to the following parameters was tested: (1) the apical vertebra rotation quantified according to Cobb as + versus ++ versus +++, (2) number of curvatures - single versus double, (3) number of vertebrae within the single scoliosis curvature: above versus below the mean of 6.5 vertebrae, (4) type of management: physiotherapy alone versus physiotherapy and corrective bracing, (5) the age of patient: above or below the mean of 13.2 years. Correlation between the number of points in Beighton scale and age as well as between Beighton scale and the number of vertebrae included in the primary curvature was calculated.

Statistical analysis was performed with Statistica 8.1 (StatSoft, USA). The Kolmogorov-Smirnov test was initially applied to check normal distribution. The U Mann-Whitney test, Chi² test, Pearson and Spermann correlation rank were also conducted to. The value p = 0.05 was adopted as the level of significance.

Different authors apply various names to define increased joint mobility, despite using similar diagnostic scales. The following exemplifies this variety: Joint hypermobility (JHM), Joint hypermobility syndrome - JHS, Benign hypermobility syndrome (BHS), Benign joint hypermobility syndrome (BJHS) or Joint laxity [2‐8, 12, 25‐31]. This lack of consistency is an obstacle to direct comparison of results [4]. Gedalia et al., with the use of Beighton scale, reported that JHM occurs in 12% of American pupils aged 5-17, where in 18% of girls and in 7% of boys [2]. The study by Vougiouka et al. assessing right limbs and lumbar spine flexion proved that Benign Hypermobility Syndrome occurs in 8.78% of children aged 5-14 [3]. The authors emphasize that BHS occurs in a significantly smaller percentage of boys than girls (7.1% versus 10.7%). Hakim and Grahame analyzed the results of 7 examinations conducted on children aged 11-17 [4]. Their analysis shows that JHM occurs in approximately 10-15% of boys and 20-40% of girls. Observations made with the use of Beighton scale in Brazilian children aged 4-7 indicated that as many as 64.6% of children in that group presented JHM [5]. Similar figures were obtained by de Inocencio Arocena et al. (Beighton scale), who confirmed JHM occurrence in 55% of children aged 4-14 [6]. It seems that important differences between publications result basically from the diverse methodologies used.

The most frequent method used in clinical screening is Beighton 9-point scale [4, 23, 32, 33]. It consists of exclusively assessing joint mobility: extension of the fifth MPC joint to 90°, thumb abduction to front forearm, hyperextension of elbow joint above 10°, hyperextension of knee joint above 10° as well as capability to stand bend and place one's palms flat on the ground. Each hypermobile joint gets one point. To diagnose JHM, at least 4 points have to be obtained [4, 23, 34].

Carter and Wilkinson published a method of assessment of generalized joint hypermobility about a decade before Beighton score was introduced [29‐31]. Their scale takes into account similar articulations: the thumb, elbow and knee joints. The first difference from Beighton scale concerns the assessment of passive hyperextension of all four II-V fingers to a position parallel to the forearm extensor aspect (Figure 6), instead of the assessment of the fifth finger only. The second difference involves assessing the range of ankle dorsiflexion beyond 45° in the Carter and Wilkinson's method (Figure 7), instead of assessing the ability to touch the ground with one's palms as adopted by Beighton. Both scales are considered to be reliable in the assessment of joint hypermobility [30]. It is also important to notice that ankle dorsiflexion beyond 45° may be limited even in hypermobile subjects by relative shortening of the triceps muscle which is quite a common condition in a growing population. On the other hand, touching the floor with one's hands, as proposed by Beighton, can be executed by subjects who are not hypermobile but who have experienced intensive stretching of the hamstring muscles.

Marshall test is another method of assessing joint hypermobility based on the thumb motion range measured in the forearm direction [35]. The following figures show scores obtained during this test: I - < 45° of passive thumb abduction (Figure 8); II - 45° of abduction (Figure 9); III - 90° of abduction (Figure 10); IV - 135° of abduction (Figure 11), and V - thumb can be opposed to the forearm (Figure 12). Marshall test is widely used for its simplicity. The disadvantage of this test is the fact that it focuses only on one joint.

Bulbena scale is a 10-point scale used to assess generalized joint hypermobility [36]. It takes into consideration the signs in upper extremities, lower extremities and skin. The above described methods are summarized in Table 12.

Apart from various methods used to assess generalized joint hypermobility, within a given method, different cut-off values have been proposed. For the purposes of diagnosing JHM in children, Lamari et al. adopted 3 of the 9 points values from Beighton scale [5]. A similar threshold was adopted by Gedalia et al. [2]. According to Grahame and Bird 87% of British rheumatologists consider that at least 3 hypermobile joint suffice to diagnose JHM [37]. However, Smits-Engelsman et al. suggests that the minimal threshold value should be at least 7 points in the 9-point scale [32]. In 1998, criteria from Brighton were introduced to diagnose BJHS, comprising the so-called major and minor criteria. However, the Brighton scale requires specific examinations which are difficult in screening conditions, for example rectal prolapse or joint inflammation [4, 28]. Hakim and Grahame as well as Hakim and Sahota suggest that the diagnosis of JHM should be supplemented with a 5-point questionnaire [24, 34]. It would allow a fast clinical overview as its questions refer to symptoms observed both at present and in the past, thus accounting for mobility changes that occur with age. A positive answer to two or more questions would indicate hypermobility with sensitivity of 80-85% and specificity of 80-90% [4, 24, 34].

The authors of this study used a 9-point Beighton scale to assess JHM prevalence with the threshold level of minimum 4 points, supplemented with a 5-part questionnaire to avoid false positive results, as suggested by Hakim and Grahame [23, 24].

The studies determining the prevalence of JHM usually refer to healthy population of children and adolescents [2, 3, 5, 6, 38, 39]. The second group of papers compares occurrence of accompanying symptoms and signs in children with JHM versus their prevalence in the control group [4, 7‐9, 40, 41]. The authors, however, are not aware of any reports on JHM prevalence in children suffering from idiopathic scoliosis. Binns found de-creased thumb-to-forearm distance measured during passive apposition in Chinese patients with adolescent idi-opathic scoliosis [42]. This suggests that they are relatively hyperlax; however, the study was based solely on thumb mobility.

The issue of JHM in idiopathic scoliosis seems clinically important as these children may be subjected to intensive physiotherapy according to various approaches ("schools"). These therapeutic activities include, for example, flexibility exercises, passive stretching exercises, as well as proprioception exercises [16‐21, 43]. Thus, it seems crucial to assess JHM prevalence in these patients. According to Keer and Grahame, physiotherapy is a mainstay of musculoskeletal aspects of JHS treatment [44]. However, those authors emphasize the fact that patients often complain about physiotherapy, claiming that their condition seems to have deteriorated, especially when their hypermobile joints are not handled with due care while treated manually. Moreover, physiotherapists often admit that their knowledge about JHM patients is not sufficient [44]. Such knowledge has great significance since it is typical of JHM patients to be particularly vulnerable to soft tissue injuries which occur during physical exercises. Additionally the increased sensitivity of pain receptors as well as recurrent micro injuries can increase the risk of joint injuries [4, 41].

The results indicate significantly higher JHM prevalence in children with idiopathic scoliosis, supported by the commonly recognized negative impact of this impairment on the locomotor system of children treated with physiotherapy. Assessing JHM needs to be taken into consideration when planning physiotherapy for IS children. This is even more true given the fact that JHM diagnosis is not time consuming and is based on screening tests that are easy to perform.