Generalised joint hypermobility (GJH) is defined as increased mobility in several joints caused by an increased laxity of connective tissue.1 In epidemiological studies, the Beighton score is used to identify GJH,2 as it has exhibited a moderate to high repeatability3,4 and is a self-report measure that offers a valid and reliable assessment of joint hypermobility.5 The prevalence of GJH (Beighton score≥4/9) in children ranges from 19.2% to 58.9%6–10 and is associated with female sex,9–11 younger age,2,8,10,12–14 non-white race,10 malnutrition,6,9,13 a greater level of physical activity,6,9 higher socioeconomic status6 and greater maternal educational attainment.9 The prevalence of functional gastrointestinal disorders (FGIDs) as defined by the current Rome IV criteria in children aged 4–18 years ranges from 21.2% to 25.0%, and the most prevalent among them are functional constipation and functional dyspepsia.15,16

In adults, there is evidence of an association between functional gastrointestinal manifestations and GJH, joint hypermobility syndrome (JHS) and Ehlers–Danlos syndrome-hypermobility type (EDS-HT).17–24 Most adults with EDS-HT report gastrointestinal symptoms, most commonly epigastric pain and constipation.22 In children, the evidence supporting an association between the presence of FGIDs and JHS is scarce and controversial.25,26 Most studies that assessed the association between FGIDs and JHS in children and adults were conducted in patients that sought care for gastrointestinal manifestations.18,19,21–24 Only 1 study has analysed the association between FGIDs and JHS in the non-patient population,24 and it did not find an association between JHS and FGIDs. The study has not yet been replicated, and no study to date has analysed this association applying the Rome IV criteria. In light of recent studies that have found an association between certain glycoproteins, such as glycoprotein-tenascin-X, and changes in gastrointestinal function,27 we believe it would be relevant to perform studies to determine whether FGIDs and GJH are associated in the overall paediatric population, which could reflect commonalities in the pathophysiology of connective tissue and functional gastrointestinal disorders, and help elucidate the pathogenesis of FGIDs, which still remains unclear. The potential implications of such an association would transcend the paediatric population, as a proportion of children with FGIDs become adults with FGIDs. The aim of our study was to assess the association between GJH and FGIDs and identify risk factors for GJH in girls attending a public school in Tuluá, Colombia.

We conducted a matched case-control study between February 27 and May 16, 2018 in a public school for girls in Tuluá, a city at the centre of the Valle del Cauca, Colombia, with 219138 inhabitants. We started by administering the Spanish version of the Questionnaire on Pediatric Gastrointestinal Symptoms (QPGS) based on the Rome IV criteria, which had been previously validated by our research group28 to female students aged 10 to 18 years. We excluded girls that reported organic disorders in this questionnaire. We then included girls with any form of FGIDs (cases), whom we matched to other girls of similar age and school year that did not meet the criteria for a FGID (controls). Both groups of girls were assessed for GJH by means of the Beighton hypermobility score.2 We also collected data on sociodemographic variables such as race, age (taking into account whether they were school-aged or adolescent), family history of FGIDs and anthropometric measures including weight, height and waist circumference.

Applying the Rome IV criteria, we classified FGIDs as cyclic vomiting syndrome, functional nausea and functional vomiting, rumination syndrome, aerophagia, functional dyspepsia, irritable bowel syndrome, abdominal migraine, functional abdominal pain not otherwise specified, functional constipation and nonretentive faecal incontinence. We classified girls with a Beighton score of 4/9 as cases of GJH. We divided the girls by age group into schoolchildren (10–12 years) and adolescents (13–18 years); by race into mixed, of African descent, white and indigenous; by World Health Organization (WHO) obesity criteria into obese, overweight, with normal weight, moderately underweight and severely underweight; and based on the Vargas classification29 as having or not having abdominal obesity.

We have expressed results as absolute frequencies and percentages or as mean and standard deviation. We analysed the data using Stata 15, using the chi square test or Fisher exact test as applicable. We performed univariate, bivariate and multivariate regression analysis to assess potential effects on the variables of interest and the diagnosis of GJH. We assessed the strength of associations by calculating odds ratios (OR) with their corresponding 95% confidence intervals (CIs). Results with a P-value of less than 0.05 were considered statistically significant.

We obtained signed informed consent from the legal guardians of participants as well as the participants themselves. The study was approved by the Ethics Committee of the Universidad del Valle and the administration of the public school for girls.

Our study in girls aged 10 to 17 years did not found an association between the presence of FGIDs, identified by means of the Spanish version of the QPGS (Rome IV criteria), and GJH. These results contrast with the findings of Kovacic et al.,25 whose study in a group of children and adolescents with FGIDs managed at a gastroenterology clinic, most frequently with a diagnosis of irritable bowel syndrome or functional dyspepsia, found a 10.03-times greater probability of having GJH (95% CI, 5.26–19.13; P=.001) compared to healthy adolescents in the general population, which led the authors to suggest that an abnormal connective tissue matrix may be involved in the pathophysiology of FGIDs. These data are consistent with our previous findings, although they were obtained in children of both sexes and with a diagnosis of FGID according to the Rome III criteria.9 Our results cannot be compared to the findings of Chelimsky et al.,21 who did not find autonomic dysfunction in girls with benign joint hypermobility syndrome. The cumulative data of the study by Kovacic et al.25 and studies in adults that sought care for gastrointestinal complaints19,20,22–24 suggest that when an association is found between GJH and FGIDs, it is only in patients that seek care because the severity of their comorbidities accounts for the development of FGIDs, and also that there is no organic explanation for this association.

In our sample of girls with and without FGIDs, we found a prevalence of GHJ of 44.1%, a greater proportion compared to 32.9% of the girls included a previous case-control study we conducted in Cali, Colombia (2445281 inhabitants)9 and smaller compared to the prevalence of 56.0% described by Kovacic et al. in the United States.25 These differences in prevalence can be attributed not only to differences in population, ethnicity and regional characteristics, but also to differences in the sex and age of participants and the questionnaire used to identify FGIDs. This overall prevalence of GJH is within the range reported in previous studies at the international level (19.2–60.7%).6–10

In our study, we did not find an association between GJH and any of the analysed variables. As for the association between GJH and FC, which was the main form of FGID found in this sample of girls with a prevalence of 29.0%, Reilly et al.20 found that GJH was much more prevalent in boys with slow transit constipation (STC), which suggests that abnormalities in the synthesis of connective tissue may play a role in the aetiology of STC. However, it is important to take into account that STC is an organic disorder rather than a functional one. Similarly, a study by Kajbafzadeh et al.26 in children with another form of organic disease, in this case urological (urinary dysfunction), found that constipation was more frequent in male patients with GJH compared to those without GJH. In our previous case-control study in children in Cali, Colombia,9 we found that female sex and younger age were associated with GJH. In the general population, different authors have found associations between GJH and higher socioeconomic level,6 female sex,9,10 younger age,2,8,10,12–14 non-white race,8 malnutrition,6,9,13 higher level of physical activity6 and higher maternal educational attainment.10

Some of the strengths of our study is that we used the methodology proposed for FINDERS (Functional International Digestive Epidemiological Research Survey Group) applied to our previous study with the former version of the Rome criteria,11 in which girls were evaluated by staff trained on the performance of the Beighton score and the use of questionnaires in Spanish validated for diagnosing FGID in a context that allowed us to obtain data on children in the community without the selection bias inherent in the profile of individuals that seek medical care. Some of the weaknesses are that our study only included female students and was conducted in a single public school without comparison with a private school, that we did not ask about other potential comorbidities that could be useful to assess the impact of GJH and FGIDs, and that these were all girls that attended school and did not go to a health care facility.

In conclusion, our study in girls aged 10–17 years enrolled in one public school did not find an association between GJH and the presence of FGIDs identified by means of the Spanish version of the QPGS based on the Rome IV criteria. We were also unable to identify any risk factor for GJH among the variables analysed, which suggests the need for different types of studies enrolling children managed in specialty clinics able to assess whether connective tissue laxity is associated with the presence of severe FGID and the presence of comorbidities, which leads to seeking medical care.

The authors have no conflicts of interest to declare.