Congenital dysplasia of the hip (CDH; MIM 142700) is one of the most common congenital skeletal anomalies. CDH is an abnormal seating of the femoral head in the acetabulum 1. CDH acts as a significant risk factor for the development of hip osteoarthritis 234. Shallow acetabulum and lax capsule around the hip joint are the main causes of CDH 56. Former epidemiological investigations show that CDH has a considerable genetic component. Several family studies of CDH have showed that its prevalence was significantly higher in first-degree relatives of probands 789. A study of identical twins indicated hereditary factors are of prime importance in CDH 10, and a genome-wide screening of a Japanese family with acetabular dysplasia identified a linkage on a limited location of the specific chromosome 11.

Growth differentiate factor 5 (GDF5; also known as cartilage-derived morphogenetic protein-1) is a member of the transforming growth factor-β (TGF-β) super-family. GDF5 is expressed in the regions between skeletal elements where joints will later form 1213. It plays a crucial role in the morphogenesis of tendon, ligament and bone. A null mutation of GDF5 causes developmental failure of skeletal structure and intra-articular ligaments in mice 1415. Type C brachydactyly (MIM 113100) is a skeletal disorder caused by GDF5 mutation 1617, and some patients with type C brachydactyly also present with dysplasia of hip joints 1819.

Recently, a functional single nucleotide polymorphism (SNP) in the 5'-untranslated region of GDF5 (rs143383; +104T/C) was found to be significantly associated with osteoarthritis in people of Japanese and Han Chinese origin 20. This SNP was located in the GDF5 core promoter and exerted allelic differences in promoter activity of the GDF5 gene. The susceptibility allele (+104T) showed reduced transcriptional activity of GDF5 in chondrogenic cells 20. Association of this SNP with osteoarthritis has been replicated in people of European origin 21. These findings suggest that GDF5, especially the functional SNP rs143383, may play a key role in the aetiology and pathogenesis of CDH. To evaluate this possible association, we examined the genetic association of the GDF5 SNP with CDH in people of Han Chinese origin and found a compelling association between GDF5 and CDH.

The ages of patients with CDH and controls (mean ± SD) were 21.6 ± 12.4 months (range 2 to 72 months) and 58.1 ± 11.0 years (range, 39 to 94 years), respectively. More than 50% of the CDH cases were delivered by caesarean section. The ratio of female to male was about six to one in patients with CDH. Distributions of genotypes in the CDH and control groups were conformed to Hardy-Weinberg equilibrium (p = 0.77 and 0.50, respectively) (Table 1). The distribution of the severity of the hip disease was 6% with score 1, 16% with score 2 and 78% with score 3 (Table 2). Significant differences in allele frequency was detected between CDH and control groups (p = 0.0037) (Table 3). Significant differences in the genotype frequency were observed in the comparison of TT (T allele homozygote) and other genotypes combined (p = 0.013), and in a comparison of CC (C allele homozygote) and other genotypes combined (p = 0.029) (Table 3). No significant difference was found between different delivery methods (p = 0.78).

We stratified subjects by gender and compared the genotype distribution and allele frequency. In female samples, the most significant difference was observed in the allele frequency (p = 0.0053) (Table 3). The genotype distribution and allele frequency in male members of the CDH and control groups were similar to that in the female samples and all samples as a whole. No significant difference was detected in the comparison of genotype and allele frequency between male CDH and control subjects (Table 3). No significant difference was detected in any comparisons between female and male cases or female and male controls. A significant difference was found between samples with hip dislocation when stratified by severity (p = 0.0078) and no significant difference was found in subjects with hip instability and subluxation (Table 4). When all subjects were stratified by severity (0 = control; 1 = instability; 2 = subluxation; 3 = dislocation), a significant increasing linear trend (p = 0.020) was seen in the T allele frequency as the severity worsened.

To the authors' knowledge this is the first demonstration of a compelling association betweenn functional GDF5 SNP rs143383 and CDH in the Han Chinese population. Significant differences were observed in allele frequency, and in comparisons of TT versus other genotypes combined and CC versus other genotypes combined. Significant differences were also observed in females after stratification of gender. Distribution of genotype in males was similar to that in females and the group as a whole, although no significant differences were detected in genotype and allele frequencies. No significant difference was found in any comparison between female and male subjects. The lack of significance in male subjects may be due to the limited sample number, although a large sex bias of CDH incidence obviously exists. To clarify this possible association, further research should be conducted with a larger sample number.

We discovered the significant association with hip dislocation when stratified by severity, but not with subluxation and instability. A significant increasing linear trend in the T allele frequency as the severity worsens was also observed. This indicates that the SNP may be associated with severity of CDH, but a definite conclusion could not be made because the sample number was so limited and no significant association was detected among groups of different severity.

GDF5 has been found to play an indispensable role in joint morphogenesis and GDF5 can promote the condensation of mesenchymal cells, which is the initiate step of developing cartilage element. GDF5 can also enhance chondrogenic differentiation of mesenchymal cells 2425262728. The T allele of rs143383 was overrepresented in CDH, and it showed a reduced transcriptional activity of GDF5 in vitro and in vivo 2021. Reduced expression of GDF5 would decrease the condensation and chondrogenic differentiation of mesenchymal cells and result in a reduction in the amount of chondrogenic cells in hip joints. It leads to a developmental deficiency of the acetabulum and proximal femoral element, especially the femoral head. As mentioned above, the absence of GDF5 can cause developmental failure of intra-articular ligaments in mice 14, so we suspected that a reduction of GDF5 expression may also lead to developmental deficiency of the ligaments and capsule in and around the human hip joint. Insufficiency of osteal elements and soft tissues in and around hip joints could contribute to susceptibility to CDH simultaneously or individually. Further study on local expression of GDF5 is needed to explore detailed mechanisms between reduced GDF5 expression and CDH.

Several association studies have been carried out to detect the susceptibility gene for CDH 2930313233, and most of them produced negative results 293031. One study found that a MSX1 polymorphism was associated with limb deficiency defects including CDH 32, but it no individual data for CDH was shown. Two polymorphisms in type II collagen and vitamin D receptor genes were reported to be associated with osteoarthritis secondary to hip dysplasia 33, but another study showed a negative association of these two polymorphisms with nonsyndromic CDH 29. Whether these two polymorphisms are associated with hip dysplasia or with osteoarthritis is still disputed. Our study is the first report of association between SNP and clearly defined CDH. Further studies should be conducted with larger sample numbers in different ethnic groups.

Our study suggested that there is an association between GDF5 and CDH susceptibility in a Chinese Han population.

CDH: congenital dysplasia of the hip; CI: confidence interval; GDF5: growth differentiate factor 5; OR: odds ratio; SNP: single nucleotide polymorphism; TGF-β: transforming growth factor-β.

The authors declare that they have no competing interests.

All authors contributed to the final manuscript. In addition, JD and DQS genotyped the samples and participated in the design and analysis of the study. PZ, JQ, HN, YX, CY, LZ, HZ, BZ and JW evaluated the patients and genotyped these samples. BL and SI coordinated the study. QJ and YD supervised the whole study.