Developmental dysplasia of the hip (DDH) is the most common skeletal developmental disease. However, its genetic architecture is poorly understood. We conduct the largest DDH genome-wide association study to date and replicate our findings in independent cohorts. We find the heritable component of DDH attributable to common genetic variants to be 55% and distributed equally across the autosomal and X-chromosomes. We identify replicating evidence for association between GDF5 promoter variation and DDH (rs143384, effect allele A, odds ratio 1.44, 95% confidence interval 1.34–1.56, P = 3.55 × 10−22). Gene-based analysis implicates GDF5 (P = 9.24 × 10−12), UQCC1 (P = 1.86 × 10−10), MMP24 (P = 3.18 × 10−9), RETSAT (P = 3.70 × 10−8) and PDRG1 (P = 1.06 × 10−7) in DDH susceptibility. We find shared genetic architecture between DDH and hip osteoarthritis, but no predictive power of osteoarthritis polygenic risk score on DDH status, underscoring the complex nature of the two traits. We report a scalable, time-efficient recruitment strategy and establish for the first time to our knowledge a robust DDH genetic association locus at GDF5.

Background: Developmental dysplasia of the hip (DDH) is a common, heritable condition characterised by abnormal formation of the hip joint, but has a poorly understood genetic architecture due to small sample sizes. We apply a novel case-ascertainment approach using national clinical audit (NCA) data to conduct the largest DDH genome-wide association study (GWAS) to date, and replicate our findings in independent cohorts. Methods: We used the English National Joint Registry (NJR) dataset to collect DNA and conducted a GWAS in 770 DDH cases and 3364 controls. We tested the variant most strongly associated with DDH in independent replication cohorts comprising 1129 patients and 4652 controls. Results: The heritable component of DDH attributable to common variants was 55% and distributed similarly across autosomal and the X-chromosomes. Variation within the GDF5 gene promoter was strongly and reproducibly associated with DDH (rs143384, OR 1.44 [95% CI 1.34-1.56], p=3.55x1022). Two further replicating loci showed suggestive association with DDH near NFIB (rs4740554, OR 1.30 [95% CI 1.16-1.45], p=4.44x10-6) and LOXL4 (rs4919218, 1.19 [1.10-1.28] p=4.38x10-6). Through gene-based enrichment we identify GDF5, UQCC1, MMP24, RETSAT and PDRG1 association with DDH (p<1.2x10-7). Using the UK Biobank and arcOGEN cohorts to generate polygenic risk scores we find that risk alleles for hip osteoarthritis explain <0.5% of the variance in DDH susceptibility. Conclusion: Using the NJR as a proof-of-principle, we describe the genetic architecture of DDH and identify several candidate intervention loci and demonstrate a scalable recruitment strategy for genetic studies that is transferrable to other complex diseases.

two years in the mean score of four of the five KOOS subscales, covering pain, symptoms, activities of daily living, and quality of life (KOOS 4); scores ranging from 0 (worst) to 100 (best). Results: In the trial of TKR, 18 (36%) patients randomized to nonsurgical treatment alone underwent TKR during the two years (five between one and two years), while eight patients randomized to nonsurgical treatment and ten patients randomized to usual care underwent TKR during the two years in the other trial. Patients randomized to TKR followed by non-surgical treatment had greater improvement in KOOS 4 than patients randomized to non-surgical treatment alone in either trial (Fig. 1; 34.6 vs. 16.1 and 18.5 respectively; adjusted mean difference of the trial (95% confidence interval (95% CI)) of 18.0 (À24.1 to À11.9)). Patients randomized to non-surgical treatment alone in either trial improved more than patients randomized to written information and treatment advice (Fig. 1; 16.1 and 18.5 respectively vs. 11.6; adjusted mean difference of the trial (95% CI) of À5.8 (À11.0 to À0.6)). Conclusions: TKR followed by non-surgical treatment resulted in greater improvements in pain and function than non-surgical treatment alone and non-surgical treatment resulted in greater improvements than written information after two years in patients with knee OA. Nearly two out of three patients with moderate to severe knee OA eligible for TKR postponed surgery for at least two years following nonsurgical treatment. These results should encourage clinicians and patients to discuss benefits and harms of both surgical and non-surgical treatment options to decide what treatment best meets the need and expectations of the individual patient.

No abstract