DXA-derived hip shape is related to osteoarthritis: findings from in the MrOS cohort.

B.G Faber,E Orwoll,J.H Tobias,J Lynch,C.L Gregson,N.E Lane,D Baird,J.S Gregory,R.M Aspden,R.J Barr,M.C Nevitt

doi:10.1016/j.joca.2017.09.006

Abstract

SummaryObjectiveStatistical shape modelling (SSM) of radiographs has been used to explore relationships between altered joint shape and hip osteoarthritis (OA). We aimed to apply SSM to Dual-energy X-ray Absorptiometry (DXA) hip scans, and examine associations between resultant hip shape modes (HSMs), radiographic hip OA (RHOA), and hip pain, in a large population based cohort.MethodSSM was performed on baseline hip DXA scans from the Osteoporotic Fractures in Men (MrOS) Study. Associations between the top ten HSMs, and prevalent RHOA from pelvic radiographs obtained 4.6 years later, were analysed in 4100 participants. RHOA was defined as Croft score ≥2. Hip pain was based on pain on walking, hip pain on examination, and Western Ontario and McMaster Universities Arthritis Index (WOMAC).ResultsThe five HSMs associated with RHOA showed features of either pincer- or cam-type deformities. HSM 1 (increased pincer-type deformity) was positively associated with RHOA [1.23 (1.09, 1.39)] [odds ratio (OR) and 95% CI]. HSM 8 (reduced pincer-type deformity) was inversely associated with RHOA [0.79 (0.70, 0.89)]. HSM 10 (increased cam-type deformity) was positively associated with RHOA [1.21 (1.07, 1.37)]. HSM 3 and HSM 4 (reduced cam-type deformity) were inversely associated with RHOA [0.73 (0.65, 0.83) and 0.82 (0.73, 0.93), respectively]. HSM 3 was inversely related to pain on examination [0.84 (0.76, 0.92)] and walking [0.88, (0.81, 0.95)], and to WOMAC score [0.87 (0.80, 0.93)].ConclusionsDXA-derived measures of hip shape are associated with RHOA, and to a lesser extent hip pain, possibly reflecting their role in the pathogenesis of hip OA.

Highlights

Hip osteoarthritis (OA) is an increasingly important cause of morbidity as the mean age of the population increases[1]
To establish whether Dual-energy X-ray Absorptiometry (DXA)-derived hip shape represents a useful phenotype for future genome wide association studies (GWAS) studies intended to identify novel genetic risk factors for hip OA, in the present study, we aimed to examine whether hip shape derived from a Statistical shape modelling (SSM) applied to hip DXA scans is associated with radiographic hip OA (RHOA) in the Osteoporotic Fractures in Men (MrOS) Study; in this population-based cohort, hip radiographs were performed a mean of 4.6 years following baseline DXA scans
At visit 2, right hip radiographs were read for RHOA, which were available for 4100 (69.9%) of these participants, who formed the basis of the present study, of whom 90.7% were white, 3.3% Asian, 3.2% African American and 2.8% multiracial/unknown/other

Summary

Introduction

Hip osteoarthritis (OA) is an increasingly important cause of morbidity as the mean age of the population increases[1]. One of the most important is abnormalities of hip development leading to alterations in hip shape, exemplified by developmental dysplasia of the hip which is screened for routinely in neonates[2,3]. More subtle alterations in hip shape have been reported to be associated with hip OA. Camtype deformities, caused by extra bone growth around the anterolateral aspect of the femoral headeneck junction resulting in a non-spherical femoral head, leading to femoro-acetabular impingement (FAI)[4], are associated with premature onset of OA5,6. FAI may result from a pincer-type deformity where the acetabulum overhangs and encroaches on the lateral aspect of the

Methods

Results

Discussion

Conclusion