Participants
Individuals over the age of 45 years were recruited between June 2011 and October 2014 through advertisements, word of mouth, and hospital orthopaedic waiting lists. Volunteers were screened using the Harris Hip Score (HHS) (18) and radiographic examination. From 420 individuals who volunteered, 60 individuals with hip pain were eligible to complete the HHS and subsequently undergo radiographic screening evaluation for possible participation in the study. All participants were of Caucasian background. Weight-bearing anterior-posterior radiographs of the pelvis and hips were performed with feet internally rotated by 15+5 degrees (19). An experienced radiologist electronically scored radiographs for both hips based on the Kellgren-Lawrence grades, and the presence of osteophytes using the Osteoarthritis Research Society International grading criteria (20), and electronically measured hip joint space width at the supero-medial, apical and supero-lateral regions.
Inclusion and exclusion criteria
Participants with hip pain in the last 3 months, HHS <95 points, Kellgren-Lawrence grade 2 or 3 and/or joint space width <3 mm in one or both hips were defined as having hip OA (n=27). Individuals with Kellgren-Lawrence grade 4 and joint space width <1 mm or any major lower limb musculoskeletal or neurological conditions besides hip OA were excluded.
Data for the affected unilateral or most affected bilateral limb of individuals with hip OA were used for statistical analysis (12). The most affected limb in those with bilateral hip OA was determined by the least joint space width. Ethical approval was granted by Griffith University Human Research Ethics Committee (GU Ref No: PES/23/08/HREC), Queensland Health, Health and Medical Research Human Research Ethics Committee (HREC/13/QPAH/207), and Monash University Human Research Ethics Committee (10754), and all participants provided written informed consent prior to commencement of the study.
Anthropometric measures
Height was measured using a stadiometer with the removal of footwear. Weight was measured via an electronic scale with the removal of footwear and heavy clothing. Body mass index (BMI) was calculated.
Hip pain and function
Hip pain and function were evaluated using the validated Hip Disability and Osteoarthritis Outcome Score (HOOS) (21). The HOOS is composed of 40 items and assesses patient-relevant outcomes in five subscales: pain (10 items), symptoms (5 items), activity of daily living (17 items), sport and recreation function (4 items), and hip-related quality of life (4 items). Each question is scored from 0 to 4 (5 Likert boxes). For each subscale, the scores are summarized and transformed into a worst to best scale ranging from 0-100 scale. Higher scores refer to better outcome: 0 representing extreme problems and 100 representing no problems.
Hip muscle CSA and fat infiltration
Participants underwent magnetic resonance imaging (MRI) of the pelvis and leg (starting from above iliac crest down to knee, bilateral) using a 3.0 T MRI unit (Phillips Healthcare Ingenia). Participants were positioned in supine position with body coil arrays superiorly placed on lower limbs and legs in 15° of hip internal rotation, secured together with a strap. Hip muscle CSA was measured on axial images obtained using a T1 weighted 2-dimensional fast spin echo sequence (repetition time 731.6 msec, echo time 6.5 msec, flip angle 90°, slice thickness 10 mm, pixel matrix 0.47mm × 0.47mm, and 960 × 960 matrix). The CSA of hip muscles was measured from five regions (Figure 1), adapted from a previous study (22): (A) Iliac crest: psoas major; (B) Upper border of the acetabulum: gluteus maximus, gluteus medius and gluteus minimus; (C) Lower border of the acetabulum: obturator internus; (D) Ischial tuberosity: obturator externus; (E) Just below the gluteus maximus muscle: adductor longus, adductor magnus, and rectus femoris. Muscle CSA was measured by tracing the border of each muscle using the software Osirix (University Hospital of Geneva, Geneva, Switzerland) on an independent workstation. A reader, trained by a radiologist, who was blinded to the participant characteristics, hip pain and function, measured the hip muscle CSA twice with one week interval, and the average was taken as the muscle CSA. The intra-observer reproducibility (intra-class correlation coefficient) ranged from 0.78 to 1.00. Fat infiltration of hip muscles was measured on axial images and categorised into grade 0: no fat infiltration, grade 1: 1-10% fat infiltration, grade 2: 11-50% fat infiltration, and grade 3: >50% fat infiltration. The intra-observer reproducibility (intra-class correlation coefficient) of our measurement was 0.99.
Statistical analysis
Participant characteristics were tabulated. Multiple linear regression was used to examine the associations of hip muscle CSA with hip pain and function (HOOS scores) of the target hip, adjusted for age and gender. A p-value of less than 0.05 (two-tailed) was considered statistically significant. All statistical analyses were performed using the IBM Statistical Package for the Social Sciences (SPSS, Chicago, IL) software, version 24.