Cindy Boer

84 | Chapter 2.1 Materials and Methods Ethics statement The participating studies were approved by the medical ethics committees of all partici- pating centres, and all participants gave their written informed consent before entering the study Discovery GWAS, replication and meta-analysis We conducted genome-wide association studies of mJSW for each cohort of the discov- ery stage: Rotterdam Study I (RS-I), Rotterdam Study II (RS-II), TwinsUK, SOF and MrOS using standardized age-, gender and population stratification (four principal compo- nents) adjusted residuals from linear regression. Cohort description and details of the single GWAS studies are given in Supplementary Text 1 and Supplementary Table 1. The 6 cohorts used in the discovery stage were combined in a joinedmeta-analysis using inverse variance weighting with METAL[39]. Genomic control correction was applied to the standard errors and P-values before meta-analysis. SNPs with a p-value<=5×10 −6 were selected for replication. The top SNPs for each independent locus were taken for replication in seven studies: the Genetics of Osteoarthritis and Lifestyle (GOAL) study, the Chingford study, CHECK (Cohort Hip & Cohort Knee), Genetics osteoARthritis and progression (GARP) study, the Genetics of Generalized Osteoarthritis (GOGO), the John- ston County Osteoarthritis Project (JoCo) and additionally the NottinghamOA case-con- trol study for association with Hip OA (see Supplemental material for detailed infor- mation of the cohorts). Association of the SNPs with mJSW was additionally adjusted for height to test its independence. Secondary analyses included: association of the top SNPs with hip OA using logistic regression analysis (age and gender adjusted and by study centres an/or relatedness when it was pertinent). We used conditional analyses to investigate whether there are any independent signals in the identified associated loci, which were implemented using GCTA-COJO analysis[40]. Phenotype description of minimal Joint Space width (mJSW) The mJSWwas assessed at pelvic radiographs in anterior-posterior position. The mJSW was measured in mm, along a radius from the center of the femoral head, and defined as the shortest distance found from the femoral head to the acetabulum. Within the Rotterdam Study, we used a 0.5 mm graduated magnifying glass laid directly over the radiograph to measure the minimal joint space width of the hip joints[41]. Within SOF and MrOS, a handheld caliper and reticule was used to measured mJSW to the nearest 0.1mm between the acetabular rim and proximal head of the femur[42]. For CHECK, mJSW was measured semi-automatic with the Software tool HOLY[43].