Cindy Boer

22 | Chapter 1.1 Genome-wide osteoarthritis association studies The research field of osteoarthritis was an early adaptor of GWASs and has proven to be relatively successful in finding underlying genetic variations, results of which (for 2013) are summarized in Figure 4. The first GWAS performed for osteoarthritis was done in 2008, and identified a single region/locus of the DNA [47]. As stated before, several hun- dreds of genes are probably involved in osteoarthritis. Each common SNV, as identified in GWAS, will only pose a small risk for osteoarthritis , meaning that probably hundreds if not thousands of SNVs will be associated. Due to these small effect sizes of individual SNVs, very large numbers of osteoarthritis cases and controls are needed for successful and ro- bust SNV discovery through GWAS [41]. For this reason, GWAS researchers are forced to collaborate in order to collect such large sample sizes. The first osteoarthritis collaboration, the TREAT-OA (Translational Research in Eu- rope Applied Technologies for OsteoArthritis) consortium, included 6,709 osteoarthritis cases, and 44,439 controls [48]. This consortium only confirmed the previously found sin- gle genetic locus in 2008, a far cry from the hundreds thought to exist. Although this was a massive effort in 2008, the sample size was still too small for robust GWAS findings. Thus, even larger collaborative efforts were needed. Chapter 4 presents the results from the Genetics of Osteoarthritis (GO) consortium. This is the most recent and largest collabora- tive effort on osteoarthritis genetics, including ~180.000 osteoarthritis cases and 600.000 controls, a ~26 fold increase in sample size from the first collaborative effort. Phenotypes for osteoarthritis Increasing the number of cases and controls involved in a study is not the only way to increase the power to robustly detect associated genetic variants. To reduce the hetero- geneity seen in complex diseases(such as in osteoarthritis), the use of endophenotypes or stratified phenotypes of the disease is possible ( Figure 6 ). The use of such phenotypes to reduce osteoarthritis heterogeneity will be presented in Chapter 2 . Endophenotypes are phenotypes (characteristics, traits) of the disease that are more closely related to the underlying genetics, than to the disease itself [49]. For example, the thickness of cartilage in the joint is considered an endophenotype for osteoarthritis[50]. This is because the amount of articular cartilage is fixed, there is limited replacement of the collagen matrix after skeletal maturity, even with the occurrence of disease. Meaning that the total carti- lage thickness in a joint is largely determined by underlying genetics[51]. Chapter 2.1 presents the results of using cartilage thickness in the hip joint as an endophenotype for hip osteoarthritis. This and other osteoarthritis endophenotypes are depicted in Figure 6 .