Skip to main content
. 2017 Jun 9;76(7):1199–1206. doi: 10.1136/annrheumdis-2016-210373

Table 3.

Power gains associated with dichotomous versus specific radiographic endophenotype descriptors

Study parameters
Estimated parameters
SNP Definition of cases Definition of controls N cases N controls EAF OR Power (%) N cases N controls N total
rs754106 Hip OA with superior JSN Hip OA with non-superior JSN 1265 552 0.54 0.72 18.45 2416 1054 3470
Hip OA with superior JSN Population-based 1265 6500 0.54 0.91 0.05 10 669 54 821 65 490
Hip OA Population-based 1817 6500 0.54 1.01 0.00 >1 000 000 >3 577 325 >4 577 325
rs73023563 Hip OA with superior JSN Hip OA with non-superior JSN 1265 552 0.19 1.63 57.78 1571 686 2257
Hip OA with superior JSN Population-based 1265 6500 0.20 1.07 0.00 31 513 161 925 193 438
Hip OA Population-based 1817 6500 0.19 0.94 0.00 43 995 157 384 201 379
rs17050727 Hip OA with axial/medial JSN Hip OA with non-axial/medial JSN 425 1392 0.14 1.65 28.62 705 2309 3014
Hip OA with axial/medial JSN Population-based 425 6500 0.14 1.43 4.89 1168 17 864 19 032
Hip OA Population-based 1817 6500 0.14 1.00 0.00
rs6766414 Hip OA with hypertrophic response Hip OA with non-hypertrophic response 531 1286 0.24 1.42 12.40 1139 2758 3897
Hip OA with hypertrophic response Population-based 531 6500 0.24 1.25 0.92 2199 26 918 29 117
Hip OA Population-based 1817 6500 0.24 1.00 0.00
rs61837881 Hip OA with hypertrophic response Hip OA with non-hypertrophic response 531 1286 0.24 1.39 7.69 1298 3144 4442
Hip OA with hypertrophic response Population-based 531 6500 0.24 1.33 7.15 1323 16 195 17 518
Hip OA Population-based 1817 6500 0.24 1.06 0.00 40 187 143 762 183 949
rs16869403 Hip OA with atrophic response Hip OA with non-atrophic response 267 1550 0.10 2.11 58.36 330 1916 2246
Hip OA with atrophic response Population-based 267 6500 0.10 1.93 42.25 383 9324 9707
Hip OA Population-based 1817 6500 0.10 1.05 0.00 115 540 413 324 528 864

For each variant and stratum we present power (%) based on study-specific effect size, effect allele frequency and case/control ratio from the GWAS on Illumina610k. Similarly we estimated the sample size required to have 80% power to detect the associated signals with genome-wide significance (p=5×10−8).

EAF, effect allele frequency; GWAS, genome-wide association studies; JSN, joint space narrowing; OA, osteoarthritis; SNP, single nucleotide polymorphism.