IDDF2019-ABS-0253 Explained variance and predictability of inflammatory bowel diseases by genetic risk score in five asian populations (results from the international IBD genetics consortium). (June 2019)

• Record Type: Journal Article
• Authors: Abedian, Shifteh; Wong, Sunny H; Sommeren, Suzanne Van; Takahashi, Atsushi; Cheon, Jae Hee; sood, Ajit; Vahedi, Homayoon; Yamazaki, Keiko; Kim, Won Ho; BK, Thelma; Daryani, Nasser E; Kubo, Michiaki; Yang, Suk-Kyun; Banerjee, Rupa; Malekzadeh, Reza; Weersma, Rinse K; Ng, Siew C; Alizadeh, Behrooz Z
• Is Part Of: Gut. Volume 68(2019)Supplement 1
• Page Start: A110
• Page End: A110
• Publication Date: 2019-06
• Subjects: Gastroenterology -- Periodicals; 616.33
• DOI: 10.1136/gutjnl-2019-IDDFAbstracts.211
• Languages: English
• ISSNs: 0017-5749

Abstract

Abstract : Background: In the absence of properly designed studies, the clinical implication of genetic findings in Inflammatory Bowel Disease (IBD) is a matter of persistent debate especially in Asian population where the prevalence of IBD including Crohn's Disease (CD) and Ulcerative Colitis (UC) is rising. We aimed to investigate the predictability of IBD, CD, and UC by the means of Genetic Risk Score (GRS), in yet unaffected high-risk individuals from East Asia (EA) and Central Asia (CA). Methods: This present study included 9, 698 subjects, consisting of 2, 003 CD, 2, 730 UC and 4, 965 countries, age and gender-matched controls, genotyped on the Immunochip array of three EA (Japan, South-Korea and China) and two CA countries (India and Iran). We generated a multi-locus GRS for each population by combining information from up to 201 known genome-wide significant IBD associated variants to summarize a total load of genetic risk for each phenotype. We estimated explained variance and predictability of IBD, CD, and UC by GRS. We shuffled the EA data into: training set including two out of the three EA populations to build a model to calculate the odds ratio (OR) for each IBD variants, and a test set including the third population for the validation of predictive model built in the training set. For Indian and Iranian populations, we used the previously estimated ORs for the Caucasian population, to build GRS and test the predictive model in these two populations. Results: GRS of IBD could significantly explain up to 4.40% and 4.14% of IBD variance in EA and CA populations but given a prevalence of 0.01% and 0.04% for IBD it yields to a negligible predictive probability up to 8.8x10 -4 and 5.52x10 -4 . GRS of CD and UC could significantly explain CD and UC to a lesser extent compared to IBD given a lower prevalence of CD and UC (figure 1 ). Conclusions: The present study shows that genetic findings based on Trans-ethnic analyses are applicable across Asian populations. GRS alone can explain a limited percentage of disease occurrence in general population (<5% of disease susceptibility) and may not predict IBD in the Asian populations.