Arthritis & Rheumatism, Volume 60,
October 2009 Abstract Supplement
The 2009 ACR/ARHP Annual Scientific Meeting
Philadelphia October 16-21, 2009.
Admixture Composition in African-Americans and Its Impact On SLE
Molineros1, Julio, Kim1, Xana, Kelly1, Jennifer A., Bruner2, Gail R., Oksenberg3, Jorge R., James1, Judith A., Gilkeson4, GS
Oklahoma Medical Research Foundation, Oklahoma City, OK
Oklahoma Medical Rsrch, Oklahoma City, OK
School of Medicine, University of California, San Francisco, San Francisco, CA
Medical University of South Carolina, Charleston, SC
Oklahoma Medical Research Foundation, OK
Systemic lupus erythematosus (SLE) disproportionately affects individuals with African ancestry with respect to disease prevalence, manifestation, severity and organ damage. It can be hypothesized that genetic factors together with individual ancestry or admixture proportion could explain these disparities. Since recent admixture can create substantial increase of linkage disequilibrium (LD) in magnitude and range, these individuals can provide a unique opportunity for admixture mapping (AM) for SLE. We compared and contrasted (a) admixture dynamics, (b) risk of developing SLE, (c) ancestry-phenotype correlation, (d) effect of admixture on 3 associated genes, IRF5 (rs2004640), TNPO3 (rs2280714), and ITGAM (rs11436709), and (e) feasibility and power of AM, between a high admixed (AA) and a low admixed "Gullah" (GH) population with African ancestry.
To estimate individual admixture and ancestry proportion we used 114 ancestry informative markers (AIM) on 1558 AA (722 cases, 836 controls),and 303 GH (163 cases, 140 controls). We applied structured association test (SAT) and logistic regression adjusted for individual admixture to account for population stratification. The power of AM was estimated analytically in AA and GH.
Both AA and GH were best explained by admixture of European and African ancestral populations. Mean European ancestry was significantly higher in AA than in Gullah (19.5% vs 8%, P=×031). Case and control ancestry proportion was significantly different only in AA (P=2.×05). We also found that African ancestry was a risk factor for SLE only in AA (OR=3.1; P=×05). While 3 previously reported SLE associated variants were also replicated the association with SLE (P <0.05) in GH (rs2004640 P=0.001; rs2280714 P=0.04 and rs1143679 P=0.003), only 2 were associated in AA (rs2004640 P=0.0003; rs2280714 P=0.61 and rs1143679 P=0.001), these results were unchanged even after admixture correction. Haplotype and conditional regression analysis using rs2004640 and rs2280714 (16kb apart) identified these variants might be independently associated with SLE. Ancestry association identified no AIMs consistently associated with SLE in both AA and GH. For AM, as expected, case-only design is more powerful than case-control design. For examples, to detect genetic association (80% power) of an allele with ancestry relative risk of 2.0, we need 1474 GH samples as opposed to half (705) the samples required in AA with an affected only design, and we need 4 times more samples in a case-control design.
The admixture effect on SLE increased with admixture proportion. SLE risk was greater in AA that GH. Except TNPO3, both ITGAM and IRF5 associations were replicated in AA and GH. AA population is more suited for admixture mapping than GH.
To cite this abstract, please use the following information:
Molineros, Julio, Kim, Xana, Kelly, Jennifer A., Bruner, Gail R., Oksenberg, Jorge R., James, Judith A., et al; Admixture Composition in African-Americans and Its Impact On SLE [abstract]. Arthritis Rheum 2009;60 Suppl 10 :755