Objective: To investigate a high prevalence of systemic sclerosis (SSc; scleroderma) in a well-defined population of 21,255 Choctaw Indians residing in 8 southeastern Oklahoma counties who were "users" of Indian Health Services.
Methods: A case-control study of 12 SSc cases and 48 matched non-SSc controls (4 per case) was conducted to investigate potential occupational, residential, and infectious exposures, as well as genetic factors which might predispose to SSc. HLA class II alleles were determined by DNA oligotyping, and class I and III alleles were defined serologically.
Results: The prevalence of SSc in full-blooded Choctaws was at least 8/1,704, or 469/100,000 (95% confidence interval [95% CI] 203-930) over the 4-year interval 1990-1994 and was significantly higher than that among non-full-blooded Choctaws (6/19,551, or 31/100,000) (P = 0.00001, odds ratio [OR] = 15.4, 95% CI 4.9-49.8). The overall prevalence of SSc in Oklahoma Choctaws (66/100,000) also was significantly higher than that in other Native Americans in Oklahoma (9.5/100,000) (P = 10(-6), OR = 6.95, 95% CI 3.3-13.7), who showed a prevalence similar to that reported for whites (2.1-25.3/100,000). Among the SSc cases, there was striking homogeneity of disease expression with the majority exhibiting diffuse scleroderma, pulmonary fibrosis, and autoantibodies to topoisomerase I. No environmental exposures were found to be in excess among cases versus controls. The strongest risk factor for SSc in cases (100%) versus controls (54%) was an HLA haplotype bearing the alleles B35, Cw4, DRB1*1602 (DR2), DQA1*0501, and DQB1*0301 (DQ7) (P = 0.002, Pcorr = 0.036, OR = 21, 95% CI 2.9-437). Survey of another group of Choctaws residing in another state revealed no cases of SSc despite a high frequency of the same HLA haplotype.
Conclusion: Full-blooded Choctaw Native Americans living in southeastern Oklahoma have the highest prevalence of SSc yet found in any population. A major risk factor for disease is a uniquely Amerindian HLA haplotype; however, additional genes and/or an as-yet-unidentified environmental exposure seem likely.