Inheritance of the shared epitope and long-term outcomes of rheumatoid arthritis among community-based Caucasian females

• Published in: Genetic epidemiology Vol. 15; no. 1; pp. 61 - 72
• Main Authors: Criswell, Lindsey A., Mu, Hua, Such, Carol L., King, Mary-Claire
• Format: Journal Article
• Language: English
• Published: New York Wiley Subscription Services, Inc., A Wiley Company 1998; Wiley-Liss
• Subjects: Alleles; Arthritis, Rheumatoid - genetics; Arthritis, Rheumatoid - immunology; Biological and medical sciences; California; Community Health Services; Disabled Persons; Diseases of the osteoarticular system; Epitopes - genetics; European Continental Ancestry Group - genetics; Female; Follow-Up Studies; Genetic Heterogeneity; Genotype; heterogeneity; HLA-DRB1 alleles; Humans; Inflammatory joint diseases; long-term outcomes; Medical sciences; Middle Aged; Models, Genetic; Multivariate Analysis; rheumatoid arthritis; shared epitope; Treatment Outcome; California; Human; Immunopathology; HLA-System; Major histocompatibility system; Diseases of the osteoarticular system; Class II histocompatibility antigen; Autoimmune disease; Inflammatory joint disease; HLA-DR-Locus; Chronic; Association; Epidemiologic genetics; Rheumatoid arthritis; Risk factor; Female; Genetics; Inheritance(genetics)
• ISSN: 0741-0395; 1098-2272
• DOI: 10.1002/(SICI)1098-2272(1998)15:1<61::AID-GEPI5>3.0.CO;2-7

Multiple HLA‐DRB1 alleles encoding a shared epitope (SE) at amino acid positions 70–74 are associated with susceptibility and severity of rheumatoid arthritis (RA). We examined the relationship between the number and DRB1 genotype of SE alleles inherited and long‐term outcomes of 180 community‐based, Caucasian female RA patients followed annually for up to 12 years. Outcomes examined were physician assessment of RA course; annual measures of pain, function, and number of painful joint groups; history of joint surgery; and resource utilization. Models accounted for correlation among serial observations for the same patient and adjusted for patient age and disease stage. We examined two genetic models: a SE model in which patients were classified according to the number of SE copies inherited and a genotype model in which patients were categorized into one of six groups based on the inherited DRB1 genotype. We used likelihood ratio tests to compare these genetic models and to compare alternative model specifications. Our results demonstrate strong associations between inheritance of the SE and long‐term outcomes of community‐based Caucasian females with RA. However, the pattern of results is not consistent across the outcomes. An additive model of risk is apparent for history of joint surgery and RA hospitalization. In contrast, a near reversal of this pattern is apparent for function, joint pain, pain rating, and RA physician visits. Finally, although the genotype model did not appear to be a better predictive model for RA outcomes overall, it did reveal some striking heterogeneity of SE alleles that was masked by the more parsimonious SE model. For example, the odds ratio (OR) for joint surgery for patients with 2 SE copies (OR = 3.16) reflects an average of 2 very different ORs when patients are further categorized according to genotype groups 4 and 5 (OR = 1.3 and 11.9, respectively). Genet. Epidemiol. 15:61–72,1998. © 1998 Wiley‐Liss, Inc.