Background Rheumatoid arthritis (RA) is a complex inflammatory disease known to aggregate within families, implicating the crucial role of shared genetic and/or environmental factors in its etiology. Despite the many advances from genome-wide association studies, family history is an easily obtained aggregate of an individual's complete genetic risk, and is accepted to be important in clinical diagnosis.
Objectives Our study is aimed to explore how much of the RA familial risk can be explained by established genetic and environmental risk factors; and to find out whether identified genetic markers enough to explain the burden of familial risk.
Methods We used data from two Swedish population registers and the Epidemiological Investigation of Rheumatoid Arthritis (EIRA) project. First degree relatives (FDRs) of the index person (EIRA cases and controls) were identified through the Swedish Multi-Generation register. The index individual was considered “exposed” to familial risk if any FDRs had a RA diagnosis from the Swedish Patient Register. Analysis was run divided by ACPA status in cases and by RF status in relatives. Established genetic risk factors (45 SNPs according to GWAS catalog and the shared epitope) were selected. Environmental risk factors (smoking, drinking, BMI, socio-economic status, silica exposure, fish consumption) were determined based on previous publications. Due to missingness for individual risk factors, the complete case analysis was followed by Multiple Imputation. All analysis was adjusted for age, gender and residential area.
Results In general, 11.9% of ACPA positive cases had RA affected FDRs, a higher proportion than the 6.5% in ACPA negative cases and 3.5% in controls. We found a stronger aggregation pattern in ACPA/RF positive RA (OR=4.09) than ACPA/RF negative RA (OR=2.30); while the ORs for cross-phenotype familial risk was 3.18 in ACPA positive cases/RF negative relatives, and 1.62 in ACPA negative cases/RF positive relatives. Further, we found that established non-genetic risk factors do not visibly explain anything of the familial risk, while SE and the identified 45 SNPs only explain a small proportion for ACPA positive RA (ACPA+ cases/RF+ relatives: ORCrude=4.09, ORadjust for SE=3.85, OR adjust for 45SNPs=3.66, OR adjust for both=3.55; ACPA+ cases/RF- relatives: ORCrude=3.18, ORadjust for SE=2.45, ORadjust for 45 SNPs=2.29, ORadjust for both=2.07) but not for ACPA negative RA.
Conclusions We found that established genetic and environmental risk factor do not explain much of the familial aggregation of RA, suggesting that most heritability remain to be elucidated. It is likely that a large amount of additional loci with individually extremely small magnitude remain to be identified. In particular for ACPA negative RA, it seems that we have only found factors for a very minor (or no) part of the familial aggregation. This suggests a need for studies focusing specifically on ACPA negative RA, trying to find genetic risk factors or identify disease subtypes with a stronger familial risk.
Disclosure of Interest None declared