Familial correlations of Rheumatoid Arthritis and Autism Spectrum Disorders

Introduction: Autism Spectrum Disorders (ASD) and Rheumatoid Arthritis (RA) share important risky genes, i.e. HLA-DRB1 genes, and environmental risk factors, i.e. preterm. However, only a few studies examined the relationship between ASD and RA and got mixed results.

Objective: To examine the familial correlations between RA and ASD and whether the correlations are genetic origins. Method: We followed a birth cohort born from 1995 to 2015 with Swedish-born parents. The diagnoses of related disorders were obtained from Swedish Patient Register. Children’s parental relatives (mothers, fathers, maternal/paternal aunts&uncles) and peer relatives (full siblings, maternal/paternal half siblings&cousins) were identified by Swedish registries. To examine the familial correlations, we fitted separate cox models for different parental relatives. We compared peer relatives’ Relative Recurrence Risk (RRR) of ASD in two nested cox models: one with maternal RA and one without. All cox models took the length of follow-up time of ASD as outcome, corresponding relatives’ RA as exposure, adjusted by parental age, birth year, parental psychiatric history. To examine the correlation origin, we fitted liability threshold models (ACE model) using full siblings and cousins’ data to compare the change of variance of genetic component of ASD before and after adjusting maternal RA/parental RA. The adjustments applied to ACE model included birth year and sex.

Results: There were a total of 1,471,265 children, of which ASD cases were 28,609(1.94%) and 12,526(0.85%) children had RA mothers. The adjusted hazard ratio was 1.30(95%CI: 1.17-1.44) for maternal RA and was 1.30(95%CI: 1.04–1.62) for maternal aunts. No significant results were found in other parental relatives. RRRs of peer relatives’ ASD and variance estimate of genetic component did not changed after adjusting RA.

Conclusion: No familial correlations were found between RA and ASD, but maternal RA showed as an independent risk factor of children’s ASD. Maternal aunts’ RA could be false significant due to strong correlation to maternal RA. By examining the origins, no familial correlations were confirmed, and the risk of maternal RA is not likely to be genetic origins.