Development and External Validation of a Genetic Risk Score for Pain in Rheumatoid Arthritis

Several single-nucleotide polymorphisms (SNPs) have been associated with chronic pain syndromes. Our objective was to determine whether genetic variants are associated with pain and disease activity in rheumatoid arthritis (RA).

Participants were included from two independent RA cohorts: FORWARD (National Databank for Rheumatic Diseases, training data set) and the Veterans Affairs Rheumatoid Arthritis Registry (VARA; validation data set). Multivariable linear regression was used to estimate the relationship between cross-sectional pain scores and 36 fibromyalgia (FM)-associated SNPs in FORWARD. SNP alleles were summed and weighted by these regression coefficients to generate a genetic risk score (GRS) for pain for each participant in both cohorts. Linear regressions and generalized estimating equations were used to determine the relationship between this GRS, an existing pain intensity GRS, and pain and self-reported disease activity.

The sample comprised 756 participants from FORWARD (mean age 56.8 years, 89.4% female) and 2,176 participants from VARA (mean age 64.3 years, 11.0% female) who had pain and genotyping data. Participants in the validation data set (VARA) with FM GRS in the highest quartile had more baseline pain than those in the lowest quartile (+0.55 [95% confidence interval 0.16-0.93], P = 0.006). This was also true for the existing pain intensity GRS. VARA participants in the highest quartile of both GRS had more pain throughout follow-up and higher disease activity scores.

GRS based on pain-related SNPs were associated with RA pain and disease activity, suggesting that the genetic risk of pain may have clinical impacts in RA, such as the likelihood of achieving remission.