Background Rheumatoid factor and several other autoantibodies occur more often among women, but the role of female sex for anti-citrullinated peptide antibodies (ACPA) occurrence and concentration have not previously been explored in general population settings.

Objectives To study the relationship between ACPA (occurrence and concentration) and female sex in a population based cohort.

Methods We used a population based subset of the Swedish Twin register including 12590 individuals (mean age 65 (range 48-93) all of whom had been screened for ACPA as determined by Anti-CCP2 semi-quantitative ELISA (Eurodiagnostica, Malmö; Sweden). A subgroup of the cohort (n=623) was also screened for antibody reactivity to four candidate autoantigens. Further, we used self-reported smoking, imputed HLA-SE, and retrieved information about rheumatoid arthritis (RA) diagnosis from the National Patient Register. Individuals with ACPA but without a registered RA diagnosis were considered as individuals at risk for RA. We analyzed the association between female sex and ACPA concentration with t-tests, and presence of ACPA, and ACPA-positive RA, respectively, by logistic regression modeling with robust standard errors, adjusted for age, sex, smoking, and HLA-SE.

Results The prevalence of positive ACPA overall was 2.8% (350/12590), 1.0% of ACPA positive RA (124/12590), and 1.8% of ACPA-positivity without RA (226/12590). Among those with ACPA in the absence of RA, the mean ACPA concentration was 161 IU/ml (95% CI 105-216) for females versus 247 IU/ml (95% CI 128-367) for males. Among those with ACPA positive RA, the concentrations were 785 IU/ml (95% CI 614-956) for females and 662 IU/ml (95% CI 411-912) for males. These sex differences in concentrations were not statistically significant.

The proportion of females increased from 55% in the entire cohort, through 61% among those with ACPA in the absence of RA, to 71% among those with ACPA positive RA. Female sex was associated with ACPA overall (OR=1.6, 95% CI 1.2-2.0), ACPA positivity in the absence of RA (OR=1.3, 95% CI 1.0-1.8), and ACPA-positive RA (OR=2.1, 95% CI 1.4-3.3). Among those with positive ACPA, the risk of also having RA was higher among females (OR=1.8, 95% CI 1.1-3.1).

Among the candidate auto-antigens, anti-citrullinated alpha-enolase was most frequent (24% positive) in the analyzed subgroup of individuals (n=350 ACPA-positive and 273 ACPA negative) and was associated with female sex (OR=2.0, 95% CI 1.3-3.2).

Conclusions In this population-based cross-sectional study we found that female sex was associated with ACPA positivity, even in the absence of concomitant RA, but not with ACPA concentrations. Among those with ACPA positivity, female sex was associated with presence of RA. We controlled for confounding by several established risk factors but residual confounding may be present, e.g. from socioeconomic factors. Our results suggest that female sex is a risk factor not just for RA development but also for ACPA, and point to the need of further research on risk factors for ACPA and ACPA-positive RA with a differential distribution by sex.

Disclosure of Interest : None declared

DOI 10.1136/annrheumdis-2014-eular.5147