Emily C Somers, Seetha U Monrad, Jeffrey S Warren, Maritsa Solano, Lourdes Schnaas, Mauricio Hernandez-Avila, Martha Maria Tellez-Rojo, Howard Hu

To characterize antinuclear antibody (ANA) prevalence according to distinct assay methodologies in a pediatric cohort from Mexico City, and to further examine associations with age and sex. Serum ANA were measured by indirect immunofluorescence assay (IFA) and multiplex immunoassay in 114 children aged 9-17 years. IFA was considered positive at a cutoff titer of ≥1:80. Agreement between assay methods was assessed by kappa statistic. Sensitivity, specificity, and 95% confidence intervals (CIs) of the multiplex were computed with IFA as the reference standard. Of the 114 children (mean age 14.7 [standard deviation 2.1] years; 54 [47%] female), 18 of 114 (15.8%) were ANA positive by IFA, and 11 of 114 (9.6%) by 11-antigen multiplex assay. ANA prevalence was higher in females compared with males by both of the methods (ratios 1.6-1.9 to 1). Agreement between tests was classified as slight by kappa (κ=0.177 [95% CI -0.051, 0.406]). The multiplex immunoassay had sensitivity of 22.2% (95% CI 6.4, 47.6) and specificity of 92.7% (95% CI 85.6, 97.0), and failed to capture 3 of 4 (75%) of the high-titer (≥1:1280) IFA-positives. Up to 15% of children in this general population cohort were ANA positive, with a higher rate of positivity among females according to both assay methods. Substantial discordance in ANA results was found between IFA and multiplex methods, even for high-titer IFA positives. These findings underscore the need to sufficiently account for assay characteristics when interpreting ANA test results, and support IFA as the more appropriate assay for studies of subclinical autoimmunity.