Islet autoantibody (IAb) assays for a general population screening should ideally be high-troughput, low-cost, and offer high-sensitivity and high positive predictive value in a low-risk population. The Autoimmunity Screening for Kids (ASK) study has compared the performance of the ’gold standard’ radiobinding assays (RBAs) and a multiplex electrochemiluminesence assay (ECL) in a large unselected pediatric population. Single IAb by RBA but negative by ECL have been previously shown to be low-affinity and of low predictive value for progression to clinical diabetes. From 2017-2019, ASK screened for IAbs 23,400 Colorado children ages 1-17 y. A high-throughput multiplex ECL assay combining IAbs to insulin, GAD and IA-2 in one single well was used in parallel with standard non-multiplex RBAs. At the initial screening with RBA, 3.0% (697/23400) of children were positive for at least one IAb, including 0.4% (98/23400) for multiple IAbs and 2.6% (599/23400) for a single IAb. Among children with ≥2 IAbs, ECL assay results were 89.8% congruent with RBA. In contrast, only 21.5% (129/599 p<0.0001) of children with single IAb by RBA were ECL positive (i.e., high-affinity). During a median follow-up of 0.5 y (range 0.1-2.5 y), 90% (98/109) of children with a single IAb by both RBA and ECL remained persistently positive and 9.2% of those progressed to ≥2 IAbs. In contrast, 74% (242/328, p=0.0003) of children with a single IAb by RBA but ECL negative remained persistently positive; only 1.8% (p=0.001) of those converted to ≥2 IAbs (half became ECL positive). ASK results suggest that applying the ECL assay as the confirmation test may help to remove nearly 80% of subjects with single IAb from further follow-up and monitoring, with little loss of sensitivity. Alternatively, the multiplex ECL could be used as the primary screening tool to leverage its high throughput and low cost, compared to RBAs.


X. Jia: None. L. He: None. D. Miao: None. K. Waugh: None. B.I. Frohnert: None. A. Steck: None. M. Rewers: None. L. Yu: None. C.R. Geno Rasmussen: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at