Clinically Relevant Circulating Protein Biomarkers for Type 1 Diabetes: Evidence From a Two-Sample Mendelian Randomization Study

OBJECTIVE To identify circulating proteins influencing type 1 diabetes susceptibility using Mendelian randomization (MR). RESEARCH DESIGN AND METHODS We used a large-scale two-sample MR study, using cis genetic determinants (protein quantitative trait loci [pQTL]) of up to 1,611 circulating proteins from five large genome-wide association studies, to screen for causal associations of these proteins with type 1 diabetes risk in 9,684 case subjects with type 1 diabetes and 15,743 control subjects. Further, pleiotropy-robust MR methods were used in sensitivity analyses using both cis and trans-pQTL. RESULTS We found that a genetically predicted SD increase in signal regulatory protein gamma (SIRPG) level was associated with increased risk of type 1 diabetes risk (MR odds ratio [OR] 1.66 [95% 1.36–2.03]; P = 7.1 × 10−7). The risk of type 1 diabetes increased almost twofold per genetically predicted standard deviation (SD) increase in interleukin-27 Epstein-Barr virus–induced 3 (IL27-EBI3) protein levels (MR OR 1.97 [95% CI 1.48–2.62]; P = 3.7 × 10−6). However, an SD increase in chymotrypsinogen B1 (CTRB1) was associated with decreased risk of type 1 diabetes (MR OR 0.84 [95% CI 0.77–0.90]; P = 6.1 × 10−6). Sensitivity analyses using MR methods testing for pleiotropy while including trans-pQTL showed similar results. While the MR-Egger suggested no pleotropic effect (P value MR-Egger intercept = 0.31), there was evidence of pleiotropy in MR-PRESSO (P value global test = 0.006). CONCLUSIONS We identified three novel circulating protein biomarkers associated with type 1 diabetes risk using an MR approach. These biomarkers are promising targets for development of drugs and/or of screening tools for early prediction of type 1 diabetes.