Optimal Cutoff Point of Glutamate Decarboxylase Antibody Titers in Differentiating Two Subtypes of Adult‐Onset Latent Autoimmune Diabetes

Abstract

The optimal cutoff point of glutamate decarboxylase antibody (GAD-Ab) titers for differentiating two latent autoimmune diabetes (LADA) subtypes remains unclear. One hundred and forty-five GAD-Ab-positive patients screened from phenotypic type 2 diabetes were diagnosed as LADA. The clinical features were compared among LADA patients with different GAD-Ab titers. The receiver-operating characteristic (ROC) curve was used to evaluate the diagnostic value of GAD-Ab titers and to define the optimal cutoff point. The heterogeneity of clinical features in LADA could be discriminated by five GAD-Ab titers, with maximal differences at the titer of 175 U/mL. The ROC curve analysis showed that the optimal cutoff point for discriminating two LADA subtypes was at the titer of 175 U/mL, with sensitivity and specificity of 54.5% and 92.1%, respectively. These findings demonstrated that the two clinically distinct subtypes of LADA can be optimally discriminated by the GAD-Ab titers.