TJM, JMD, RAO, ATH and KRO discussed and contributed to study design and provided support for the analysis and interpretation of results. secondary care. Participants 1352 (model development) and 582 (external validation) participants diagnosed with diabetes between the age of 18 and 50 years of white Western origin. Main outcome steps Type 1 diabetes was defined by quick insulin requirement (within 3 years of analysis) and severe endogenous insulin deficiency (C-peptide 200?pmol/L). Type 2 diabetes was defined by either a lack of quick insulin requirement or, where insulin treated within 3 years, retained endogenous insulin secretion (C-peptide 600?pmol/L at 5 years diabetes period). Model overall performance was assessed using area under the receiver operating characteristic curve (ROC AUC), and internal and external validation. Results Type 1 diabetes was present in 13% of participants in the development cohort. All five predictor variables were discriminative and self-employed predictors of type 1 diabetes (p 0.001 for those) with individual ROC AUC ranging from 0.82 to 0.85. Model overall performance was high: ROC AUC range 0.90 (95% CI 0.88 to 0.93) (clinical features only) to 0.97 (95% CI 0.96 to 0.98) (all predictors) with low prediction error. Results were consistent in external validation (medical features and Zinc Protoporphyrin GADA ROC AUC 0.93 (0.90 to 0.96)). Conclusions Clinical diagnostic models integrating medical features with biomarkers have high accuracy for identifying type 1 diabetes with quick insulin requirement, and could aid clinicians and experts in accurately identifying individuals with type 1 diabetes. strong class=”kwd-title” Keywords: Type 1 diabetes, Type 2 diabetes, Classification, C-peptide, GADA, IA-2A, Type 1 Diabetes Genetic Risk Score Advantages and limitations of this study Diabetes type is definitely robustly defined using direct measurement of endogenous insulin secretion, an end result closely related to treatment, education and monitoring requirements. A combination of a large development dataset and small number of predictors minimises risk of model overfitting, a common problem with diagnostic models of this nature. Models are robustly internally and externally validated. The cross-sectional nature of the development and validation cohorts means that time to insulin was self-reported and measurement of model predictors was not undertaken at analysis: both body mass index and islet autoantibody prevalence may switch over time. Models have been developed in white Western populations with young adult onset diabetes: further work is required to extend this work to other age groups and ethnicities. Intro Making the correct analysis of type 1 and type 2 diabetes is vital for appropriate management, with recommendations for these conditions recommending very different glucose-lowering treatment and education.1C3 These differences are predominantly driven by the quick development of severe endogenous insulin deficiency in type 1 diabetes.1 This means that individuals with type 1 diabetes need rapid insulin treatment and are at risk of life-threatening ketoacidosis without insulin treatment. They develop a requirement for physiological insulin Zinc Protoporphyrin alternative (eg, multiple injections, carbohydrate counting, pumps) due to the very high glycaemic variability associated with severe insulin deficiency4 5 and Rabbit polyclonal to ZMAT3 have poor glycaemic response to most adjuvant glucose-lowering therapies.6 In contrast, individuals with type 2 diabetes continue to help to make substantial endogenous insulin even many decades after analysis.7 Glycaemia is therefore usually managed initially with way of life switch or oral agents4 8 and, if insulin treatment is needed, a combination of simple insulin regimens and adjuvant non-insulin therapies.4 5 8 9 Correctly distinguishing between diabetes subtypes at analysis is often difficult and misclassification is therefore common.10C12 Current recommendations focus on etiopathological meanings without giving obvious criteria for clinical use.1 13 In clinical practice, clinical features are predominantly used to determine diabetes subtype but only age at analysis and body mass index (BMI) have evidence for power at diabetes onset, whereas other features used by clinicians such as symptoms at analysis, excess weight loss or ketosis do not have an evidence foundation.14 Increasing obesity rates mean that many individuals with type 1 diabetes will be obese and type 2 diabetes is occurring in Zinc Protoporphyrin the young.15 Type 1 Zinc Protoporphyrin diabetes has been recently shown to happen at similar rates in those aged above and below 30 years.16 Therefore, simple cut-offs based on age at analysis and BMI are unlikely to accurately diagnose diabetes type for many individuals.1 10 Similarly, there is no single diagnostic test that can be used to classify diabetes robustly at analysis. While measurement of islet autoantibodies can assist classification, many individuals with type 1 diabetes are islet autoantibody bad and many individuals with the medical phenotype of type 2 diabetes, without quick insulin requirement, are islet autoantibody-positive.17 A Type 1 Diabetes Genetic Risk score (T1D GRS) has been recently shown to assist analysis of diabetes type but this provides imperfect discrimination in isolation.18 To classify diabetes, a suitable gold standard is necessary. As the key factor driving variations in treatment decisions between the two subtypes.