Prediction Of Type 1 Diabetes Research Articles

Genetic Discovery and Risk Prediction for Type 1 Diabetes in Individuals Without High-Risk HLA-DR3/DR4 Haplotypes.

More than 10% of patients with type 1 diabetes (T1D) do not have high-risk HLA-DR3 or -DR4 haplotypes with distinct clinical features, such as later onset and reduced insulin dependence. We aimed to identify genetic drivers of T1D in the absence of DR3/DR4 and improve prediction of T1D risk in these individuals. We performed T1D association and fine-mapping analyses in 12,316 non-DR3/DR4 samples. Next, we performed heterogeneity tests to examine differences in T1D risk variants in individuals without versus those with DR3/DR4 haplotypes. We further assessed genome-wide differences in gene regulatory element and biological pathway enrichments between the non-DR3/DR4 and DR3/DR4 cohorts. Finally, we developed a genetic risk score (GRS) to predict T1D in individuals without DR3/DR4 and compared with an existing T1D GRS. A total of 18 T1D risk variants in non-DR3/DR4 samples were identified. Risk variants at the MHC and multiple other loci genome wide had heterogeneity in effects on T1D dependent on DR3/DR4 status, and non-DR3/DR4 T1D had evidence for a greater polygenic burden. T1D-assocated variants in non-DR3/DR4 were more enriched for regulatory elements and pathways involved in antigen presentation, innate immunity, and β-cells and depleted in T cells compared with DR3/DR4. A non-DR3/DR4 GRS outperformed an existing risk score GRS2 in discriminating non-DR3/DR4 T1D from no diabetes (area under the curve 0.867; P = 7.48 × 10-32) and type 2 diabetes (0.907; P = 4.94 × 10-44). In total, we identified heterogeneity in T1D genetic risk dependent on high-risk HLA-DR3/DR4 haplotype, which uncovers disease mechanisms and enables more accurate prediction of T1D across the HLA background.

A Framework for Prediction of Type II Diabetes through Ensemble Stacking Model

In order to prevent long term complications of diabetes its early diagnosis is crucial. With Increasing advances in Artifical Intelligence (AI) and Machine Learning(ML) researchers are increasingly focusing on using them for early diagnosis of diseases.AI and ML has significant potential for early prediction of type 2 diabetes. In this article we have described a ML based framework for prediction of type 2 diabetes -Improved Ensemble Learning with Dimensionality Reduction Model (IELDR) and discussed its result. An IELDR algorithm is an Auto encoder-based feature extraction method with ensemble learning. The experiments were carried out using the LS_diabetes dataset. LS_diabetes dataset containing 374 records with 35 features related to lifestyle and stress. Accuracy, precision, specificity, sensitivity, f1 score, roc and Mathew correlation coefficient (MCC) were measured. After this results were tested and validated using Diabetes_2019 dataset and PIMA diabetes dataset. The IELDR model showed results in terms accuracy, precision, specificity, sensitivity, f1 score, roc and Mathew correlation coefficient (MCC) of 98.67%, 95.24%, 100%, 98.18%, 97.56%, 99.09% and 0.97 respectively. In comparison with PIMA diabetes dataset, LS_diabetes dataset showed an accuracy, precision, sensitivity, specificity, f1-score,roc and mcc value by 17.96%,13.15% 40.22%,5.59%,28.38%,22.09% and 0.4 respectively. The IELDR model achieved the best result on the LS_diabetes dataset showed an accuracy, sensitivity, roc and mcc value improved by 1.82%, 1.58%, 3.01%and 0.04 % compared to the Diabetes_2019 dataset .This proposed IELDR system predicts the risk of type 2 diabetes in a healthy person based on the person’s current lifestyle pattern. This system can be helpful for early prediction of type2 diabetes.

Genetic association and machine learning improves discovery and prediction of type 1 diabetes.

Type 1 diabetes (T1D) has a large genetic component, and expanded genetic studies of T1D can lead to novel biological and therapeutic discovery and improved risk prediction. In this study, we performed genetic association and fine-mapping analyses in 817,718 European ancestry samples genome-wide and 29,746 samples at the MHC locus, which identified 165 independent risk signals for T1D of which 19 were novel. We used risk variants to train a machine learning model (named T1GRS) to predict T1D, which highly differentiated T1D from non-disease and type 2 diabetes (T2D) in Europeans as well as African Americans at or beyond the level of current standards. We identified extensive non-linear interactions between risk loci in T1GRS, for example between HLA-DQB1*57 and INS, coding and non-coding HLA alleles, and DEXI, INS and other beta cell loci, that provided mechanistic insight and improved risk prediction. T1D individuals formed distinct clusters based on genetic features from T1GRS which had significant differences in age of onset, HbA1c, and renal disease severity. Finally, we provided T1GRS in formats to enhance accessibility of risk prediction to any user and computing environment. Overall, the improved genetic discovery and prediction of T1D will have wide clinical, therapeutic, and research applications.

Genome-Wide Polygenic Risk Score Predicts Incident Type 2 Diabetes in Women With History of Gestational Diabetes.

Women with a history of gestational diabetes mellitus (GDM) are at increased risk of developing type 2 diabetes (T2D). It remains unclear whether genetic information improves prediction of incident T2D in these women. Using five independent cohorts representing four different ancestries (n = 1,895), we investigated whether a genome-wide T2D polygenic risk score (PRS) is associated with increased risk of incident T2D. We also calculated the area under the receiver operating characteristics curve (AUROC) and continuous net reclassification improvement (NRI) following the incorporation of T2D PRS into clinical risk models to assess the diagnostic utility. Among 1,895 women with previous history of GDM, 363 (19.2%) developed T2D in a range of 2 to 30 years. T2D PRS was higher in those who developed T2D (-0.08 vs. 0.31, P = 2.3 × 10-11) and was associated with an increased risk of incident T2D (odds ratio 1.52 per 1-SD increase, 95% CI 1.05-2.21, P = 0.03). In a model that includes age, family history of diabetes, systolic blood pressure, and BMI, the incorporation of PRS led to an increase in AUROC for T2D from 0.71 to 0.74 and an intermediate improvement of NRI (0.32, 95% CI 0.15-0.49, P = 3.0 × 10-4). Although there was variation, a similar trend was observed across study cohorts. In cohorts of GDM women with diverse ancestry, T2D PRS was significantly associated with future development of T2D. A significant but small improvement was observed in AUROC when T2D PRS was integrated into clinical risk models to predict incident T2D.

Predicting type 2 diabetes via machine learning integration of multiple omics from human pancreatic islets

Type 2 diabetes (T2D) is the fastest growing non-infectious disease worldwide. Impaired insulin secretion from pancreatic beta-cells is a hallmark of T2D, but the mechanisms behind this defect are insufficiently characterized. Integrating multiple layers of biomedical information, such as different Omics, may allow more accurate understanding of complex diseases such as T2D. Our aim was to explore and use Machine Learning to integrate multiple sources of biological/molecular information (multiOmics), in our case RNA-sequening, DNA methylation, SNP and phenotypic data from islet donors with T2D and non-diabetic controls. We exploited Machine Learning to perform multiOmics integration of DNA methylation, expression, SNPs, and phenotypes from pancreatic islets of 110 individuals, with ~ 30% being T2D cases. DNA methylation was analyzed using Infinium MethylationEPIC array, expression was analyzed using RNA-sequencing, and SNPs were analyzed using HumanOmniExpress arrays. Supervised linear multiOmics integration via DIABLO based on Partial Least Squares (PLS) achieved an accuracy of 91 ± 15% of T2D prediction with an area under the curve of 0.96 ± 0.08 on the test dataset after cross-validation. Biomarkers identified by this multiOmics integration, including SACS and TXNIP DNA methylation, OPRD1 and RHOT1 expression and a SNP annotated to ANO1, provide novel insights into the interplay between different biological mechanisms contributing to T2D. This Machine Learning approach of multiOmics cross-sectional data from human pancreatic islets achieved a promising accuracy of T2D prediction, which may potentially find broad applications in clinical diagnostics. In addition, it delivered novel candidate biomarkers for T2D and links between them across the different Omics.

Modeling type 1 diabetes progression using machine learning and single-cell transcriptomic measurements in human islets

Type 1 diabetes (T1D) is a chronic condition in which beta cells are destroyed by immune cells. Despite progress in immunotherapies that could delay T1D onset, early detection of autoimmunity remains challenging. Here, we evaluate the utility of machine learning for early prediction of T1D using single-cell analysis of islets. Using gradient-boosting algorithms, we model changes in gene expression of single cells from pancreatic tissues in T1D and non-diabetic organ donors. We assess if mathematical modeling could predict the likelihood of T1D development in non-diabetic autoantibody-positive donors. While most autoantibody-positive donors are predicted to be non-diabetic, select donors with unique gene signatures are classified as T1D. Our strategy also reveals a shared gene signature in distinct T1D-associated models across cell types, suggesting a common effect of the disease on transcriptional outputs of these cells. Our study establishes a precedent for using machine learning in early detection of T1D.

Development and validation of a machine learning model for prediction of type 2 diabetes in patients with mental illness.

Type 2 diabetes (T2D) is approximately twice as common among individuals with mental illness compared with the background population, but may be prevented by early intervention on lifestyle, diet, or pharmacologically. Such prevention relies on identification of those at elevated risk (prediction). The aim of this study was to develop and validate a machine learning model for prediction of T2D among patients with mental illness. The study was based on routine clinical data from electronic health records from the psychiatric services of the Central Denmark Region. A total of 74,880 patients with 1.59 million psychiatric service contacts were included in the analyses. We created 1343 potential predictors from 51 source variables, covering patient-level information on demographics, diagnoses, pharmacological treatment, and laboratory results. T2D was operationalised as HbA1c ≥48 mmol/mol, fasting plasma glucose ≥7.0 mmol/mol, oral glucose tolerance test ≥11.1 mmol/mol or random plasma glucose ≥11.1 mmol/mol. Two machine learning models (XGBoost and regularised logistic regression) were trained to predict T2D based on 85% of the included contacts. The predictive performance of the best performing model was tested on the remaining 15% of the contacts. The XGBoost model detected patients at high risk 2.7 years before T2D, achieving an area under the receiver operating characteristic curve of 0.84. Of the 996 patients developing T2D in the test set, the model issued at least one positive prediction for 305 (31%). A machine learning model can accurately predict development of T2D among patients with mental illness based on routine clinical data from electronic health records. A decision support system based on such a model may inform measures to prevent development of T2D in this high-risk population.

Multi-Feature Map Integrated Attention Model for Early Prediction of Type 2 Diabetes Using Irregular Health Examination Records.

Type 2 diabetes (T2D) is a worldwide chronic disease that is difficult to cure and causes a heavy social burden. Early prediction of T2D can effectively identify high-risk populations and facilitate earlier implementation of appropriate preventive interventions. Various early prediction models for T2D have been proposed. However, these methods do not consider the following factors: 1) health examination records (HER) containing health information before diagnosis; 2) rating information containing clinical knowledge; and 3) local and global information of time-series features. These diagnostically relevant factors need to be considered. It is challenging due to irregular and multivariate time series. In this paper, we propose the multi-feature map integrated attention model (MFMAM) for early diabetes prediction using HER. Specifically, HER is converted into the multi-feature map to capture local and global volatility, as well as the sequence order of high-dimensional features. We concatenate rating indicators to introduce clinical knowledge. In addition, considering missing and temporal patterns, we utilize missing and time embedding to learn the complex transition of health status. We adopt attention mechanisms to capture essential features (channels) and time points (spatial). To evaluate the proposed model, we conducted experiments on real-world long-term HER. The results demonstrated that MFMAM outperformed baseline models on tasks of varying sequence lengths and prediction windows. Moreover, we applied our designs to baseline models, and their performance was considerably improved. The proposed model contributes to the short-term and long-term early prediction of T2D in individuals with varying information richness.

A novel interpretable deep transfer learning combining diverse learnable parameters for improved T2D prediction based on single-cell gene regulatory networks

Accurate deep learning (DL) models to predict type 2 diabetes (T2D) are concerned not only with targeting the discrimination task but also with learning useful feature representation. However, existing DL tools are far from perfect and do not provide appropriate interpretation as a guideline to explain and promote superior performance in the target task. Therefore, we provide an interpretable approach for our presented deep transfer learning (DTL) models to overcome such drawbacks, working as follows. We utilize several pre-trained models including SEResNet152, and SEResNeXT101. Then, we transfer knowledge from pre-trained models via keeping the weights in the convolutional base (i.e., feature extraction part) while modifying the classification part with the use of Adam optimizer to deal with classifying healthy controls and T2D based on single-cell gene regulatory network (SCGRN) images. Another DTL models work in a similar manner but just with keeping weights of the bottom layers in the feature extraction unaltered while updating weights of consecutive layers through training from scratch. Experimental results on the whole 224 SCGRN images using five-fold cross-validation show that our model (TFeSEResNeXT101) achieving the highest average balanced accuracy (BAC) of 0.97 and thereby significantly outperforming the baseline that resulted in an average BAC of 0.86. Moreover, the simulation study demonstrated that the superiority is attributed to the distributional conformance of model weight parameters obtained with Adam optimizer when coupled with weights from a pre-trained model.

Complete HLA genotyping of type 1 diabetes patients and controls from Mali reveals both expected and novel disease associations.

HLA genotyping was performed on 99 type 1 diabetes (T1D) patients and 200 controls from Mali. Next-generation sequencing of the classical HLA-A, -B, -C, -DRB1, -DRB3, -DRB4, -DRB5, -DQA1, -DQB1, -DPA1, and -DPB1 loci revealed strong T1D association for all loci except HLA-C and -DPA1. Class II association is stronger than class I association, with most observed associations predisposing or protective as expected based on previous studies. For example, HLA-DRB1*03:01, HLA-DRB1*09:01, and HLA-DRB1*04:05 predispose for T1D, whereas HLA-DRB1*15:03 is protective. HLA-DPB1*04:02 (OR = 12.73, p = 2.92 × 10-05 ) and HLA-B*27:05 (OR = 21.36, p = 3.72 × 10-05 ) appear highly predisposing, although previous studies involving multiple populations have reported HLA-DPB1*04:02 as T1D-protective and HLA-B*27:05 as neutral. This result may reflect the linkage disequilibrium between alleles on the extended HLA-A*24:02~HLA-B*27:05~HLA-C*02:02~HLA-DRB1*04:05~HLA-DRB4*01:03~HLA-DQB1*02:02~HLA-DQA1*02:01~HLA-DPB1*04:02~HLA-DPA1*01:03 haplotype in this population rather than an effect of either allele itself. Individual amino acid (AA) analyses are consistent with most T1D association attributable to HLA class II rather than class I in this data set. AA-level analyses reveal previously undescribed differences of the HLA-C locus from the HLA-A and HLA-B loci, with more polymorphic positions, spanning a larger portion of the gene. This may reflect additional mechanisms for HLA-C to influence T1D risk, for example, through expression differences or through its role as the dominant ligand for killer cell immunoglobulin-like receptors (KIR). Comparison of these data to those from larger studies and on other populations may facilitate T1D prediction and help elucidate elusive mechanisms of how HLA contributes to T1D risk and autoimmunity.

Validation of genetic risk scores for type 2 diabetes on a Russian population sample from the biobank of the National Medical Research Center for Therapy and Preventive Medicine

Aim. To validate and evaluate the accuracy of 14 genetic risk scores (GRSs) for type 2 diabetes (T2D), created earlier in other countries, using a Russian population sample from the biobank of the National Medical Research Center for Therapy and Preventive Medicine.Material and methods. For genetic analysis, next generation sequencing data was used on a sample from the Russian population (n=1165) based on the biobank collection. The study included 14 GRSs associated with T2D.Results. The study demonstrated that the predictive power of 12 out of 14 GRSs for T2D was replicated in the Russian population. As quality metrics, we used the area under the ROC curve, which for models including only GRS varied from 54,49 to 59,46%, and for models including GRS, sex and age — from 77,56 to 78,75%.Conclusion. For the first time in Russia, a study of 14 T2D GRSs developed on other populations was conducted. Twelve GRSs have been validated and can be used in the future to improve risk prediction and prevention of T2D in Russia.

Understanding Islet Autoantibodies in Prediction of Type 1 Diabetes.

As screening studies and preventive interventions for type 1 diabetes (T1D) advance rapidly, the utility of islet autoantibodies (IAbs) in T1D prediction comes with challenges for early and accurate disease progression prediction. Refining features of IAbs can provide more accurate risk assessment. The advances in islet autoantibodies assay techniques help to screen out islet autoantibodies with high efficiency and high disease specificity. Exploring new islet autoantibodies to neoepitopes/neoantigens remains a hot research field for improving prediction and disease pathogenesis. We will review the recent research progresses of islet autoantibodies to better understand the utility of islet autoantibodies in prediction of T1D.

Type-2 Diabetes prediction with cuttlefish search optimization and hybrid model

In India, there are over 30 million individuals with diabetes and many more who are at risk. Diabetes and related health issues must be detected and treated early to be prevented. This study looks at a person's lifestyle and family history to determine their risk of having diabetes. The suggested hybrid machine learning approach, which combines the Support Vector Machine (SVM) classifier and Naive Bayes (NB) classifier, was used to forecast the probability of Type 2 diabetes. We incorporate Cuttlefish Search Optimization (CFSO) to increase prediction accuracy. After the model has been successfully trained, individuals may evaluate their own risk of developing diabetes. 952 times the information required for the question using an online and offline survey along with 18 questions about your habits, health, and family history. Preprocessing was done using min-max normalization, and feature extraction was done using linear discriminant analysis (LDA). The recommended approach was also examined using the Pima Indian Diabetes database. The performance of the provided approach is proved to be the most accurate for both datasets. The term "Type 2 Diabetes prediction with Cuttlefish Search optimization and a hybrid model" refers to the employment of a particular optimization technique called Cuttlefish Search in conjunction with a hybrid model to forecast a person's chance of getting Type 2 Diabetes.

OGTT Metrics Surpass Continuous Glucose Monitoring Data for T1D Prediction in Multiple-Autoantibody-Positive Individuals.

The value of continuous glucose monitoring (CGM) for monitoring autoantibody (AAB)-positive individuals in clinical trials for progression of type 1 diabetes (T1D) is unknown. Compare CGM with oral glucose tolerance test (OGTT)-based metrics in prediction of T1D. At academic centers, OGTT and CGM data from multiple-AAB relatives were evaluated for associations with T1D diagnosis. Participants were multiple-AAB-positive individuals in a TrialNet Pathway to Prevention (TN01) CGM ancillary study (n = 93). The intervention was CGM for 1 week at baseline, 6 months, and 12 months. Receiver operating characteristic (ROC) curves of CGM and OGTT metrics for prediction of T1D were analyzed. Five of 7 OGTT metrics and 29/48 CGM metrics but not HbA1c differed between those who subsequently did or did not develop T1D. ROC area under the curve (AUC) of individual CGM values ranged from 50% to 69% and increased when adjusted for age and AABs. However, the highest-ranking metrics were derived from OGTT: 4/7 with AUC ∼80%. Compared with adjusted multivariable models using CGM data, OGTT-derived variables, Index60 and DPTRS (Diabetes Prevention Trial-Type 1 Risk Score), had higher discriminative ability (higher ROC AUC and positive predictive value with similar negative predictive value). Every 6-month CGM measures in multiple-AAB-positive individuals are predictive of subsequent T1D, but less so than OGTT-derived variables. CGM may have feasibility advantages and be useful in some settings. However, our data suggest there is insufficient evidence to replace OGTT measures with CGM in the context of clinical trials.

138-OR: Characterization and Prediction of Genetic Risk of T1D in Individuals without HLA-DR3 or HLA-DR4

The Major Histocompatibility (MHC) class II haplotypes HLA-DR3 and HLA-DR4 confer substantial genetic risk for type 1 diabetes (T1D), where 90% of individuals with T1D carry at least one copy of DR3 and/or DR4. We sought to understand whether the remaining 10% of T1D individuals without DR3 or DR4 haplotypes (non-DR3/DR4) have differences in genetic risk and whether current risk scores are effective for these individuals. We tested for T1D association using 12,716 non-DR3/DR4 samples from five European ancestry cohorts using genotypes imputed into the Michigan HLA and TOPMed v2 reference panels and identified independent signals using stepwise conditional analyses. There were 21 signals with significant effects on T1D, including 13 at the MHC locus and 8 genome-wide, of which 5 were novel. Across all T1D signals, 14 lead variants exhibited significant heterogeneity in effects on T1D in the non-DR3/DR4 subset compared to the full T1D population, including HLA-B*39 and the PTPN22 locus. Nearly 50% of non-DR3/DR4 T1D would be misclassified by a current T1D genetic risk score (T1D GRS1), which is weighted heavily by DR3/DR4 tag variants. We developed a new GRS using the 21 independent signals and determined whether this GRS could better discriminate T1D in individuals with non-DR3/DR4 haplotypes. The non-DR3/DR4 T1D GRS effectively differentiated T1D case and control samples (AUC=0.87) including in samples not in the initial discovery set (AUC=0.86), and improved prediction compared to GRS1 (AUC=0.52 using all controls, AUC=0.79 using only non-DR3/DR4 controls). These findings reveal a subset of T1D cases with heterogeneity in genetic risk and which are not as well represented by current risk scores, and argue that risk scores developed within heterogeneous subsets may enable even more accurate prediction of T1D. Disclosure C.Mcgrail: None. K.J.Gaulton: Consultant; Genentech, Inc., Stock/Shareholder; Neurocrine Biosciences, Inc., Vertex Pharmaceuticals Incorporated. Funding National Institutes of Health (DK120429, DK122607)

Anti-Islet Autoantibodies in Type 1 Diabetes.

Anti-islet autoantibodies serve as key markers in immune-mediated type 1 diabetes (T1D) and slowly progressive T1D (SPIDDM), also known as latent autoimmune diabetes in adults (LADA). Autoantibodies to insulin (IAA), glutamic acid decarboxylase (GADA), tyrosine phosphatase-like protein IA-2 (IA-2A), and zinc transporter 8 (ZnT8A) are currently employed in the diagnosis, pathological analysis, and prediction of T1D. GADA can also be detected in non-diabetic patients with autoimmune diseases other than T1D and may not necessarily reflect insulitis. Conversely, IA-2A and ZnT8A serve as surrogate markers of pancreatic β-cell destruction. A combinatorial analysis of these four anti-islet autoantibodies demonstrated that 93-96% of acute-onset T1D and SPIDDM cases were diagnosed as immune-mediated T1D, while the majority of fulminant T1D cases were autoantibody-negative. Evaluating the epitopes and immunoglobulin subclasses of anti-islet autoantibodies help distinguish between diabetes-associated and non-diabetes-associated autoantibodies and is valuable for predicting future insulin deficiency in SPIDDM (LADA) patients. Additionally, GADA in T1D patients with autoimmune thyroid disease reveals the polyclonal expansion of autoantibody epitopes and immunoglobulin subclasses. Recent advancements in anti-islet autoantibody assays include nonradioactive fluid-phase assays and the simultaneous determination of multiple biochemically defined autoantibodies. Developing a high-throughput assay for detecting epitope-specific or immunoglobulin isotype-specific autoantibodies will facilitate a more accurate diagnosis and prediction of autoimmune disorders. The aim of this review is to summarize what is known about the clinical significance of anti-islet autoantibodies in the pathogenesis and diagnosis of T1D.

Expanded feature space-based gradient boosting ensemble learning for risk prediction of type 2 diabetes complications

As a metabolic disease, diabetes is a serious threat to human health, in which type 2 diabetes (T2D) constitutes about 90% of diabetic cases. As the disease progresses, T2D often comes with various complications such as kidney disease and retinopathy, which may cause huge threats to patients’ health. In recent years, complication prediction of T2D has gained wide research interest as early disease detection and management can effectively reduce the risk of death. In this paper, to break through the limitation of the available data and achieve higher overall prediction accuracy, an improved feature space-based gradient boosting regression tree ensemble (IFS-GBRTE) approach is proposed for the risk prediction of developing T2D complications. Specifically, the original feature space is expanded based on existing feature generation theories. Then the new feature space is refined by comparing the total contribution of each feature during the construction of a classification and regression tree (CART) under a cross-validation scheme. Finally, a gradient boosting ensemble algorithm using CART as base learners is utilized for model training. The prediction model is validated on the T2D data provided by the China National Clinical Medical Science Data Center. The experimental results show that the proposed IFS-GBRTE achieves 82.49% accuracy and realizes better generalization ability than compared single models and ensemble learning models, meanwhile it is helpful to obtain more effective predictive variables for improving the accuracy of risk prediction of T2D complications, which is of great significance for achieving the early prevention, screening and nursing of complications, and further reducing mortality and save medical resources.

Effective assay technologies fit for large-scale population screening of type 1 diabetes.

While worldwide prevention efforts for type 1 diabetes (T1D) are underway to abrogate or slow progression to diabetes, mass screening of islet autoantibodies (IAbs) in the general population is urgently needed. IAbs, the most reliable biomarkers, play an essential role in prediction and clinical diagnosis of T1D. Through laboratory proficiency programs and harmonization efforts, a radio-binding assay (RBA) has been well established as the current 'gold' standard assay for all four IAbs. However, in view of the need for large-scale screening in the non-diabetic population, RBA consistently faces two fundamental challenges, cost-efficiency and disease specificity. While all four IAbs are important for disease prediction, the RBA platform, with a separate IAb test format is laborious, inefficient and expensive. Furthermore, the majority of IAb positivity in screening, especially from individuals with single IAb were found to be low risk with low affinity. It is well documented from multiple clinical studies that IAbs with low affinity are low risk with less or no disease relevance. At present, two non-radioactive multiplex assays, a 3-assay ELISA combining three IAbs and a multiplex ECL assay combining all four IAbs, have been successfully used as the primary methods for general population screenings in Germany and the US, respectively. Recently, the TrialNet Pathway to Prevention study has been organizing an IAb workshop which aims to analyze the 5-year T1D predictive values of IAbs. A T1D-specific assay with high efficiency, low cost and requiring low volume of sample will definitely be necessary to benefit general population screening.

Prediction of type 1 diabetes with machine learning algorithms based on FTIR spectral data in peripheral blood mononuclear cells.

The incidence of autoimmunity is increasing, to ensure timely and comprehensive treatment, there must be a diagnostic method or markers that would be available to the general public. Fourier-transform infrared spectroscopy (FTIR) is a relatively inexpensive and accurate method for determining metabolic fingerprint. The metabolism, molecular composition and function of blood cells vary according to individual physiological and pathological conditions. Thus, by obtaining autoimmune disease-specific metabolic fingerprint markers in peripheral blood mononuclear cells (PBMC) and subsequently using machine learning algorithms, it might be possible to create a tool that will allow the diagnosis of autoimmune diseases. In this preliminary study, it was found that the peak shift at 1545 cm-1 could be considered specific for autoimmune disease type 1 diabetes (T1D), while the shifts at 1070 and 1417 cm-1 could be more attributed to the autoimmune condition per se. The prediction of T1D, despite the small number of participants in the study, showed an inverse AUC = 0.33 ± 0.096, n = 15, indicating a stable trend in the prediction of T1D based on FTIR metabolic fingerprint data in the PBMC.

The modification of individual factors on association between serum 25(OH)D and incident type 2 diabetes: Results from a prospective cohort study.

Several epidemiological studies have suggested an association between low vitamin D status and increased risk for type 2 diabetes (T2D). This study aimed to explore the dose-response relationship of serum 25-hydroxyvitamin D [25(OH)D] concentrations with incident T2D and the interaction between serum 25(OH)D with individual factors on T2D risk. A total of 1,926 adults without diabetes (mean age: 52.08 ± 13.82 years; 42% men) were prospectively followed for 36 months. Cox proportional hazards model and restricted cubic spline analysis were performed to assess the association and dose-response relationship between serum 25(OH)D and T2D incidence. Both additive and multiplicative interactions were calculated between serum 25(OH)D and individual factors. The net reclassification index (NRI) was used to evaluate the improvement of risk prediction of T2D by adding serum 25(OH)D to traditional risk factors. There were 114 new T2D cases over a mean follow-up of 36 months. Serum 25(OH)D was not associated with T2D incidence, and no significant dose-response relationship was found in the total population. However, stratified analyses suggested a non-linear inverse relationship among individuals with baseline fasting plasma glucose (FPG) <5.6 mmol/L (P overall = 0.061, P non-linear = 0.048). And a significant multiplicative interaction was observed between serum 25(OH)D and FPG on T2D risk (P = 0.005). In addition, we found a significant additive interaction of low serum 25(OH)D with older age (RERI = 0.897, 95% CI: 0.080-1.714; AP = 0.468, 95% CI: 0.054-0.881), male (AP = 0.441, 95% CI: 0.010-0.871), and insufficient physical activity (RERI = 0.875, 95% CI: 0.204-1.545; AP = 0.575, 95% CI: 0.039-1.111) on T2D risk. Significant additive interactions were also observed between vitamin D deficiency/insufficiency with male, overweight/obesity, and insufficient physical activity on T2D risk. Moreover, adding low serum 25(OH)D to a model containing established risk factors yielded significant improvements in the risk reclassification of T2D (NRI = 0.205, 95% CI: 0.019-0.391). Our results indicated a non-linear relationship of serum 25(OH)D concentrations with T2D risk among individuals with normal FPG and additive interactions of serum 25(OH)D with gender, overweight/obesity, and physical activity on T2D risk, suggesting the importance of outdoor exercise.