Construction of Risk Prediction Model of Type 2 Diabetic Kidney Disease Based on Deep Learning.

BACKGRUOUND: This study aimed to develop a diabetic kidney disease (DKD) prediction model using long short term memory (LSTM) neural network and evaluate its performance using accuracy, precision, recall, and area under the curve (AUC) of the receiver operating characteristic (ROC) curve. METHODS: The study identified DKD risk factors through literature review and physician focus group, and collected 7 years of data from 6,040 type 2 diabetes mellitus patients based on the risk factors. Pytorch was used to build the LSTM neural network, with 70% of the data used for training and the other 30% for testing. Three models were established to examine the impact of glycosylated hemoglobin (HbA1c), systolic blood pressure (SBP), and pulse pressure (PP) variabilities on the model's performance. RESULTS: The developed model achieved an accuracy of 83% and an AUC of 0.83. When the risk factor of HbA1c variability, SBP variability, or PP variability was removed one by one, the accuracy of each model was significantly lower than that of the optimal model, with an accuracy of 78% (P<0.001), 79% (P<0.001), and 81% (P<0.001), respectively. The AUC of ROC was also significantly lower for each model, with values of 0.72 (P<0.001), 0.75 (P<0.001), and 0.77 (P<0.05). CONCLUSION: The developed DKD risk predictive model using LSTM neural networks demonstrated high accuracy and AUC value. When HbA1c, SBP, and PP variabilities were added to the model as featured characteristics, the model's performance was greatly improved.

Annual decline in ß-cell function in patients with type 2 diabetes in China.

BACKGROUND: The aim of this study was to investigate the annual decline of ß-cell function correlated with disease duration in patients with type 2 diabetes in China. METHODS: This cross-sectional study included 4792 adults with type 2 diabetes who were recruited from four university hospital diabetes clinics between April 2018 and November 2018. Baseline data were collected from electric medical records. Participants were divided into 21 groups with 1-year diabetes duration interval to assess the decline rate of ß-cell function. Homeostatic model assessment model (HOMA 2) model was applied to assess ß-cell function. Multiple linear regression model was used to evaluate the association between biochemical and clinical variables and ß-cell function. RESULTS: In Chinese patients with type 2 diabetes, ß-cell function declined by 2% annually. Using angiotensin receptor blockade (ARB) (ß = .048; P = .011), metformin (ß = .138; P = .021), or insulin (ß = .142; P = .018) was associated with increased ß-cell function. However, increased BMI (ß = -.215; P = .022), alcohol consumption (ß = -.331; P < .001), haemoglobin A1c (ß = -.104; P = .027), or increased diabetes duration (ß = -.183; P = .003) was significantly and negatively associated with ß-cell function. CONCLUSIONS: We determined that the annual rate of the ß-cell function decline was 2% in patients with type 2 diabetes in China. Moreover, we confirmed a positive relationship between ARB treatment and ß-cell function, while BMI and alcohol consumption were significantly and negatively associated with the ß-cell function.