• Mingyu Hao, Xiaohong Huang, Xueting Liu, Xiaokang Fang, Haiyan Li, Lingbo Lv, Liming Zhou, Tiecheng Guo, and Dewen Yan.
• Department of Endocrinology, Shenzhen Clinical Research Center for Metabolic Diseases, Shenzhen Second People's Hospital, the First Affiliated Hospital of Shenzhen University, Health Science Center of Shenzhen University, Shenzhen, China.
• Ann. Med. 2023 Dec 1; 55 (1): 766777766-777.

ObjectiveDiabetes mellitus complicated with heart failure has high mortality and morbidity, but no reliable diagnoses and treatments are available. This study aimed to develop and verify a new model nomogram based on clinical parameters to predict diastolic cardiac dysfunction in patients with Type 2 diabetes mellitus (T2DM).Methods3030 patients with T2DM underwent Doppler echocardiography at the First Affiliated Hospital of Shenzhen University between January 2014 and December 2021. The patients were divided into the training dataset (n = 1701) and the verification dataset (n = 1329). In this study, a predictive diastolic cardiac dysfunction nomogram is developed using multivariable logical regression analysis, which contains the candidates selected in a minor absolute shrinkage and selection operator regression model. Discrimination in the prediction model was assessed using the area under the receiver operating characteristic curve (AUC-ROC). The calibration curve was applied to evaluate the calibration of the alignment nomogram, and the clinical decision curve was used to determine the clinical practicability of the alignment map. The verification dataset was used to evaluate the prediction model's performance.ResultsA multivariable model that included age, body mass index (BMI), triglyceride (TG), creatine phosphokinase isoenzyme (CK-MB), serum sodium (Na), and urinary albumin/creatinine ratio (UACR) was presented as the nomogram. We obtained the model for estimating diastolic cardiac dysfunction in patients with T2DM. The AUC-ROC of the training dataset in our model was 0.8307, with 95% CI of 0.8109-0.8505. Similar to the results obtained with the training dataset, the AUC-ROC of the verification dataset in our model was 0.8083, with 95% CI of 0.7843-0.8324, thus demonstrating robust. The function of the predictive model was as follows: Diastolic Dysfunction = -4.41303 + 0.14100*Age(year)+0.10491*BMI (kg/m2) +0.12902*TG (mmol/L) +0.03970*CK-MB (ng/mL) -0.03988*Na(mmol/L) +0.65395 * (UACR > 30 mg/g) + 1.10837 * (UACR > 300 mg/g). The calibration plot diagram of predicted probabilities against observed DCM rates indicated excellent concordance. Decision curve analysis demonstrated that the novel nomogram was clinically useful.ConclusionDiastolic cardiac dysfunction in patients with T2DM can be predicted by clinical parameters. Our prediction model may represent an effective tool for large-scale epidemiological study of diastolic cardiac dysfunction in T2DM patients and provide a reliable method for early screening of T2DM patients with cardiac complications.KEY MESSAGESThis study used clinical parameters to predict diastolic cardiac dysfunction in patients with T2DM. This study established a nomogram for predicting diastolic cardiac dysfunction by multivariate logical regression analysis. Our predictive model can be used as an effective tool for large-scale epidemiological study of diastolic cardiac dysfunction in patients with T2DM and provides a reliable method for early screening of cardiac complications in patients with T2DM.

12 keywords...

hide…