The present study aims to screen and evaluate the early clinical predictors for type 2 diabetes mellitus (T2DM) patients with mild cognitive impairment (MCI) and dementia in Hunan province. A cross-sectional study was conducted from May 2023 to October 2023 to collect data on long-term T2DM patients who settled in Hunan province and were treated in the Department of Geriatrology at Xiangya Hospital of Central South University. The patients were grouped according to the Montreal Cognitive Assessment (MoCA) scale. Basic patient information and multiple serum markers were collected, and differences between groups were compared using one-way ANOVA or Kruskal-Wallis (KW) tests. The multivariate logistic regression analysis was utilized to assess risk factors and Nomogram models were constructed. The logistic regression analysis showed that years of education and serum levels of 1, 5-AG were related factors for the progression of T2DM to T2DM with MCI, and body weight, years of education and FPN levels affected the progression of T2DM with MCI to T2DM with dementia. Based on this, two Nomogram risk prediction models were established. The area under the curve (AUC) of the Nomogram model predicting T2DM progression to T2DM combined with MCI was 0.741, and the AUC of the Nomogram model predicting T2DM combined with MCI progression to T2DM combined with dementia was 0.734. The calibration curves (DCA) of the two models in the training and validation sets were symmetrically distributed near the diagonal line, indicating that the models in the training and validation sets could match each other. In summary, body weight, years of education, and serum HDL-3, FPN, and 1, 5-AG levels are associated with the development of MCI and dementia in T2DM patients. The Nomogram models constructed based on these factors can predict the risk of MCI and dementia in T2DM patients, providing a basis for clinical decision-making.
本研究旨在筛选及评估湖南地区2型糖尿病（T2DM）合并轻度认知障碍（MCI）和痴呆的早期临床预测指标。采用横断面研究，于2023年5至10月收集中南大学湘雅医院老年内分泌科进行治疗的长期定居湖南的T2DM患者数据，根据蒙特利尔认知量表（MoCA）分组。收集患者基本资料及多项血清指标，利用单因素ANANO或KW检验比较各组间差异，通过logistic回归分析相关因素，并构建Nomogram模型。多因素logistic回归分析分别显示，受教育年限、血清1，5-AG水平是T2DM进展为T2DM合并MCI的相关因素，体重、受教育年限及FPN水平影响T2DM合并MCI进展为T2DM合并痴呆的相关因素。据此分别建立的2个Nomogram风险预测模型，预测T2DM进展为T2DM合并MCI的Nomogram模型的曲线下面积（AUC）为0.741，T2DM合并MCI进展为T2DM合并痴呆的Nomogram模型的AUC为0.734。两模型在训练集和验证集中的校准曲线（DCA）均对称分布在对角线附近，表明模型在训练集与验证集中相互符合。综上，体重、受教育年限、血清HDL-3、FPN及1，5-AG水平是T2DM患者发生MCI及痴呆的相关因素，基于这些因素构建的Nomogram模型在预测T2DM患者发生MCI及痴呆风险具有一定的临床应用价值。.