基于血浆miRNA表达谱构建糖尿病并发MCI预测模型：前瞻性队列研究

申请人前期发现全球36%的糖尿病患者患有轻度认知功能障碍（MCI）。MCI除增加不可逆性痴呆发生的风险外，还可导致一系列不良生存结局。因此，构建预测模型以早期识别高风险人群，从而实现糖尿病并发MCI的一级预防十分重要，但已报道的模型推广价值受限。miRNA通过参与调控细胞生命活动的众多信号转导途径对机体的各种生理和病理过程发挥作用。既往基于小样本普通人群的病例对照研究发现血浆miRNA表达谱与MCI有关，其在预测其它疾病发生发展中的价值也得到了广泛验证，但目前尚不清楚血浆miRNA表达谱是否参与了糖尿病患者MCI的发生。因此，本项目拟采用前瞻性队列研究设计，通过随访前期已建立的认知功能正常的糖尿病患者队列并检测血浆miRNA表达谱，探索并验证潜在预测因素，进一步采用机器学习法构建基于血浆miRNA表达谱的糖尿病并发MCI的预测模型，为实现一级预防提供科学依据，从而最终改善糖尿病患者的预后。

The applicant has found that 36% of diabetes patients suffer from mild cognitive impairment (MCI) globally. MCI can not only increase the risk of irreversible dementia, but also lead to a series of adverse outcomes. Therefore, it is imperative to build a prediction model for early identification of those at high risk for MCI, and then realize the primary prevention of MCI among diabetes patients. However, the generalizability of the only one reported prediction model was limited. It has been noted that miRNAs are involved in numerous signal pathways which regulate cell activities and therefore, play an important role in various physiological and pathological processes. Previous case-control studies with limited sample sizes showed that plasma miRNA expression profile was related to MCI among the general population. In addition, the important role of miRNA profiling in predicting the occurrence and development of diseases has been well established. However, it remains unknown whether plasma miRNA expression profile is involved in the occurrence of MCI among diabetes patients. Accordingly, by following up an established cohort of diabetes patients with normal cognitive function, the current study aims to explore the predictive factors for the occurrence of MCI based on plasma miRNA profiling, and further build a prediction model of MCI among diabetes patients using machine learning methods. This model could provide a scientific foundation for primary prevention measures that could ultimately improve the prognosis of diabetes.