Scholarlink

<p>Cognitive impairment is a&#xa0;frequent and disabling complication of cerebral small vessel disease (CSVD). This study aimed to identify risk factors associated with cognitive impairment in CSVD patients and to develop predictive models for clinical application. We retrospectively analyzed 482 patients with CSVD treated at our hospital between January 2021 and August 2024. Patients were randomly divided into training (<i>n</i> = 337) and test (<i>n</i> = 145) sets. Multivariate logistic regression identified hypertension, elevated HbA1c (≥ 6.2%), homocysteine (Hcy ≥ 21.1 μmol/L), cystatin&#xa0;C (Cys‑C ≥ 0.9 mg/L), lacunar infarction, and Fazekas score ≥ 3 as independent risk factors for cognitive impairment (all <i>P</i> &lt; 0.05). Based on these variables, we developed logistic regression and random forest prediction models. The logistic regression model achieved area under the curve (AUC) values of 0.851 (training set) and 0.801 (test set), while the random forest model showed AUCs of 0.882 and 0.818, respectively. Both models demonstrated good calibration and predictive performance. These findings indicate that combining clinical, laboratory, and imaging parameters allows for accurate identification of CSVD patients at high risk for cognitive impairment, facilitating early intervention and optimized management.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Identifying risk factors and developing predictive models for cognitive impairment in cerebral small vessel disease

  • 期刊论文

  • 发表日期：

    2024年12月1日

查看全文
  • Xiaoyang Wang,
  • Xianmei Wu,
  • Yanchun Jiang,
  • Yanbo Zhang,
  • Hongmei Lin

摘要

Cognitive impairment is a frequent and disabling complication of cerebral small vessel disease (CSVD). This study aimed to identify risk factors associated with cognitive impairment in CSVD patients and to develop predictive models for clinical application. We retrospectively analyzed 482 patients with CSVD treated at our hospital between January 2021 and August 2024. Patients were randomly divided into training (n = 337) and test (n = 145) sets. Multivariate logistic regression identified hypertension, elevated HbA1c (≥ 6.2%), homocysteine (Hcy ≥ 21.1 μmol/L), cystatin C (Cys‑C ≥ 0.9 mg/L), lacunar infarction, and Fazekas score ≥ 3 as independent risk factors for cognitive impairment (all P < 0.05). Based on these variables, we developed logistic regression and random forest prediction models. The logistic regression model achieved area under the curve (AUC) values of 0.851 (training set) and 0.801 (test set), while the random forest model showed AUCs of 0.882 and 0.818, respectively. Both models demonstrated good calibration and predictive performance. These findings indicate that combining clinical, laboratory, and imaging parameters allows for accurate identification of CSVD patients at high risk for cognitive impairment, facilitating early intervention and optimized management.