数据日期 2026-10-29 · 生成时间 2026-10-29 06:03
研究问题
We aimed to conduct a meta-analysis evaluating the various available prediction scales, including their predictive performance in short, medium, and long follow-up durations and in middle-aged and older age groups.
方法
Predictive performance measures included AUC (area under the receiver operating characteristic curve), sensitivity, specificity, and accuracy.；We included all observational studies with con- secutive enrollment and follow-up.；The com- ponents and scoring systems of the three risk scores included in the meta-analysis are presented in Online Resource 1, Appendices 7–9.
核心结论
Four [31–33] cohorts AUC estimates were included in the short follow-up subgroup, with a pooled AUC of 0.71 (95% CI 0.65–0.77).；Eight [31–36] were included in the medium follow-up subgroup, with a pooled AUC of 0.65 (95% CI 0.61–0.69).；The pooled AUC was 0.63 (95% CI 0.54–0.72) in the medium follow-up/middle-aged subgroup [33–35, 37], 0.55 (95% CI 0.52–0.57) in the medium fol- low-up/older subgroup [32, 34], and 0.74 (95% CI 0.73–0.76) in the long follow-up/middle-aged sub- group [34, 38–41]. ‖ 结论：Therefore, the choice of risk score should consider both the patient’s age and the intended pre- diction period.
局限
On the other hand, we did not find enough stud- ies to analyze all subgroups and scores (e.g., LIBRA with long follow-up), and the heterogeneity was high in some analyses.；The high between-study heterogeneity in some analyses may also reduce the generalizability of the pooled estimates.；The high heterogeneity among studies may reflect differ- ences in risk score performance across populations and clinical settings.
对本项目的价值
★★★★★ 脑老化是最大未满足需求，也是终点最难量化的领域；biomarker 与认知读数选择最敏感。队列：UK Biobank、FHS、Rotterdam Study；最大样本量约 19,114；设计 RCT、孟德尔随机化。
全文来源
Springer OA PDF（正文 45,018 字符，已过标题/DOI 逐字归属校验）
原文题录
- 英文标题：How well do dementia risk scores predict dementia? A systematic review and meta-analysis.
- 期刊与标识：Geroscience · PMID 42934475 · DOI 10.1007/s11357-026-02615-y**
C 级证据分级：C 级（预印本 / 观点评论，非原始研究）
本文属观点/评论类文献，不是原始研究。其论点依赖于作者对既有证据的梳理与判断，不构成新的实验证据。本站标注 C 级即为此意——阅读时请把它当作「领域内的思考框架」，而非「研究发现」。
关于本文的文献来源：本站依据 IMA 知识库「五刊监控·深度文献卡片」撰写，依据已获取的期刊全文（方法、样本量与结论已回到原文核对）；未能获取全文的同批条目另行标注「摘要级」，全程未做任何数值推算补全。