A Color Atlas of Morphologic Hematology

形态学血液学彩色图谱

阅读:1

Abstract

BACKGROUND: Preclinical Alzheimer's disease (AD) develops 15–20 years before symptom onset. This study aimed to develop models to enhance the detection and monitoring of preclinical AD. METHOD: A retrospective cohort study was conducted based on data from the A4 study, spanning February 28, 2014, to June 8, 2023. A total of 420 A4 participants with preclinical AD were analyzed to examine associations between hematology/serum biochemistry markers and brain amyloid‐PET SUVR or plasma pTau‐217 levels. Multivariable linear regression models were used for these analyses, and machine learning models were developed to predict these AD biomarkers over a 4.5‐year period. RESULT: Increased monocyte count was associated with elevated brain amyloid‐PET SUVR, while higher mean corpuscular volume (MCV) was linked to elevated plasma pTau‐217 levels. Additionally, elevated AST‐to‐ALT ratios and creatinine levels were associated with increased amyloid‐PET SUVR and plasma pTau‐217 levels. Hematology and serum biochemistry markers notably improved the predictive performance of models for plasma pTau‐217 levels. However, monocyte count was not a strong predictor of amyloid‐PET SUVR. CONCLUSION: The machine learning models demonstrated promising potential for predicting AD biomarkers, emphasizing the need for future research to evaluate their applicability in both clinical and research contexts.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。