Protein importance ranking and SHAP visualization of modeling across all incident dementia populations. A, Sequential forward selection from pre-selected candidate proteins. The bar graph indicates the importance of proteins sorted based on their contributions to predicting future ACD (as judged by information gain). The line graph illustrates the cumulative AUC values (right axis) when including proteins one by one in each iteration. The best proteins we finally selected are marked in red. Shaded regions represent standard errors derived from cross-validation. b, SHAP visualization plot of selected proteins. The width of the range of horizontal bars can be understood as the extent of contribution to ACD prediction; the broader their scope, the greater the contribution. The color of the horizontal bars indicates the magnitude of plasma proteins, which was coded in a gradient from blue (low) to red (high), represented by the color bar on the right side. The direction on the X The axis indicates the probability of developing dementia (right) or being healthy (left). Credit: Natural aging (2024). DOI: 10.1038/s43587-023-00565-0
In the largest study of its kind, scientists showed how protein “biomarkers” predict dementia 15 years before diagnosis.
Research shows how protein profiles in the blood accurately predict dementia up to 15 years before clinical diagnosis. These are known as biomarkers, which are molecules present in blood, other body fluids or tissues that indicate a normal or abnormal process or condition or disease.
In the study, scientists from the University of Warwick and Shanghai Fudan University used the largest blood proteomics and dementia cohort to date, including blood samples from 52,645 healthy participants. health recruited from the UK Biobank, a population-based study cohort.
Blood samples collected between 2006 and 2010 were frozen and then analyzed 10 to 15 years later by the research team, who analyzed them between April 2021 and February 2022. Through March 2023, a total of 1,417 participants developed dementia. blood showed dysregulation of protein biomarkers.
Of 1,463 proteins analyzed, using machine learning, 11 proteins were identified and combined to form a panel of proteins, which researchers showed were highly accurate in predicting future dementia. Further incorporation of the conventional risk factors of age, gender, education level and genetics showed for the first time the high accuracy of the predictive model, measured at over 90%, indicating its potential future use in community dementia screening programs.
Proteins (e.g. glial fibrillary acidic protein, GFAP) had previously been identified as potential dementia biomarkers in smaller studies, but this new research was much larger and conducted over several years. Known as a longitudinal analysis (a study conducted on a sample of participants over several years), researchers were able to show differences and trajectories between people with dementia and controls over 15 years.
Early diagnosis is essential for people with dementia. New pharmaceutical technologies can slow or even reverse the progression of Alzheimer’s disease, but only if the disease is detected early enough. The drug lecanemab is one of two new treatments for the disease.
Lead author Professor Jianfeng Feng, from the University of Warwick’s Department of Computer Science, hopes that future drugs can be developed to interact with the proteins identified in the study.
Professor Feng highlighted that the combination of artificial intelligence and protein analysis offers a promising avenue for precision medicine. This is very important for screening middle-aged to older people in the community who are at high risk of dementia. “This model could be seamlessly integrated into the NHS and used as a screening tool by GPs,” Professor Feng said.
Co-corresponding author Professor Wei Cheng from Fudan University explained that this research builds on the dementia prediction model previously developed by the team, which used variables such as age, presence of a certain gene and the age of the mother at death. “Compared to our previous work, the newly developed protein-based model is obviously a major breakthrough,” he said.
Another corresponding author, Professor Jintai Yu, a specialist in neuro-vegetative diseases from Fudan University, added: “Proteomic biomarkers are more easily accessible and non-invasive, and they can significantly facilitate the application of large-scale screening. population scale. »
He also highlighted the drawbacks of previous risk models, which relied primarily on complex biomarkers that were difficult to obtain using procedures such as lumbar puncture or complex imaging methods, meaning their widespread use is hampered. due to invasive procedures and the high cost of performing them.
The results are published in the journal Natural aging.
More information:
Jian-Feng Feng, Plasma proteomic profiles predict future dementia in healthy adults, Natural aging (2024). DOI: 10.1038/s43587-023-00565-0. www.nature.com/articles/s43587-023-00565-0
Provided by the University of Warwick
Quote: Protein biomarkers predict dementia 15 years before diagnosis, according to new study (February 12, 2024) retrieved February 12, 2024 from
This document is subject to copyright. Apart from fair use for private study or research purposes, no part may be reproduced without written permission. The content is provided for information only.
