In the long journey of exploration of medicine, human beings have always been full of challenges for the prevention and development of various diseases. Among them, dementia, a disease that seriously affects the quality of life of the elderly, is a goal that scientists have been striving for decades. Recently, the scientific research team of Fudan University brought us an exciting news: they used large-scale proteomics data and artificial intelligence algorithms to successfully discover important plasma biomarkers that can lead to the risk of dementia in the future, and can reveal the haze of dementia 15 years in advance!
This landmark study has been published in the journal Nature Aging and has been highly praised by the main journal of Nature. It not only provides us with a new perspective on the difficult problem of dementia, but also points the way to early prevention and strategies for the future.
So, how exactly did this study do it?
The team sifted through key information from blood samples from more than 50,000 healthy adults and followed them meticulously for 14 years. During these long years, they are like guardians, always keeping an eye on the health of these participants. Eventually, they found that 1,417 of them had developed dementia. Through an in-depth analysis of blood samples from these patients, the researchers identified four proteins that are closely associated with dementia and Alzheimer's disease: GFAP, NEFL, GDF15, and LTBP2.
Surprisingly, for some participants who later developed dementia, their blood levels of these proteins were already out of the normal range more than a decade before the onset of symptoms. This discovery undoubtedly provides us with a valuable window of opportunity to take active interventions before the shadow of dementia creeps in.
Among these four proteins, GFAP is particularly striking. The study found that people with high blood GFAP levels were more than twice as likely to develop dementia and nearly three times as likely to develop Alzheimer's as well! This data is undoubtedly a wake-up call for us to pay close attention to changes in this biomarker.
However, it is clearly not enough to rely solely on the levels of these four proteins to judge dementia risk. In order to improve the accuracy and reliability of the algorithm, the researchers further applied machine learning technology to design an efficient algorithm. They combined the levels of these four protein biomarkers with demographic factors such as age, gender, education level, and family history to construct a comprehensive model.
How does this model work? The researchers trained it with information from two-thirds of the study's participants, acting like a rigorous mentor, constantly honing and refining its performance. They then tested it with data from the remaining 17,549 people to test its combat capabilities. The results are exciting: the model is 90% accurate in terms of the risk of developing dementia 15 years in advance**!
Behind this achievement is the hard work and unremitting efforts of the team led by Cheng Wei, a researcher at the Institute of Brain-inspired Intelligence Science and Technology of Fudan University. Researcher Cheng Wei said: "The fully phenotypic dementia model constructed by our team in the early stage can advance the risk of onset by 10 years with an accuracy of 85%, and this study further advances the number of years to 15 years before the onset of the disease and the accuracy exceeds 90%; This study shows the important role of proteomics in the early precision intervention of brain diseases, and provides new ideas for future brain disease research. ”
The success of this study not only provides us with a powerful tool to protect against dementia risk, but also reveals the great potential of proteomics in the study of brain diseases. It shows us how the power of science can demystify disease step by step and lead to a healthier and better future. Let's look forward to more such scientific research results in the near future!
Technology**