Over the years, science has revealed numerous Alzheimer’s risk factors. Some of them, like vision loss or hearing loss, can be addressed by wearing glasses or a hearing aid. Others, like higher genetic risk with an APOE4 gene, are virtually impossible to change. However, we’ve reported on research that following a whole-food, nutritious diet and living a healthy lifestyle with regular exercise and sleep can influence certain genetic risk factors for illness.
Other risk factors we are unaware of are bound to contribute to Alzheimer’s, but scientists are still identifying and getting cutting-edge help with artificial intelligence (AI).
As it turns out, AI can uncover links to the disease no one had ever considered.
AI Revolutionizing Alzheimer’s Research
Artificial Intelligence is moving into the world of medicine because it can analyze massive amounts of data and make connections that would otherwise leave humans in the dark.
For instance, language impairment is a common symptom in Alzheimer’s patients, and, as we previously reported, AI was able to pick up subtle linguistic changes to predict the disease with an accuracy of 80.3 percent.
Scientists at the University of California San Francisco (UCSF) are also using AI to detect Alzheimer’s, but they’re coming at it from a different angle.
A Predictive Ability of 72 Percent
The researchers used UCSF’s clinical database, which contains health records of more than five million patients from 1980 to 2021. The database includes health conditions, blood tests, lab reports, and imaging results.
They ran a machine-learning program with an AI algorithm trained to identify common features and conditions shared by those who went on to be diagnosed with Alzheimer’s disease over a seven-year period.
Once AI uncovered the conditions that predicted Alzheimer’s disease, the next stage was discovering their biological reasons.
To this end, they used a tool called SPOKE (Scalable Precision Medicine Knowledge Engine). SPOKE is a database of over 25 databases. It allows researchers to tie together data on interconnected pathways covering almost every aspect of biology. The information identifies patterns and potential molecular targets for therapy.
The findings showed they could identify with 72 percent predictive power who would develop the disease up to seven years prior to onset. Some of the conditions linked to Alzheimer's disease were expected, but others came as a surprise.
Osteoporosis Linked to Alzheimer’s in Women
The conditions that most influenced the prediction were high cholesterol and, for women, the bone-weakening disease osteoporosis.
The study’s lead author, Alice Tang said: “This is a first step towards using AI on routine clinical data, not only to identify risk as early as possible, but also to understand the biology behind it. The combination of diseases allows our model to predict Alzheimer's onset. Our finding that osteoporosis is one predictive factor for females highlights the biological interplay between bone health and dementia risk.”
The complete list of conditions linked to Alzheimer's include:
- High cholesterol
- High blood pressure
- Vitamin D deficiency
- Erectile dysfunction
- Enlarged prostate
- Major depressive disorder
- Osteoporosis – women only
The study’s senior author, Marina Sirota, added, “This is a great example of how we can leverage patient data with machine learning to predict which patients are more likely to develop Alzheimer’s and also to understand why that is so.”
What Did AI “See” To Make The Discoveries?
SPOKE identified the association between Alzheimer’s and high cholesterol through a variant form of the apolipoprotein E Alzheimer’s-linked gene, APOE4. As for the link between osteoporosis in women, it was through a variant in a lesser-known gene called MS4A6A.
Whether treating any of these predictive conditions lowers the risk of Alzheimer’s is unknown, so the researchers intend to mine the database of health records further to see if they can find out.
