Objective
We aimed to find out what factors can predict cognitive decline in older adults who are currently cognitively healthy over a 5-year period.
Background
Alzheimer’s disease varies greatly among individuals, making it challenging to design clinical trials and select the right participants. Understanding how to identify more similar groups can help improve trial outcomes.
Study Design and Methods
The study involved 1,169 participants who tested positive for amyloid (a protein linked to Alzheimer’s) and were part of the A4 trial, where some received the drug solanezumab and others a placebo. Additionally, 538 participants who tested negative for amyloid were part of the LEARN study. We used statistical models to analyze various factors such as age, gender, genetic markers, cognitive tests, and specific protein levels in the blood to predict cognitive decline.
Results
The average age of participants was about 70 years, with a majority being female. Adding specific protein levels (P-tau217) and cognitive scores to our initial analysis significantly improved our ability to predict cognitive decline. The best predictive models showed strong performance across all groups involved in the study.
Conclusions
This study highlights the importance of initial cognitive tests and specific blood protein levels in predicting cognitive decline. Using these practical measures can enhance the design of clinical trials and improve participant selection, leading to better treatment outcomes.
Opportunities for Clinics and Patients
Based on our findings, clinics can:
- Define clear goals for predicting cognitive decline using baseline data.
- Select AI tools that cater to specific clinical needs.
- Implement changes gradually, starting with pilot projects and tracking results.
Contact Us for AI Solutions in Medical Management
If you’re interested in AI solutions tailored for healthcare, reach out to us:
