Strokes are increasingly impacting younger populations, with approximately 10 to 15 per cent of all cases occurring in people under the age of 50, particularly affecting young women. While classic risk factors like high blood pressure, diabetes, smoking, being overweight, dyslipidaemia, and alcohol or drug use drive this rising trend in ischaemic strokes among young adults, medical research is pushing further. Scientists are examining how genetic predisposition interacts with lifestyle variables, an intricate puzzle that modern technology is uniquely equipped to solve.
Artificial Intelligence and Genomic Analysis
To untangle the complex relationship between genetics and stroke risk, a team of experts at the University of Technology Sydney developed a specialized artificial intelligence model. Led by first author Yue (Eric) Yang, the researchers utilized the same underlying technology that powers large language models like ChatGPT to analyze thousands of genetic variants.
According to Yang, this algorithm stands out because it does more than simply pinpoint genetic variants linked to strokes. It evaluates their specific locations within the genome and analyzes how they interact with one another. When tested using genetic and health data from the UK Biobank database and compared against medical records from the US All-of-Us database, the prototype demonstrated higher accuracy than previous risk assessment tools in differentiating between healthy individuals and stroke survivors.
Key Findings and Identified Genetic Markers
The AI-driven analysis successfully identified 18 distinct genetic variants associated with ischaemic stroke. Notably, individuals with a higher genetic risk score faced a 22 per cent greater likelihood of suffering a stroke compared to the control group.
Among the discoveries, researchers highlighted a specific gene located in the cerebral cortex called EDEM2. This gene appears to serve as a key factor in susceptibility to ischaemic stroke, offering a promising target for future biological and medical research.
Integrating Lifestyle and Next Steps
Genetics represents only one segment of an individual's overall medical profile. Experts emphasize that combining a genomic risk score with clinical metrics and lifestyle factors—such as age, blood pressure, cholesterol, atrial fibrillation, and type 2 diabetes—can significantly enhance monitoring and prevention strategies.
Collaborators Kairui Guo and Dennis Cordato, working with the dedicated unit at Liverpool Hospital in Sydney, envision utilizing AI to better understand the true risks faced by relatives of stroke patients, especially families affected by early-onset strokes. A preventive clinical trial has been proposed to recruit an initial cohort of patients, sequence their genomes, evaluate their stroke risk, and explore potential gender differences between men and women.
At the Cardarelli Hospital in Naples, Vincenzo Andreone, director of the Complex Operational Unit of Neurology and Stroke Unit, notes that while the 22 per cent increased risk for unfavorable genetic profiles is notable, similar figures were observed previously using simpler tools. Instead, the true breakthrough lies in the innovative methodology of the AI model. Rather than just tallying at-risk genes, the technology evaluates how those genes interact, akin to understanding a sentence through the contextual order of its words rather than isolated terms.
Ultimately, specialists remind the public that while DNA offers valuable predictive insights, daily lifestyle habits remain paramount. Blood pressure management, avoiding smoking, maintaining a healthy weight, and exercising regularly continue to hold the greatest influence over stroke prevention.
