The risk of someone dying from cardiovascular disease can now be predicted by artificial intelligence (AI) using survey responses, according to new research using the 45 and Up Study.
The research from the University of New South Wales, published in the International Journal of Cardiology, may lead to the development of risk prediction tools that reduce the need for expensive tests and can be used in rural and remote areas.
Researchers used survey answers from 187,000 participants in the 45 and Up Study, who had no history of cardiovascular disease, along with linked data on mortality and hospitalisation rates between 2006 and 2020. This information helped develop machine learning-based risk prediction models for both cardiovascular mortality and hospitalisation from ischemic heart disease.
Machine learning is a type of AI where a computer system learns to make accurate predictions when fed data. The survey answers given to the risk-prediction models held information on a person’s age, gender, income and medical history as well as height, weight, physical activity and dietary patterns.
The survey-based models had similar success in predicting high-risk individuals as previous models that used results from blood tests and cardiac imaging.
“These health survey questions are easily applied at the individual patient level,” wrote the study authors. “Our models may have the potential to be used in initial screening tests to identify high-risk individuals before undergoing costly investigation.”
The researchers noted that for cardiovascular mortality, the top five features that were associated with a greater risk were higher age, lower household income, no employment, use of anti-hypertensive medication, and less time spent standing.
This paper adds to the growing body of research using the 45 and Up Study for AI-based prediction of chronic disease. Previous research has used the Study to predict the factors involved in the development of type 2 diabetes, as well as investigate the shared risk factors for cancer and mental disorders.
The 45 and Up Study is one of the biggest ongoing longitudinal studies in the world. Administered by the Sax Institute, the Study has followed more than 250,000 Australian men and women aged 45 years and over since 2005. Study participants are followed up approximately every five years with a survey on their health and wellbeing and have consented to have their survey data linked to a range of administrative datasets.
