XGBOOST MACHINE LEARNING TO IDENTIFY PREDICTIVE VALUES OF CARDIOMETABOLIC RISK FOR COGNITIVE DECLINE AND MORTALITY
2024

Using Machine Learning to Find Key Health Indicators for Cognitive Decline and Death

Sample size: 6814 publication Evidence: high

Author Information

Author(s): Liu Longjian, Cui Saishi, Wang Ruifeng, Xu Kaidi, Eisen Howard

Primary Institution: Drexel University

Hypothesis

The study aims to identify crucial biomarkers of cardiometabolic disorders that contribute to the risk of cognitive decline and mortality among adults with cognitive decline and dementia.

Conclusion

Machine learning techniques effectively identified important health indicators related to cognitive decline and mortality risks.

Supporting Evidence

  • The study followed participants from 2000-2002 to 2015.
  • 227 participants with cognitive decline or dementia died during the follow-up.
  • XGBoost models showed an accuracy of 74% in classifying cognitive decline.
  • The C-index for mortality risk was 0.86%.

Takeaway

Researchers used computers to find important health signs that can help predict if someone with memory problems might get worse or die.

Methodology

The study analyzed a cohort of 6814 participants aged 45 to 84 years, assessing cognitive function and using machine learning to identify biomarkers.

Participant Demographics

Participants were aged 45 to 84 years, with a follow-up on cognitive function in 4,591 individuals.

Statistical Information

Confidence Interval

95%CI for mortality rate: 3.8 (3.2-4.5) per 1000 person-years in males, and 2.1 (1.7-2.5) per 1000 person-years in females.

Digital Object Identifier (DOI)

10.1093/geroni/igae098.3293

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication