A new UC Irvine-led study shows that machine learning algorithms can predict dementia risk in American Indian/Alaska Native elders, providing a valuable resource for healthcare systems serving resource-limited populations.
New research led by the University of California, Irvine has shown that machine learning algorithms can successfully predict the two-year risk of dementia among American Indian and Alaska Native adults aged 65 and older. This pioneering study, published in The Lancet Regional Health – Americas, offers a significant leap forward in healthcare for a community often underrepresented in medical research.
This is the first time machine learning models have been harnessed to predict dementia risk among the American Indian/Alaska Native population, as defined by the U.S. Census Bureau.
Machine learning, a form of artificial intelligence, empowers computers to predict outcomes or make decisions by analyzing vast datasets without explicit programming. This technology enhances efficiency and precision, making the analysis of large datasets scalable and reliable.
The study addresses a critical need. The demographic of older American Indian and Alaska Native adults is expected to nearly triple from 2020 to 2060. With dementia posing a significant threat as a leading cause of disability and death in this age group, timely and accurate risk prediction tools are increasingly vital.
Dementia not only leads to cognitive decline and depression but also places emotional and financial burdens on families and significantly diminishes quality of life.
The researchers utilized seven years of data from the Indian Health Service’s National Data Warehouse. They divided this data into a five-year baseline period (2007-2011) and a two-year prediction period (2012-2013) and analyzed records from nearly 17,400 American Indian/Alaska Native adults aged 65 and older who were dementia-free at the study’s outset.
Over a two-year follow-up period, 611 individuals (3.5%) were diagnosed with dementia.
The team evaluated four machine-learning algorithms to determine their effectiveness. They found that 12 out of 15 top predictors for dementia were consistently identified across the three best-performing models, uncovering several novel predictors such as health service utilization.
“Public health researchers play a significant role in helping clinicians and policymakers make informed decisions about population health,” corresponding author Luohua Jiang, a professor of epidemiology and biostatistics at the UC Irvine Joe C. Wen School of Population & Public Health, said in a news release. “If future studies confirm these results, our findings could prove valuable to the Indian Health Service and Tribal health clinicians in identifying high-risk individuals, facilitating timely interventions and improving care coordination.”
Source: University of California, Irvine