Machine learning models predict dementia risk among American Indian/Alaska Native adults

Machine Learning-Algorithms using electronic health files can effectively predict two-year dementia risks in American Indian/Alaska-inhemid adults aged 65 and older, according to a study led by the University of California, by Irvine. The findings offer a valuable framework for other health care systems, in particular those services for resource-restricted populations.

The computer modeling results also found various new predictors for the diagnosis of dementia that were consistently identified in different models for learning the machine. Findings are published in the Lancet Regional Health – Americas. The National Institutes of Health supported the research.

Until now, no other study has looked at utilizing the power of Machine Learning models to predict the risk of dementia in the historically investigated Indian/Alaska resident of the population, as defined by the US Census Bureau.

Machine Learning models, with which computers can make predictions or decisions using huge data sets without explicit programming for each task, improve efficiency, accuracy and scalability when analyzing large data sets.

The population of older American Indian and Alaska indigenous adults is expected to increase almost triple between 2020 and 2060. With dementia is an important cause of disability and mortality in this age group, this debilitating state is an increasing care in this community.

In addition to countless ailments such as cognitive decline, weakened immune system and depression, dementia has far -reaching social effects. It emotionally takes a toll on family members, incurs significant medical costs and contributes to a general decrease in quality of life.

Public health researchers play an important role in helping clinicians and policy makers who make informed decisions about the health of the population. If future studies confirm these results, our findings can be valuable for the Indian Health Service and StaGmen Health clinics when identifying people with a high risk, facilitating timely interventions and improving care coordination. “

Luohua Jiang, professor of epidemiology and biostatistics, UC Irvine Joe C. Wen School of Population & Public Health

Jiang and colleagues took seven years of data from the national data warehouse of the Indian Health Service and the related electronic health files and divided the data into a five-year basic line period (2007 to 2011) and a two-year dementia forecast period (2012 to 2013). The study included nearly 17,400 Indian/Alaska Indigenous adults aged 65 or older who were dementia-free at the baseline, of whom almost 60 percent were female.

During the two-year follow-up, 611 people (3.5 percent) were diagnosed with dementia. Four machine-learning algorithms were evaluated and compared on the basis of their pre-processing efforts for data and model performance. Of the three best performing models that the team developed, 12 out of 15 highly ranked predictors were customary for dementia in the three models. It is important that various new predictors of dementia were identified in these algorithms by all causes, such as the use of health care.

Extra authors include Kayleen Ports, a former UC Irvine Master’s student, and Jiahui Dai, a current graduate student researcher, both from Wen Public Health; Kyle Conniff, a recent graduate UC Irvine Promovendus in statistics; and Maria M. Corrada, a professor in neurology at the UC Irvine School of Medicine. Spero M. Manson, a prominent professor, and Joan O’Connell, a associate professor, with the Centers for American Indian & Alaska Native Health at the Colorado School of Public Health also contributed to the study.

The National Institutes of Health AIM-AHEAD (Consortium for Artificial Intelligence/Machine Learning to promote the Equity Equity and Researchers diversity, 1OT2OD032581) and the National Institute on Aging (R01AG01189) for the study.