An artificial intelligence-informed proof of concept model for an ecological framework of healthy longevity forcing factors in the United States

Abstract

Unhealthy lifestyle behaviors are a doorway to downstream health consequences characterized by the following: 1) poor quality of life and diminished mobility; 2) increased likelihood of chronic disease risk factors and diagnoses; and, ultimately, 3) a shorter lifespan and healthspan. The aim of the current study is to assess if an ecological framework can predict U.S. lifespan via the use of artificial intelligence. The current study utilized several U.S. county-level datasets representing the predictive variables of the ecologic framework. A non-linear artificial intelligence statistical approach was used to assess the ability of these variables to predict life expectancy, death rate, and years of life lost. The R² values demonstrated that the performance of Extra trees models was different across the three outcomes, however, death rate always exhibited the highest R² for each feature number, indicating superior model accuracy for this outcome. Generally, an increase in the number of features led to improved model performance. Variables from all factors included in the proposed ecological framework were retained in the final predictive models. There is a need to understand why individuals/families/community, connected by shared cultural beliefs, decide to make one lifestyle behavior decision over another.