Exploring the relationship between mortality and potential risk factors in US 50 states: A population perspective using partial least squares and geographically weighted regression models

Abstract

While Alzheimer's disease becomes prevalent in elder population and attracts investment of thousands of billions for its research, its pathogenesis remains unknown. The relationship between multiple potential risk factors (Overall Health, Caregiving, cognitive decline, Nutrition/Physical Activity/Obesity, Screenings and Vaccines, mental health, Smoking and Alcohol Use) and Alzheimer's disease mortality in the 50 US states in 2020 was explored by developing multiple linear regression models, partial least squares regression models, and geographically weighted regression models in this article. In this experiment, through multiple linear regression models, we found eight significant demographic indicator variables, and to solve the covariance problem, we successfully constructed the pls model, and using the regression coefficients in the equation, we screened out the most important variables for the model, TOC06 and sex. In addition, we then explored the characteristics of the spatial distribution of mortality in Alzheimer's disease according to TOC06 and sex variables using the GWR model.