Abstract

The obesity epidemic has increasingly spread around the world in the past two decades. Data from the Center for Disease Control and Prevention (CDC) have shown the prevalence of US adult obesity doubling during this period. Also, the obesity rate reached all-time high, with more than one-third (36.5 percent) of the adult population being obese in 2017. This rapid increase in the prevalence of adult obesity may be attributed to the changes in human behavior and environment. This study was designed to identify the important risk factors that contributed to the US adult obesity rate and quantify the effect of changing a specific key factor on the reduction of the adult obesity rate, in consideration of other risk factors. It was hypothesized that at least one of the risk factors—behavioral factors, socioeconomic status, food environment, and physical environment—contributed to US adult obesity rate. The simulation using full and reduced models of linear regression and partial least squares regression was constructed to determine the relationship between the US county-level adult obesity rate and its multiple risk factors. Of the twenty-three risk factors related to adult obesity, the top three risk factors with high correlation in tornado charts were identified in both full and reduced models in simulation as physical inactivity, median household income, and percent of children eligible for free lunch program. The study showed that adult obesity rate was most influenced by behavioral factors, followed by socioeconomic status and physical environment. Also, the results of sensitivity analysis indicated that if physical inactivity for adults decreases by 5 percent—from 30 percent to 25 percent—the adult obesity rate would drop to 28 percent from the rate of 30 percent in 2013. The research findings in this study were consistent with the literature reviews, indicating that the simulation modeling approach was robust and applicable to inform population-based intervention strategies.