Americans spend billions of dollars in personal consumption expenditures each year. The percentage of FAH expenditures in the United States has been dwindling, while the percentage of FAFH expenditures has increased. Many factors might be causing this trend. Complex interactions of such factors determining the U.S. consumer’s intake of FAH and FAFH expenditures were studied using machine learning and Directed Acyclic Graphical approaches. Employment and education status are common causes of both FAH and FAFH expenditures. Body mass index, marital status, race and sex have mixed effects. Findings will be useful for policy makers to implement social support programs.