ARS | Publication request: MULTIVARIATE ANALYSIS OF FOOD SUFFICIENCY, HEALTH, AND NUTRITION FACTORS FROM THE CSFII 1994-1996

Submitted to: Federation of American Societies for Experimental Biology Conference
Publication Type: Abstract Only
Publication Acceptance Date: November 6, 2001
Publication Date: N/A

Technical Abstract: Understanding the interrelationship of food sufficiency, nutrition and health is vital but, this interrelationship has usually been analyzed with only one outcome as the dependent varible. Using select demographics, nutrient intakes, and pyramid servings (24-hour recall), self-reported health conditions, and household food sufficiency data from adult respondents of the CSFII 1994-1996, we used multivariate statistical methods, particularly variable clustering, for variable reduction to improve the understanding of complex relationships among health, nutrition, and food insufficiency. For example, we found that health conditions including diabetes, heart disease, high cholesterol, hypertension, and stroke formed a cluster while income, food insufficiency, annual income as a percent of poverty, and food stamps in last 12 months formed another. Although there were minor differences in clusters when the sample was split by race, gender, or year of sample, many clusters were consistent. The resulting factor loadings were used to score new factor variables which were used in weighted regression analyses. Supported by ARS/USDA Project #6251-530000-0020-00D