Author
![]() |
SIMPSON, P - ACHRI - DAC |
![]() |
GOSSETT, J - ACHRI - DAC |
![]() |
HALL, R - ACHRI - DAC |
![]() |
WEBER, J - ACHRI |
![]() |
Bogle, Margaret |
Submitted to: International Society for Behavioral Nutrition and Physical Activity
Publication Type: Abstract Only Publication Acceptance Date: 3/14/2003 Publication Date: 7/19/2003 Citation: SIMPSON, P., GOSSETT, J., HALL, R., WEBER, J., BOGLE, M.L. BRANCHING OUT WITH TREES TO ANALYZE DIETARY AND BEHAVIORAL CORRELATES OF OBESITY IN COMPLEX SURVEYS. International Society for Behavioral Nutrition and Physical Activity. Quebec, Canada. 2003. p. 19. Interpretive Summary: Technical Abstract: Purpose: To show that the use of tree analysis, in conjunction with logistic regression, enables construction of better and more appropriate models to explore dietary correlates of Obesity. Background: In regression analysis, the decision about which variables to include and in which form they should be included in the model can be very difficult. Screening variables and their interactions can be very tedious. All types of variables can be included in a tree analysis, including variables with missing values and variables that are highly interrelated. Because of the tree methodology, cutpoints for variables that best optimize a function are given, so it is possible to consider new variables generated from the old variables. Trees are also useful for exploring the interaction of variables. For example, if a variable appears on one side of a tree and not on the other, it suggests that there is indeed an effect of interaction. Methods: A 24-hour dietary recall was completed in the National Health and Nutrition Examination Survey, (NHANES) 1999-2000. Healthy Eating Index (HEI) scores, composed of 10 components each representing different aspects of a healthful diet, were constructed for adults. Results: Using the HEI, demographics, income status, federal aid, obesity risk factors and behaviors, as well as their interactions, we show how trees and weighted trees shed light on the correlates of obesity. Conclusion: New statistical methodology can enhance understanding of the dietary patterns and behaviors that affect obesity |