Premium
Childhood overweight and obesity models with high correlated food consumption data
Author(s) -
Méndez-Gómez-Humarán Ignacio,
Méndez-Ramírez Ignacio,
Shamah-Levy Teresa,
Moreno-Macías Hortensia,
Murata Chiharu
Publication year - 2012
Publication title -
the faseb journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.709
H-Index - 277
eISSN - 1530-6860
pISSN - 0892-6638
DOI - 10.1096/fasebj.26.1_supplement.1011.4
Subject(s) - overweight , cluster (spacecraft) , consumption (sociology) , analysis of variance , obesity , statistics , cluster analysis , association (psychology) , regression analysis , linear regression , mathematics , environmental health , medicine , computer science , psychology , social science , sociology , psychotherapist , programming language
Objective Discus the use of cluster analysis in statistical models for association of overweight and obesity (OW+O) with high correlated food consumption variables. Methods We compare regression models for association of OW+O with nutrient consumption (energy, carbohydrates, lipids, fiber and sodium). A sample of 8,367 children with food and beverages consumption records, separated in out of home and global consumption. Ward clustering method was used to group consumption patterns based on standardized averages by state. A simple analysis of variance was used to model the association. Results Linear models fail to find significant associations between consumption patterns and OW+O due to multicolliniarity. Cluster analysis is useful to establish grouped consumption patterns, we selected 4 consumption groups. Analysis of variance and Tukey‐Kramer test shows a significant higher prevalence of OW+O in cluster 2 with respect to cluster 4. Conclusion Nutrient consumption patterns and other collinear data can be used integrally with cluster analysis for group patterns as an explanatory variable in statistical models.