Multidimensional Analysis of Food Consumption Reveals a Unique Dietary Profile Associated with Overweight and Obesity in Adolescents

Abstract

There is a significant increase in overweight and obesity in adolescents worldwide. Here, we performed a cross-sectional study to examine the potential association between food consumption profiles and overweight in a large number of adolescents from Brazil. Sampling by clusters and conglomerates was carried out in students of public schools in Salvador, Brazil, between June and December 2009 and 1496 adolescents were evaluated. Data on socio-epidemiological data, anthropometric status and food consumption were captured. Multivariate analyses, such as hierarchical clustering and correlation networks, were used to perform a detailed description of food consumption profiles. There were differences in age and anthropometric status related to sex. Four clusters of food groups were identified based on the intake profile in the study population. No disparities in food intake were observed in individuals stratified by sex or anthropometric status. Furthermore, network analysis revealed that overweight or obesity were hallmarked by a selectivity in the ingestion of food groups that resulted in the appearance of inverse correlations of consumption, which was not present in eutrophic adolescents. Thus, overweight and obesity are associated with preferential choices of ingestion of specific food groups, which result in the appearance of inverse correlations of consumption. Such knowledge may serve as basis for future targeted nutritional interventions in adolescents.

Systems Nutrology of Adolescents with Divergence between Measured and Perceived Weight Uncovers a Distinctive Profile Defined by Inverse Relationships of Food Consumption

“…This study is cross-sectional analysis of a databank previously published from our group [ 14 ]. The primary study was performed in Salvador, state of Bahia, a large city of Brazil.…”

“…The details on delineation of study sample selection are described elsewhere [ 14 ]. Briefly, to define the study sample, we used the simple random sampling technique.…”

“…In this phase, the Department of Education of the state of Bahia provided the names and locations of the public schools as well as the list of registered students. These individuals were approached to verify inclusion and exclusion criteria [ 14 ], resulting in a total of 1496 students enrolled.…”

“…Furthermore, the consumption of each food and/or preparations was standardized to read units of weight (g) and or volume (mL). This information was next used for computation of the daily ingestion of the food recorded through the FFQ as previously described [ 14 , 18 ]. The results of these calculations were deconvoluted to extrapolate the daily consumption [ 14 , 18 ].…”

“…Systems biology uses different information, concepts, and approaches to better understand the whole picture of biological processes and diseases [ 15 ]. By incorporating systems biology tools to analyze nutritional data, in what we call ‘systems nutrology’, our previous study reported that overweight and obesity are associated with preferential choices of ingestion of specific food groups [ 14 ]. In the present study, we expanded these observations to assess whether the self-perception of the body image influences the pattern of food consumption of the adolescents.…”