An empirically derived “prudent” dietary pattern is associated with lower obesity occurrence: Modeling and mapping from a national nutrition survey

Abstract

The characterization of dietary patterns linked to obesity is important to address evidence-based interventions against the obesity epidemic. We hypothesized that both dietary behaviors and environmental characteristics affect the spatial distribution of obesity. This study aimed to analyze the role of dietary patterns in the occurrence of obesity in Colombia, taking into account the hierarchical (geographical) structure of the data. A cross-sectional study based on the Colombian National Nutrition Survey (2015) was carried out. A subsample of 8750 adults who reported dietary intake data (by the 24-hour diet recall method) was extracted. Dietary patterns were identified by principal component factor analysis. Their association with obesity occurrence was assessed by using multilevel Poisson regression models (8750 subjects nested in 33 geographic units). Thematic maps were constructed. The obesity prevalence was 18.0% (95% CI, 16.7-19.4). Four dietary patterns were identified for the study group. An inverse (negative) association was found in people with high adherence (vs. low adherence) to the "prudent pattern” (dairy products, fruits, vegetables) (prevalence ratio, 0.87; 95% CI, 0.78-0.96; P = .01). People with high adherence to the "soft drinks and snacking pattern" (sugary drinks, meat products, snacks) tended to show a direct (positive) association with obesity occurrence (prevalence ratio, 1.12; 95% CI, 0.99-1.27; P = .06). Maps showed higher adherence levels to the prudent pattern in departments with higher socioeconomic conditions and lower obesity prevalence. To conclude, a prudent dietary pattern emerges as a key factor in obesity occurrence in Colombia. Obesity spatial distribution may be conditioned by contextual factors that represent social inequalities in health.