Applying Multiple Statistical Methods to Derive an Index of Dietary Behaviors Most Related to Obesity

Abstract

To evaluate the success of dietary interventions, we need measures that are more easily assessed and closely aligned with intervention messaging. An index of obesogenic dietary behaviors (e.g., task-eating, servings of fruits and vegetables, fast food, and soft-drinks) may serve this purpose and could be derived via data-driven methods typically used to describe nutrient intake. We used behavioral and physical measurement (i.e., body mass index, waist circumference) data from a subset of two independent cross-sectional samples of employees at baseline (2005–2007) (n = 561) and follow-up (2007–2009) (n = 155) enrolled in Promoting Activity and Changes in Eating. Index derivation methods including principal components regression, partial least squares, and reduced rank regression were compared. The best fitting index for predicting physical measurements included fast food, French fries, and soft-drinks. Each quartile increase in index score was associated with 5% higher BMI (Ratio = 1.053; 95% CI: 1.031, 1.075) and approximate 4% higher WC (1.036; 95% CI: 1.019, 1.054) at baseline adjusted for covariates, using linear mixed models. Results were similar at follow-up before and after baseline adjustment. This index may be useful in evaluating public health or clinic-based dietary interventions to reduce obesity, especially given the ubiquity of these behaviors in the general population.