Correlates of fruit and vegetable consumption among US Adults: Results from the Food Attitudes and Behaviors (FAB) Survey

Background: The National Cancer Institute developed the Food Attitudes and Behaviors (FAB) Survey to assess the correlates of fruit and vegetable intake (FVI) among US adults. Methods: FAB, a population-level mail survey of adult Americans was fielded October through December 2007 (N=3397) with an oversampling of African Americans and a final response rate of 57%. FAB assessed potential correlates of FVI such as sociodemographics; conventional constructs including self-efficacy, barriers, and benefits related to FVI, lifestyle factors (e.g., physical activity, and smoking); novel constructs including “healthy food influencers”, “vegetarianism”, FV preference/diversity, and intrinsic and extrinsic motivation; and promising individual variables including FVI “when growing up”, seasonality, knowledge of FV recommendations, water consumption, and others. FVI was assessed using a new validated 2-item cup FV screener. All statistical analyses were conducted using SAS 9 (v.9.1). Data were weighted, outliers removed, and imputation conducted. Hierarchical linear regression was conducted entering blocks of correlates in the following order: sociodemographics, lifestyle, conventional, novel, and promising individual variables. Results: Hierarchical regression found that sociodemographics explained 4.2% of the variance, lifestyle factors explained another 4.3%, conventional constructs explained another 13.3%, novel constructs explained another 6.9%, and individual factors explained another 5.3%, so that the total variance explained in the final model was 33.5%. Conclusions: A variety of constructs and variables were found to be strong correlates of FVI. This information can be useful for researchers who want to include new correlates of FVI in their evaluation of interventions, surveillance, and/or research surveys.