P104 Theory-Informed Predictors of Fruit and Vegetable Intake Among Cost-Offset Community Supported Agriculture Enrollees

Abstract

<h3>Background</h3> In cost-offset, community-supported agriculture (CO-CSA), low-income households receive subsidized shares of locally-grown produce with the aim of improving their access to and consumption of fruits and vegetables (FV). Social Cognitive Theory (SCT) is a commonly leveraged behavior change theory and informed the design of a CO-CSA plus nutrition education intervention trial for low-income households. <h3>Objective</h3> To examine the cross-sectional relationships between SCT-related variables and FV intake among CO-CSA enrollees at baseline. We hypothesized that nutrition-related attitudes, beliefs, and self-efficacy as well as perceived FV availability and accessibility would be significant predictors of FV intake. <h3>Study Design, Setting, Participants</h3> In Spring 2016 and 2017, adult enrollees (n = 305) completed a baseline questionnaire containing study-specific and validated questions to measure nutrition-related psychosocial (knowledge, attitudes, and beliefs), environmental (FV availability and accessibility), and behavioral factors (self-efficacy) in relation to self-reported and proxy measures of FV intake (National Cancer Institute's All-Day FV Screener and resonance Raman spectroscopy to assess dermal carotenoid status). <h3>Measurable Outcome/Analysis</h3> Descriptive statistics and multiple regression models were completed in SPSS. Models controlled for demographic characteristics. Self-efficacy to prepare and eat FV (β = 0.27, <i>P</i> ≤ .001) and in-home FV availability (β = 0.13, <i>P</i> = .04) were significant predictors of self-reported daily FV intake (R2 = 0.121, <i>P</i> ≤ .001). In-home FV availability (β = 0.20, <i>P</i> ≤ .001), accurate knowledge of FV recommendations (β = 0.15, <i>P</i> = .008), and perceived geographic access to FV (β = 0.11, <i>P</i> = .046) were significant predictors of carotenoid status (R2 = 0.227, <i>P</i> ≤ .001). <h3>Conclusions</h3> Among a set of SCT-related variables, knowledge of FV recommendations, geographic access to FV, in-home availability of FV, and self-efficacy to prepare and eat FV were significant predictors of FV intake. However, these only explained 22-23% of outcome variability among a sample of CO-CSA adult enrollees. SCT alone may not be the most appropriate theory through which to understand potential levers of FV intake among adults participating in CO-CSA programs.