Dissertations & Theses (Open Access)

Date of Award


Degree Name

Doctor of Public Health (DrPH)


Casey P. Durand

Second Advisor

Shreela V. Sharma

Third Advisor

Christine Markham


American children eat fewer fruits and vegetables (F&V) and less variety of F&V than recommended for health. Food cooperatives and other programs have become a popular way to increase F&V intake, but little is known about the variety of F&V distributed by these programs or its relationship with program attendance or child F&V intake. Brighter Bites is a national, school-based food co-op distributing rescued, donated, fresh F&V to families in low-income schools. We evaluated, for the first time, the variety of F&V Brighter Bites distributed to families in the 2018-2019 school year and the relationships between that variety and both child F&V intake and family program attendance. We categorized the F&V distributed in the 2018-2019 school year using the Brighter Bites internal variety matrix and described them in detail using frequencies and percentages. We generated a variety score for each family in a subpopulation (n=3,790) of survey respondents based on the specific F&V distributed the weeks they attended. A generalized ordinal estimation model was specified to evaluate the relationship between family variety score and parent-reported child F&V intake before and after participating in Brighter Bites. We generated a variety score for schools (n=90) based on the specific F&V distributed at each school across 16 weeks of programming, then specified a multilevel negative binomial model to assess the relationship between school variety score and family program attendance. Additional post hoc analyses were completed. Across six cities, Brighter Bites distributed 109 types of F&V in the 2018-2019 school year. Families most frequently received starchy and root vegetables (white potatoes and carrots) and citrus fruits (limes and oranges), but they received dark leafy green vegetables and berries infrequently. Our statistical models were not significant overall, but in post hoc analyses of school F&V variety score and family program attendance we found differences between cities which may have obscured a relationship in our original model. Researchers are still in the early stages of evaluating and understanding relationships between the variety of F&V programs distribute and desired program and behavioral outcomes. Counting only the variety of F&V distributed by a program is inadequate to describe its influences on individual behaviors. Additional, more sensitive measures and variables, informed by a behavioral theory such as Social Cognitive Theory, should be used in future analyses to model better the intrapersonal, interpersonal, and environmental factors which influence desired outcomes.