Impact of School Rurality on Fruit and Vegetable Selection Differs by Sex and Eating Duration for Arizona Schools.

PURPOSE: Almost 1 in 5 children live in rural communities, which are more obesogenic than urban communities. Because rural schools have a higher proportion of students participating in the National School Lunch Program compared to urban schools, we aimed to understand rural-urban differences in student fruit and vegetable (FV) selection, consumption, and waste during lunch. METHODS: This secondary analysis used baseline data from a cluster-randomized trial testing school-lunch salad bar and FV marketing interventions between 2017 and 2019 across elementary (n = 13), middle (n = 12), and high schools (n = 12) in Arizona. Students' objective FV selection, consumption, and waste was measured using digital scales to the nearest 2 g. Mixed effects zero-inflated negative binomial models tested the effect of school rurality (rural vs. non-rural), adjusted for sociodemographic and within-school clustering. FINDINGS: Rural and non-rural schools did not differ on average FV selection, consumption, or waste. However, sex and eating duration significantly interacted with rurality (p < 0.05). In non-rural schools, girls were 33% more likely to select FVs for lunch compared to boys (OR = 0.43, z = -2.73, p = 0.009) and slower eaters in non-rural areas had 1.16 greater odds of selecting FVs compared to faster eaters (OR = 1.16, z = -5.65, p = 0.002). CONCLUSION: Unexpectedly, rurality was not a correlate of students' FV selection, consumption, or waste during school lunch. However, associations between sex, eating duration, and FV selection in non-rural schools aligned with the broader scientific literature. These relations were not observed in rural schools, highlighting the need to better understand the influence of sociodemographic factors and eating styles on children's eating behaviors in rural schools.