For example, at School A, on a day when 334 entrées (of four varieties) and 266 fruit items (of one variety) were prepared, only 42 vegetable items (of two varieties) were prepared. Analysis of the food production records showed that 10.2% of fruit and 28.7% of vegetable items served were left over

after service. Across all schools, vegetables were left over at a greater rate (range 22.0% to 34.6%) than fruits (range 5.0% to 16.4%) (Table 3). Among vegetable items, salads were prepared at the lowest quantities and left over at the highest quantities — e.g., at School B on a day when 181 meals were served, only 5 salads (of one variety) were prepared and all 5 were left over. The most frequently wasted fruit items were whole fruit (e.g., whole orange or apple), while fruit juices and

fruit cups were left over at lower rates. Plate waste data were collected for 2228 students — 35.5% of Selleckchem Ibrutinib the total meals served over buy Talazoparib five days at each of the four middle schools during the study period. Plate waste data analysis suggests that many students did not select fruit (31.5%) or vegetable (39.6%) items. Of those who did, many did not eat any, with more students wasting vegetables (31.4%) than fruits (22.6%) (Table 3). Rates of students selecting and eating fruits and vegetables differed across schools. School B had the highest rate of students selecting these items, but also high rates of wasting new them (Table 3). Results of the logistic regression suggest that rates of selecting and eating items differed by sex. A greater percentage of female students selected

fruit (51.0%) and vegetables (42.1%), than male students (41.7% and 32.2%, respectively) — odds ratio for selecting fruit (male as the referent group): 1.45 (95% CI 1.05, 2.00), odds ratio for selecting vegetable (male as the referent group): 1.52 (95% CI 1.32, 1.76). Among students who selected fruit, a greater percentage of female students ate any fruit, compared to male students (odds ratio for eating any fruit (male as the referent group): 1.41 (95% CI 1.02, 1.95)) (Table 4). Overall, rates of selecting and eating fruit and vegetable items did not differ greatly across race/ethnicities. No visible patterns were seen in aggregate production or plate waste data between schools with a greater percentage of Latino students (Table 3) and none of the logistic regression odds ratios showed statistical significance (Table 5). Our findings suggest that a significant proportion of students did not consume the fruits and vegetables offered as a component of their school lunch either because they did not select any fruits and vegetables or because they did not eat even a bite of them before throwing the lunch away. Production records showed that many vegetable and fruit items were prepared at lower rates.