Impact of school salad bars on fruit and vegetable selection, intake, and waste in Mid-Atlantic elementary schools

Background: Few studies have empirically examined the impact of school salad bars on elementary students' fruit and vegetable (FV) consumption within the National School Lunch Program (NSLP). This natural experiment evaluated the impact of salad bars on FV selection, intake, and waste within elementary schools. Methods: Seven school pairs, matched on Title I status and percentage of students from ethnic or racial minority backgrounds, were randomly selected. All schools served pre-portioned FV at baseline. One school within each pair received a salad bar; the other continued to serve pre-portioned FV (Control). Digital imagery plate waste methods were applied at baseline and 4-6 weeks after schools installed salad bars (post). Images were rated in the laboratory (ICCs = .94-.99) to determine FV selection, intake, and waste (servings [1 NSLP serving = 1/2c]). Multilevel modeling evaluated group (Salad Bar vs Control) and time (baseline vs post) differences and group-by-time interactions. Differences in outcomes by Title I status were also examined. Results: Across schools, mean NSLP participation was 54%. N = 6,623 trays were included (n = 3,273 Salad Bar; n = 3,350 Control). Students in Salad Bar schools selected (+ .44c) and consumed (+ .36c) more FV at post, compared to baseline. Control students decreased FV selection (-.05c) with no change in intake from baseline to post. Group, time, and group-by-time interactions were significant (ps < .0001). When examined separately, results suggest that these effects are driven by fruit. Salad Bar students increased fruit selection (+ .45c), intake (+ .36c), and waste (+ .09c) from baseline to post; no significant changes were observed in Controls. There was no significant change in vegetable selection, intake or waste for either group. Findings did not differ based on Title I status. Conclusions: Salad bars were effective in increasing elementary school students' fruit selection and intake, yet did not increase vegetable selection or intake. Additional efforts are needed to increase vegetable intake and minimize fruit waste from salad bars. Consistent findings across schools, regardless of Title I status, suggest potential for salad bars to yield increased fruit intake across socioeconomic groups. Findings can inform policies designed to increase FV intake within the NSLP. Trial registration: This investigation reports results of a systematic evaluation of school salad bars and does not meet criteria for a clinical trial, yet was retrospectively registered (10/28/22) in clinicaltrials.gov (NCT05605483) as an observational study.