key: cord-1000793-fkqj8qyi
authors: Cohen, Juliana F. W.; Hecht, Amelie A.; Hager, Erin R.; Turner, Lindsey; Burkholder, Kara; Schwartz, Marlene B.
title: Strategies to Improve School Meal Consumption: A Systematic Review
date: 2021-10-07
journal: Nutrients
DOI: 10.3390/nu13103520
sha: d0bebe36f0f672eb6b3e594bfb7b693aebd2137d
doc_id: 1000793
cord_uid: fkqj8qyi

School meals can play an integral role in improving children’s diets and addressing health disparities. Initiatives and policies to increase consumption have the potential to ensure students benefit from the healthy school foods available. This systematic review evaluates studies examining initiatives, interventions, and policies to increase school meal consumption. Following PRISMA guidelines, this review was conducted using four databases and resulted in a total of 96 studies. The research evidence supports the following strategies to increase school meal consumption: (1) offering students more menu choices; (2) adapting recipes to improve the palatability and/or cultural appropriateness of foods; (3) providing pre-sliced fruits; (4) rewarding students who try fruits and vegetables; (5) enabling students to have sufficient time to eat with longer (~30 min) lunch periods; (6) having recess before lunch; and (7) limiting students’ access to competitive foods during the school day. Research findings were mixed when examining the impact of nutrition education and/or offering taste tests to students, although multiple benefits for nutrition education outside the cafeteria were documented. There is some evidence that choice architecture (i.e., “Smarter Lunchroom”) techniques increase the proportion of students who select targeted meal components; however, there is not evidence that these techniques alone increase consumption. There were limited studies of the impact of increasing portion sizes; serving vegetables before other meal components; and strengthening local district and/or school wellness policies, suggesting that further research is necessary. Additionally, longer-term studies are needed to understand the impact of policies that limit students’ access to flavored milk. Several studies found increases in students’ meal consumption following the Healthy Hunger-Free Kids Act (HHFKA) and concerns regarding an increase in food waste following the HHFKA were not supported. Overall, there are a range of effective strategies to increase school meal consumption that can be implemented by schools, districts, and policymakers at the local, state, and federal levels (PROSPERO registration: CRD42021244688).

In the United States, approximately 95% of public and non-profit private elementary, middle, and high schools participate in the school meal programs administered by the United States Department of Agriculture (USDA) [1, 2] . Both the National School Lunch Program (NSLP) and School Breakfast Program (SBP) provide children and adolescents with healthy, low-cost meals throughout the school year. Prior to COVID-19, approximately 30 million children received school lunches daily, and roughly three-quarters of school lunch participants come from low-income households, with many relying on school meals for up to half of their daily energy intake [3, 4] . Schools are therefore uniquely positioned to promote healthy eating among children.

In 2010, the Healthy, Hunger-Free Kids Act (HHFKA) required the USDA to update many of its regulations for federal school nutrition programs to reflect the concurrent Dietary Guidelines for Americans [5] . Updates to the NSLP included requirements for more whole grain-rich foods; greater portion sizes for fruits and vegetables; a greater variety of vegetables offered throughout the week; limits on total calories and sodium; and the removal of trans fats. Nutrition standards were also strengthened for "competitive" foods, defined as snacks and beverages sold during school hours, but outside of the school meal programs (i.e., in vending machines, à la carte, snack bars, school stores, and fundraisers) [6] . Lastly, schools were required to update their local school wellness policies to promote healthier school environments [7] .

The evidence regarding school meal consumption suggests that rates are influenced by multiple intersecting factors, including at the individual, meal, cafeteria environment, and policy levels (Figure 1) . At the individual level, consumption varies by student age, gender, and eligibility for free or reduced-priced meals (based on family income), as well as factors such as preferences and prior exposure to foods [8] [9] [10] .

At the meal level, consumption consistently varies by food type (i.e., the federal meal components: meat/meat alternatives; grains; fruits; vegetables; and milk) with students typically consuming lower quantities of fruits and vegetables compared with the other components [11] [12] [13] . Although schools are not allowed to deviate from the federal meal pattern requirements, they may modify other meal service factors, such as offering students multiple choices of each required meal component; providing taste tests; and changing how the food is prepared (e.g., improving palatability/cultural appropriateness and preslicing fruits). Providing multiple food items from which students can choose increases the likelihood that one option will be appealing to a student. Increasing choices for students can occur on the lunch line (e.g., student can select between two fruit options) or through salad bars, which can hold a variety of produce. Palatability is also important to address as it is a primary determinant of food consumption, and familiarity with a food is strongly correlated with preferences [14, 15] . Therefore, focusing on culturally appropriate school meals that emphasize familiar flavors and taste may be a key component to ensuring high meal consumption rates. Interventions that involve collaborating with professional chefs or seasoning foods (both prior to serving or at "flavor stations" for students to add their own spices) have the potential to enhance the palatability and cultural appropriateness of school meals. Additonally, pre-slicing fruits may impact consumption because whole fruits, such as apples and oranges, may be challenging to consume if students have a limited amount of time to eat, perceive the fruit as messy, or have difficulty holding and biting the fruit (particularly among younger children or those with braces). Lastly, taste tests may provide the opportunity for repeated exposures to new foods, as well as a chance for students to provide feedback to cafeteria staff, thus influencing the options on the school menu [16] .

At the cafeteria environment level, modifications can potentially influence consumption via choice architecture strategies to nudge students towards healthier options; changing the length of the lunch periods to ensure students have sufficient time to eat; providing nutrition education; and scheduling recess before lunch. Choice architecture, also known as behavioral economics, uses strategies to "nudge" people towards the healthier options available in food environments [17] . Choice architecture strategies can include rewarding The problem of food waste in the NSLP has been documented for decades, and the consistently low levels of consumption for various food types, especially vegetables and fruits, remains a problem [24] [25] [26] [27] . However, a scoping review of research examining factors that impact school meal consumption has not been conducted. Several systematic reviews have examined techniques to nudge students towards healthier options (i.e., choice architecture), but many of the included studies only measured selection [28] [29] [30] [31] [32] [33] [34] . Other recent school-based reviews have been limited in scope, such as examining only fruits and vegetables [23, 35, 36] . A better understanding of initiatives, interventions, and policies that can improve school meal consumption more broadly is needed to help inform school food service programs and policies at the district, state, and federal levels. Therefore, the aim of this study was to systematically review the evidence regarding the impact of various strategies to improve school meal consumption.

The review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines [37] . This literature review was registered with the PROSPERO International Prospective Register of Systematic Review before data extraction (protocol registration number: CRD42021244688) [38] .

Articles were retrieved from four electronic databases: PubMed, Academic Search Ultimate, Education Resources Information Center (ERIC), and Thomson Reuters' Web of Science. The search strategy used varying combinations of the following keywords: school AND (meal OR breakfast OR lunch) AND (intake OR eat * OR consum * OR waste) AND (atmosphere OR behavioral economics OR breakfast in the classroom OR chef OR choice OR competitive food OR cultur * OR default OR duration OR engag * OR environment OR label * OR length OR marketing OR menu OR minutes OR nudg * OR nutrition education OR offer OR palatab * OR placement OR policy OR promotion OR repeat * OR slic * OR smarter lunchroom OR taste OR wellness). Articles in English published since the start of the literature through May 2021 were reviewed. Manual searches of the articles' reference lists, as well as a review of all articles citing the resultant literature (using Google Scholar), were conducted to identify other potentially relevant studies.

Eligible studies were quantitative research articles evaluating interventions, initiatives, and policies to influence school meal consumption. Our inclusion criteria were English, peer-reviewed publications or official government reports conducted in the United States within elementary, middle, and/or high schools participating in the USDA's SBP and/or NSLP during the academic year. The following types of articles were excluded: non-English articles; qualitative research; articles that did not examine school breakfast or lunch (e.g., only snacks or afterschool programs); studies conducted in private and/or charter schools not participating in the SBP or NSLP; initiatives that occurred outside of the school year (i.e., summer vacations); studies conducted in locations outside the United States; studies that did not quantify the amount of the entire school meal or school meal component consumed (i.e., assessing only school meal selection, binary (yes/no) estimates of consumption, estimates of only the nutrients consumed at lunch, and/or food frequency questionnaires or 24-hour recalls assessing only overall diets); and articles that did not examine initiatives, interventions, or policies (e.g., differences in consumption by age, gender, etc.). Additionally, studies examining universal free school meals and/or the Community Eligibility Provision were excluded as this was recently evaluated in a systematic review [39] . Titles and abstracts were screened by two independent reviewers, and full texts were screened by the lead author (JFWC) based on the eligibility criteria. The research team reviewed articles with unclear eligibility. The heterogeneity in the methods used to assess consumption and the variability in study designs precluded a meta-analysis; therefore, the results were synthesized in a narrative review.

Study quality and biases were assessed using the previously adapted Newcastle-Ottawa Scales (NOS) for cross-sectional and cohort studies, which are frequently used to assess non-randomized, community-based research [39] [40] [41] [42] . The studies were evaluated by two co-authors based on selection, comparability, and outcome and categorized as: low risk of bias (≥7 points), high risk of bias (4-6 points), and very high risk of bias (0-3 points) [43] . Supplementary Materials Table S1 presents the quality assessments of the included studies.

The initial search of the four databases (PubMed, Academic Search Ultimate, ERIC, and Thomson Reuters' Web of Science) identified 4423 articles. A total of 3367 duplicate records were removed, and primary screening excluded an additional 819 articles. The full text of the remaining 237 publications were examined in detail, and 151 studies were excluded. The primary reasons for exclusion were based on the study objectives (i.e., did not examine initiatives, interventions, or policies related to school meal consumption); the outcomes assessed (i.e., did not quantify the amount of school meals consumed or wasted); publication type (i.e., grey literature); and the study location (i.e., outside the United States). Ten additional articles were identified from the reference lists of the obtained articles or publications citing those studies, resulting in a total of 96 articles included in the review ( Figure 2 ). The quality scores for the articles included in the review ranged from 2 (very high risk of bias) to 10 (low risk of bias) (Supplementary Materials Table S1 ). Approximately a third of the included studies (n = 33) were classified as having a low risk of bias, over half of the articles (n = 55) had a high risk, and the remaining articles (n = 8) had a very high risk. The studies are presented by the different initiatives, interventions, and policies. Many studies included several components and are thus described in multiple sections. 

Of the 12 peer-reviewed publications that examined offering more options from which students could choose, the majority (n =8 studies) found significant positive associations with consumption; one found no association; one had mixed findings; and two found a decrease in consumption (Table 1) . Three studies were considered to have a low risk of bias, and of those, two found a positive association with consumption.

In a study using visual estimation in eight elementary and four middle schools in Houston, Texas, Cullen and colleagues found that when students had the ability to select from more fruit and vegetable options, elementary students consumed more vegetables (0.14 vs. 0.10 cups; p < 0.01), and middle school students consumed more whole fruits (0.19 vs. 0.10 cups; p < 0.001) and vegetables (0.17 vs. 0.10 cups; p < 0.01) [44] . Additionally, this study found significant increases in selection of these meal components. Using similar consumption measures, Just et al. examined the impact of increasing the number of fruit and vegetable options at lunch in 22 elementary schools and found that each additional option available increased the number of students who ate at least one serving of these meal components by 3.3 percentage points (p < 0.05) [45] . Hakim and associates also observed increases in both fruit and vegetable consumption by roughly 15% (p < 0.01) using weighed plate waste and visual estimation in one K-8 school when students were given a choice among three fruit or vegetable options [46] . Using the same consumption measure, Young and colleagues also found an increase in fruit and cooked vegetable consumption when students in one middle school had access to different fruits and vegetables every day of the week (this intervention also included exposure to increased health and physical education) [47] . In another multi-component study in 26 elementary schools in Minnesota, Perry et al. used visual estimation to examine the impact of offering multiple fruit and vegetable choices and found a 0.15 serving increase in combined fruit and vegetable consumption (p = 0.02). This intervention also included pre-slicing fruit and choice architecture techniques (e.g., visually appealing displays, verbal prompts, and rewards) [48] . A similar multi-component study by Greene and associates in seven middle schools in New York assessed consumption with visual estimation and found that multiple fruit and vegetable options (again, in addition to pre-slicing fruit and choice architecture techniques [e.g., attractive bowls and descriptive names]), was associated with a 23% increase in fruit consumption (p < 0.017), as well as a significant increase in selection. In the Green study, however, there was no impact on vegetable (or milk) selection or consumption [49] . While another study by Liquori et al. did find an increase in vegetable consumption (also measured using visual estimation) in two New York City elementary schools, this multi-component study only found an effect when students had a greater number of choices and cooking lessons, but not when choices were combined with traditional nutrition education compared with students in control schools [50] . Additionally, Ang and colleagues found no impact on fruit consumption using visual estimation when students in New York City elementary schools had access to multiple fruit options [51] .

The results among the four studies examining salad bars were mixed. In a study in two California elementary schools that included both salad bars and nutrition education (including gardening and cooking demonstrations), Taylor and colleagues found a significant increase in vegetable consumption measured using digital imagery: intake went from 0.09 cups pre-intervention to 0.15 cups post-intervention while consumption declined slightly in the control schools from 0.05 to 0.04 cups (p = 0.03) [52] . In contrast, Bean et al. found consumption decreased by 0.65 cups (p < 0.001) using digital imagery when students in two elementary schools in Virginia had access to a salad bar, although there was a significant increase in the number of fruits and vegetables selected (1.81 vs. 2.58; p < 0.001) [53] . Similar findings were observed in a cross-sectional study by Johnson et al. using 24-hour recalls among middle and high school students attending 21 schools in New Orleans, with less fruit consumption reported among students with a salad bar (and no impact on vegetable consumption) [54] . Interestingly, Adams et al. found that while salad bars alone did not impact the amount of fruits or vegetables that students ate, a greater number of options was associated with increased fruit and vegetable consumption using weighed plate waste in four elementary schools in San Diego, California [55] .

Of the nine peer-reviewed publications that examined palatability, six found a positive association with school meal consumption; one found no association; one had inconsistent findings; and one found an inverse association (Table 1) . Four studies had a low risk of bias; four had a high risk; and one had a very high risk. Of the studies with a low risk of bias, all found that enhanced palatability of school meals were positively associated with consumption.

Multiple studies have examined the impact of professional chefs in schools, and nearly all (four out of five) found positive associations with school meal consumption. In a pilot study conducted in four middle schools in Boston, Cohen and colleagues measured consumption using weighed plate waste; they found that students who received chefenhanced meals consumed 0.36 more servings of vegetables compared with students in control schools (p = 0.01) [56] . Additionally, this study found a significant increase in whole grain selection, resulting in more students consuming whole grains. In a study among 14 elementary and middle schools in Massachusetts using similar methods, Cohen et al. found an increase in both fruit (0.34 vs. 0.51 cups; p < 0.05) and vegetable (0.14 vs. 0.30 cups; p < 0.05) consumption among students in schools with a professional chef compared to students in control schools (as well as increased selection of these meal components) [57] . These findings of increased fruit and vegetable consumption were replicated by Cohen et al. in a third study conducted in eight elementary and middle schools in Massachusetts after implementation of the HHFKA [58] . A study conducted by Zellner and associates in two elementary schools in Philadelphia examined chef-prepared meals (in addition to familystyle service and an adult at the table as a role model) using visual estimation measurement and found that students consumed more of the targeted vegetables (sweet potato fries and cauliflower) [59] . Lastly, Just et al. conducted a study in one high school in New York with chef-enhanced pizza recipes (and taste tests) using visual estimation measurements; they found no impact on consumption, although increasing pizza consumption was challenging due to its very high baseline consumption rates [60] . However, an unintended consequence of this study was that more students received a side salad with the pizza, and vegetable consumption increased.

Other studies have used alternate strategies to enhance the palatability of the foods offered. D'Adamo and colleagues added spices and herbs to vegetable recipes and selected recipes based on taste tests with students. Using weighed plate waste, this study found the enhanced recipes were associated with an 18% increase in vegetable consumption (p < 0.001) [61] . Conversely, in another weighed plate waste study in 1 middle/high school in Pennsylvania, Fritts et al. found that when spices and herbs were added to vegetables, consumption decreased for several vegetables [62] . This study also found no impact on consumption of those vegetables with repeated exposures. Hamdi and associates used weighed plate waste in a study in three elementary schools in Illinois [63] . They evaluated a flavor station where students could add their own spices and seasoning to their lunches (in addition to taste tests and choice architecture techniques) and found inconsistent improvements with consumption with the components differing in effectiveness between the two participating schools, although the odds of vegetable selection were three times greater (95% CI, 1.3-6.5). [63] . Lastly, Bates et al. conducted a study in one middle and one high school in Utah where fruit smoothies were offered to students at breakfast; using visual estimation, they found offering smoothies was associated with a 0.45 serving increase in fruit consumption (p = 0.01) [64] .

Among the eight studies examining the impact of offering pre-sliced fruit, the majority (n = 6 studies) found a positive association with consumption, and two found no association (Table 1) . Four studies had a low risk of bias and the majorty (n = 3 studies) found pre-sliced fruit was associated with increased consumption.

In a study conducted by Smathers and colleagues in two elementary schools using aggregate plate waste, pre-slicing apples versus serving them whole was associated with increased consumption (2.48 times more apples by weight [p < 0.001]) [65] . Using similar consumption measures, McCool et al. also found that consumption increased when presliced apples were offered in an elementary school, and when middle school students were given the choice of pre-sliced or whole apples (some students preferred pre-sliced while others preferred whole fruit) [66] . Increased consumption of pre-sliced apples was also observed by Wansink and associates in six middle schools using visual estimation, with the percent of students eating at least half of an apple increasing by 73% (p = 0.02) [67] . In the previously described multi-component studies by Ang et al. and Greene et al., student consumption also increased when fruit was pre-sliced [49, 51] . Conversely, Swanson and colleagues examined both sliced apples and oranges in one elementary school in Kentucky, and found that orange (but not apple) consumption increased [68] . Lastly, two multicomponent studies found no association between serving pre-sliced apples and fruit consumption: Thompson et al. [69] used weighed plate waste in two elementary schools in Minnesota and Quinn et al. [70] used visual estimation in 11 middle and high schools in Washington. Among the six studies that measured selection, five found increases in selection with pre-sliced fruits.

Taste tests were one of the strategies used in nine multi-component studies, and were the sole intervention in one study. Four studies found positive associations with consumption; two had mixed findings; and four found no association (although in two of the studies, taste tests were inconsistently available in the participating schools [ Table 1 ]). Six studies had a low risk of bias, and most (4 out of 6 studies) found a positive association between taste tests and school meal consumption.

Mazzeo and colleagues examined the impact of taste tests (and choice architecture techniques [i.e., stickers as a reward for trying foods]) using visual estimation in two elementary schools and found they were associated with reduced fruit and vegetable waste by roughly 10% (39% waste in intervention schools vs. 52% in control schools [p < 0.05]) [71] . In a multi-component study also using visual estimation and conducted in six elementary schools in Utah, Morril et al. evaluated the impact of taste tests in addition to choice architecture techniques (e.g., a reward for eating fruits and vegetables) and found an increase in fruit and vegetable consumption [72] . Additionally, the previously mentioned multi-component study by Perry and colleagues found that taste tests combined with more fruit and vegetable choices and choice architecture techniques was associated with an increase in fruit and vegetable consumption [48] . Alaimo and colleagues examined taste tests and nutrition education using digital imagery in six elementary schools in Michigan and found increases in fruit consumption, but not in consumption of vegetables or other meal components [73] . Similarly, a government report examining taste tests and nutrition education in one elementary school in Oregon found increases in fruit consumption using weighed plate waste [74] . While no impact on vegetable or grain consumption were observed, this study found significant increases in their selection (as well as fruit selection). Another study with taste tests and nutrition education in 16 elementary schools in Little Rock, Arkansas was conducted by Blakeway et al. They found inconsistent results using aggregate plate waste; increases in consumption of whole wheat rolls and cottage cheese were observed in some (but not all) grades [75] . As previously noted, similarly inconsistent results were observed by Hamdi and colleagues in a study that evaluated taste tests in combination with a flavor station and choice architecture techniques [63] . No association was observed in three studies that involved taste tests and nutrition education in 28 elementary schools in Alabama by Reynolds and associates using visual estimation [76] , and in multiple elementary schools in Wyoming by Bontranger Yoder and colleagues using digital imagery [77, 78] , although taste tests were not available in all the participating schools in the latter two studies. Similarly, no impact on consumption was observed in the study by Just and colleagues when examining taste tests and chef-enhanced pizza recipes [60] .

Of the 23 peer-reviewed publications that primarily examined choice architecture, roughly half (n = 13 studies) found a positive association with school meal consumption; seven studies found no association; and three studies found an unexpected increase in food waste (Table 2) . Among the five studies with a low risk of bias, four were positively associated with consumption while one found an increase in food waste. When examining the ten multi-component studies (e.g., including nutrition education, taste tests, choices, and pre-sliced fruit), half found a positive association and half found no association. Seven of these studies had a low risk of bias and the results were also inconsistent. Overall, these mixed findings may be due in part to the fact that there are many different choice architecture techniques and success may vary by the specific strategy implemented.

Eleven studies tested the impact of providing rewards for students within elementary schools. Among those that required students to eat the fruits and vegetables prior to receiving the reward, nearly all (9 out of 10) found a positive association with fruit and vegetable consumption. The one study that only required selection (but not consumption) for a reward found no association with consumption. Hendy and colleagues conducted a study using visual estimation in one elementary school in Pennsylvania and found that providing children with a small prize for eating at least 1/8 cup of fruits and vegetables at lunch was associated with an increase in consumption [79] . Using the same measure of consumption, the previously described study by Perry and colleagues found that rewards (i.e., frozen yogurt for the classroom), as well as taste tests and more fruit and vegetable choices, were associated with an increase in fruit and vegetable consumption [48] . Jones et al. used aggregate plate waste in one elementary school in Utah and found that teacher encouragement to eat more fruits and vegetables combined with rewards (medals and points displayed on a board) was associated with a 39% increase in fruit consumption (p < 0.01) on the days when fruit was targeted and a 33% increase (p < 0.05) in vegetable consumption on the days when vegetables were targeted. Notably, neither fruit nor vegetable consumption increased when the other meal component was targeted [80] . In a study using digital imagery measurement in one Utah elementary school, Wengreen and colleagues found an increase in fruit and vegetable consumption when students received a reward for trying those meal components; this intervention also included educational videos and letters read by teachers promoting fruits and vegetables [81] . Using similar consumption measures in one elementary school in Oregon, Machado and colleagues provided verbal prompts, adult role modeling, and rewarded students for consuming fruits and vegetables with a classroom party and t-shirts. They found that the proportion of students consuming all their fruits increased by 11% (p < 0.01) and the proportion eating all their vegetables increased by 8.7% (p < 0.01) [82] . Increases in fruit and vegetable consumption were also observed in the previously described multi-component study conducted by Morrill et al., and decreases in waste for fruits and vegetables were observed in the study by Mazzeo et al.; both studies included taste tests as well [71, 72] .

In two studies conducted by Hoffman et al. within four elementary schools in New England, there were promotional posters, verbal encouragement, and students received a small prize (e.g., stickers) for eating fruits and vegetables. Both studies found an increase in consumption of fruits and vegetables using weighed plate waste [83, 84] . However, they found the results were not sustained after the intervention concluded [84] . In another study, Blom-Hoffman et al. used visual estimation in one elementary school and found no impact on vegetable consumption when examining the impact of nutrition education combined with sticker rewards and verbal praise [85] . Hudgens and associates used visual estimation measurement in one elementary school in Oregon and found that providing a reward for selecting (but not requiring student to taste) a meal with a fruit, vegetable, unflavored milk, and whole grain was not associated with an increase in consumption, although increases in selection of these components were observed [86] .

Sixteen studies examined the visual appeal of the cafeteria environment-such as attractive bowls, signage with creative names-as well as verbal prompts, and/or location of fruits/vegetables. Among these studies, only four found a positive association with consumption; three found an increase in food waste; eight found no association with consumption; and one found inconsistent results.

In a study conducted by Gustafson et al., children in four elementary schools in Nebraska designed posters marketing vegetables that were then displayed over a salad bar. They used digital imagery to measure plate waste and observed an increase in vegetable consumption [87] . Adams et al. used weighed plate waste for a study in middle schools in Phoenix, Arizona and tested the impact of moving the location of the salad bars. They found that students consumed 4.82 times more fruits and vegetables when the salad bar was accessible from the serving line compared with when salad bars were located after students had left the lunch line [88] . In the previously described multi-component study by Greene et al., attractive bowls and descriptive names were combined with providing pre-sliced fruit and increased fruit and vegetable choices. They found an increase in fruit consumption, but no impact on vegetable or milk intake [49] . Hanks et al. found that an intervention that included using attractive bowls and descriptive names in combination with changing the placement of foods, providing verbal prompts, and creating healthy convenience lines was associated with an 18% increase in fruit consumption (p = 0.004) and 25% increase in vegetable consumption (p < 0.001) in two middle/high schools in NY using visual estimation [89] .

Although the previous studies found impacts on student consumption, these findings were not replicated in other studies and many studies found null effects. Using consump-tion measures similar to Hanks et al. [75] , Quinn and associates used Smarter Lunchroom techniques (and pre-sliced fruit) and found no impact on fruit and vegetable intake in 11 middle and high schools in Washington [70] . Similarly, Thompson et al. found no impact of these interventions (i.e., Smarter Lunchroom strategies and pre-sliced fruit) on intake in two elementary schools in Minnesota using weighed plate waste [69] . Additionally, Cohen et al. used weighed plate waste to assess similar choice architecture techniques, as well as more prominent white milk (compared with chocolate milk) in 14 elementary and middle schools in Massachusetts, and found that these strategies had no impact on the consumption of fruits, vegetables, or milk [57] . Goto and colleagues increased the quantity and prominence of white milk in three elementary schools in California, and found no association with milk consumption using weighed plate waste [90] . Hanks and colleagues examined the exposure of only a healthy convenience line in a high school in New York using weighed plate waste and also found no impact on healthier meal consumption [91] . Additionally, placing vegetables first on the lunch line had no impact on consumption in a study conducted by Ang and colleagues in 14 elementary schools in New York City using visual estimation [51] . Promotional signage, a more open lunch line, wall art, and comfortable seating options was associated with reduced consumption of fruits and grain and no impact on vegetable consumption after a year of exposure in a study conducted by Koch et al. in seven high schools in New York City using digital imagery [92] . In a study conducted by Blondin et al. in six elementary schools using weighed plate waste, verbal encouragement in the classroom was found to increase milk waste [93] . This study also found that milk consumption was adversely impacted by the presence of juice. Schwartz et al. also examined verbal prompts (on the lunch lune) in two elementary schools in Connecticut using visual observation and found that while there were no differences in individual-level consumption rates between intervention and control schools (~70% consumption in both schools), there were significant increases in selection, and thus 70% of children consumed a serving of fruit in the intervention schools compared with less than 40% in the control schools [94] . As previously mentioned, inconsistent results were observed by Hamdi and colleagues in a study that evaluated choice architecture (in addition to taste tests and a flavor station) [63] , and another study by Wansink et al. using visual estimation in one high school with posters and student involvement in growing the salad greens was associated with a decrease in salad consumption [95] . Similarly, Reicks and colleagues found that including pictures of vegetables on lunch trays was associated with a decrease in carrot consumption (and no association with green beans) using aggregate plate waste in one elementary school in Minnesota [96] . Importantly, nearly all (n = 12) studies examining Smarter Lunchroom Strategies found increases in selection.

A small number of publications examined portion sizes (2 studies) and providing meal components at different times (3 studies). While all found positive associations with consumption, only one study had a low risk of bias. Among studies examining portion sizes, Ramsay et al. increased the portion size of chicken nuggets provided to kindergarteners in one school and found increased consumption (and selection) using weighed plate waste [97] , although processed foods such as chicken nuggets tend to have high consumption rates more broadly in schools [98] . In another study examining portion sizes, Miller and colleagues increased the portion sizes for fruits and vegetables by 50% and found a corresponding increase in consumption by about 13-42 g using weighed plate waste in one elementary school in Minnesota [99] . In this study, selection of some components increased while other components decreased. Two other studies conducted in an elementary school in Minnesota found that offering vegetables (i.e., red peppers, baby carrots, and broccoli) first while students waited in the lunch line was associated with increases in consumption using visual estimation (Elsbernd et al.) [100] and aggregate plate waste (Redden et al.) [101] . However, only the study by Eslbernd et al. found increases in selection as well. Similarly, in one elementary school in Philadelphia, Zellner and colleagues found that when offering fruit was delayed until later in the meal, students consumed more kale compared with when fruit was served at the same time as the vegetable (and entrée and milk) using visual estimation [102] .

Of the 11 studies focused on nutrition education alone, or with small additional components, such as parent newsletters, slightly over half (n = 6 studies) found a positive association with school meal consumption; four found no association; and one found a decrease in consumption (Table 3) . Five studies had a low risk of bias, and the majority (n = 4 studies) found positive associations between nutrition education and consumption. Among the 10 multi-component studies that combined nutrition education with other strategies (e.g., taste tests, more food choices, and choice architecture techniques), slightly fewer than half (n = 4 studies) found increases in consumption; four found no association; and two found mixed results. Among the three multi-component studies with a low risk of bias, only one found a positive association with consumption.

In one of the studies that found a positive association between nutrition education and consumption, Sharma et al. implemented a 16-week nutrition education program in three elementary schools in Texas. This intervention also included a parent component with recipes and demonstrations and fresh fruit from local pantries sent home with families. They used weighed plate waste and found a decrease in fruit and vegetable waste at lunch (β = −32.06; p < 0.001), and while no differences in selection were observed in the intervention schools, significant decreases in selection were observed in the control school [103] . In a study conducted in four elementary schools in Denver by Auld and colleagues, students were exposed to 16 nutrition lessons taught alternatively by teachers and special resources teachers [104] . This study examined food consumption using visual estimation and found that nutrition education was associated with an increase in fruit and vegetable consumption by 0.36 servings (p < 0.001). Auld et al. conducted another study using similar methodology in 10 elementary schools in Denver, but with 24 nutrition lessons taught exclusively by a special resources teacher and found a similar effect size for improvements in fruit and vegetable consumption (0.4 servings; p < 0.001) [105] . A study using visual estimation in two elementary schools in Minnesota conducted by Burgess-Champoux and colleagues focused on whole grains. This intervention included five lessons, cafeteria staff culinary training (e.g., quality control, staff taste tests, etc.) and a parent component (e.g., supermarket/bakery tours, newsletters). They found an average of a one serving increase in whole grain consumption (p < 0.0001) [106] . In a study conducted by Head, marginally significant (p = 0.05) increases in overall consumption were observed using weighed plate waste in four North Carolina elementary schools exposed to nutrition education with a focus on basic nutrition and dietary patterns [107] . Additionally, higher rates of vegetable consumption were observed using digital imagery among 6th grade students exposed to a curriculum that focused on sustainable food systems in a study conducted by Prescott et al. [108] . Increases in selection were also observed in the intervention schools in this study, however they were not statistically significant. Lastly, greater consumption was also observed in four previously described multi-component studies that combined nutrition education with cafeteria components including increases in fruit and vegetable choices [47, 52] and taste tests [73, 74] .

Inconsistent results were found in a multi-component study conducted by Blakeway et al. that included both nutrition education and taste tests; improvements were observed in some, but not all, grades [75] . Additionally, mixed findings were observed by Liquori and associates with a nutrition education program in two New York City elementary schools; while lessons that focused on food and the environment had no impact on consumption (using visual estimation), nutrition education that included a cooking component improved consumption of vegetables and whole grains, but only among younger students [50] . Jones et al. also evaluated a curriculum that involved both nutrition and agriculture, as well as farm to school activities (e.g., school gardens, field trips, cooking demonstrations) in 18 elementary and middle schools in South Carolina and found a small decrease in fruit consumption (−0.07 servings; p < 0.05) and no significant impact on vegetable consumption using digital imagery [109] . Another study conducted by Ishdorj et al. examined lunch consumption based on 24-hour recalls from a nationally representative sample of public schools and found that nutrition education was not associated with fruit or vegetable consumption [110] . No impact was observed in four other studies conducted in elementary schools with nutrition education that all specifically focused on fruit and vegetable consumption, with intake measured using visual estimation [85, 111, 112] or digital imagery [113] . Lastly, no associations were found in three other previously described multi-component studies that incorporated both nutrition education and taste tests [76] [77] [78] .

Of the four peer-reviewed publications that examined school lunch duration, three found a positive association between duration and school meal consumption and one found no association (Table 4) . Three studies had a low risk of bias and one had a high risk of bias; of the studies with a low risk of bias, two found a positive association with school meal consumption.

In a study conducted by Bergman et al. using weighed plate waste in two elementary schools in Washington, students with a 30 min lunch period consumed on average 72.8% of their meal, while students who had a 20 min lunch period consumed on average only 56.6% of their meal (p < 0.0001) [114] . In a similar study conducted by Cohen and colleagues using weighed plate waste in six elementary and middle schools in Massachusetts, students who had less than 20 min of seated time for lunch consumed on average between 10-13% less of their meal compared with students who had more than 25 min to eat lunch (corresponding with a 30 min lunch period; p < 0.0001) [19] . This study also found that students were more likely to select a fruit when they had more time to eat. Using digital imagery to assess consumption, Gross et al. examined 10 elementary schools in New York City and found that lunch periods that were at least 30 min long were associated with an increased odds of consuming fruits (OR = 2.0; p = 0.02) and whole grains (OR = 2.1; p < 0.05) [115] . Lastly, in another study conducted in New York City among elementary school students, Ang et al. found that lunch period duration was not associated with consumption [51] . However, less than 15% of the students had schedules with lunch periods greater than 20 min, which may have limited the ability to detect significant differences.

Of the 10 peer-reviewed publications that examined recess before lunch (Table 4) , seven found a positive association with school meal consumption, one found mixed results with different meal components, and two found no association. Of the studies examining the timing of recess that were considered to have a low risk of bias, the majority (3 out of 4) found positive associations with school meal consumption for differing meal components.

Seven studies examining recess before lunch used weighed plate waste methodology. Bergman and colleagues examined two elementary schools in Washington and found that students who had recess before lunch consumed on average 72.8% of their meal compared with students who had recess after lunch and consumed on average 59.9% of their meal (p < 0.0001) [116] . In a study conducted by Chapman et al. among students attending eight elementary schools in New Orleans, recess before lunch was associated with a 5% increase in fruit consumption, but no differences were observed for other meal components [117] . This study also found that students with lunch periods early or late in the day consumed less than students with traditional lunch hours. Getlinger and associates also found overall reduction in food waste from 34.9% to 24.3% in one elementary school in Illinois [118] Strohbehn et al. used a combination of weighed plate waste and digital imagery in three elementary schools, similarly finding reduced waste for fruits, as well as grains and meat/meat alternatives, but found increases in waste for vegetables [119] . Two additional studies, conducted by Hunsberger et al. in one elementary school in Oregon and McLoughlin et al. in two elementary schools in Illinois, found recess before lunch was associated with greater milk consumption, but found no association with other meal components [21, 120] . Lastly, Tanaka et al. examined one elementary school in Hawaii and found no association with school meal consumption using weighed plate waste [121] . Among studies using other consumption measures, research conducted by Ang et al. among elementary students in New York City examined consumption with visual estimation and found recess before lunch was associated with small increases in the amount of fruits (0.08 cups; p < 0.001) and vegetables (0.007 cups; p = 0.04) eaten [51] . Another study conducted by Price et al. in seven elementary schools in Utah using visual estimation to assess diet also found that recess before lunch was associated with greater fruit and vegetable consumption (0.16 servings; p < 0.01) [122] . However, a study conducted by Fenton and associates using diary assisted 24-hour recalls among students attending 31 elementary schools in California found no association between the timing of recess and fruit and vegetable consumption at lunch [123] .

Out of the six studies that examined the impact of the HHFKA, most (4 of 6) found increases or no impact on consumption, with two studies finding a decrease in intake for certain meal components (Table 5 ). Among the studies with a low risk of bias, two out of three found that this federal policy was associated with increases in school meal consumption.

Cohen and colleagues measured consumption using weighed plate waste in four elementary and K-8 schools in Massachusetts and found a 15.6% increase in entrée consumption (p < 0.0001) and a 16.2% increase in vegetable intake (p < 0.0001) [24] . Additionally, significantly more students selected fruit, resulting in more students consuming this meal component. Consistent results were found by Schwartz et al. in 12 middle schools using the same measure of consumption; entrée consumption increased from 71% to 84% (p < 0.05) and vegetable consumption increased by roughly 20% (p < 0.05) [25] . Greater fruit selection was also observed in this study. Ishdorj and colleagues examined consumption using aggregate plate waste in three Texas elementary schools, and found no impact on consumption [98] . Similarly, Bontranger Yoder et al. found no association with school meal intake using digital imagery in 11 elementary schools in Wisconsin [78] . Cullen et al. found no association with fruit, vegetable, or entrée consumption, and a decrease in milk consumption using visual estimation after implementation of the HHFKA, although they observed significant increases in the selection of fruits, vegetables, legumes, and protein foods [26] . Contrasting with the results of the other studies, Amin et al. found a small decrease in fruit and vegetable consumption by roughly 0.05 cups (p < 0.01) in two elementary schools (using both digital imagery and weighed plate waste) after the HHFKA went into effect, although increases in fruit selection were observed [124] .

Six studies examined the impact of limiting student access to competitive foods (either via state or local policies or based on student purchases) and nearly all (5 of 6) found that consumption of school meals was inversely associated with access to competitive foods. Only one study had a low risk of bias and found that students consumed more of their healthier school meals when access to competitive foods was limited.

Cullen and colleagues examined Texas elementary schools without access to competitive foods and a middle school where students could purchase snacks/beverages and observed that the elementary students without access to competitive foods consumed significantly more fruits and vegetables than the middle school students based on selfreported measures of intake [125] . Additionally, they found greater selection of fruit, vegetables, and milk when students did not have access to competitive foods or beverages. Similarly, Cullen and colleagues examined the transition from elementary school to middle school and found that students consumed significantly fewer fruits and vegeta-bles and less milk when they entered middle school and had new access to competitive foods [126] . Marlette et al. used weighed plate waste in three Kentucky middle schools and found similar decreases in school meal consumption when students purchased competitive foods [127] . Using a nationally representative sample with 24-hour recall data, Ishdorj et al. found that students attending schools with policies that placed restrictions on the sale of competitive foods and desserts reported greater vegetable consumption. They also found that when policies were in place limiting access to higher fat milk (whole or 2%), students reported increases in both fruit and vegetable consumption. In contrast, students reported reduced fruit intake when there were policies limiting French fries [110] . When Cullen and associates used lunch food records to examine a Texas state policy that strengthened the standards for competitive foods, they found that students in three middle schools reported greater vegetable and milk consumption [128] . Of note, when a local policy in Texas only limited access to vending machines in the cafeteria (but vending machines with competitive foods were available in other parts of the school), Cullen et al. observed that students in three middle schools compensated by increasing their competitive food purchases from the other vending machines. Students also increased their milk consumption but decreased their vegetable consumption [129] .

Limiting flavored milk is a policy that has been addressed in some locations. Four studies have examined the impact of local policies limiting access to chocolate milk, and among the two studies with a low risk of bias, one found increases in waste and one found no impact. Among the three studies examining other local policies, two studies with a high risk of bias found an increase in consumption and one study with a low risk of bias had mixed findings.

Cohen and colleagues examined weighed plate waste in four elementary/K-8 schools in Massachusetts where access to chocolate milk was limited and found a 10% decline in unflavored milk consumption within the first year of implementation (as well as reduced selection of milk) [24] . Greater unflavored milk waste was also observed by Hanks et al. using visual estimation in 25 elementary schools after chocolate milk was removed [130] . Among the studies with longer exposure to the policy (i.e., two years), Cohen et al. found that students adjusted to the policy and there was no adverse impact on average milk intakes (nor selection) in four Boston middle schools using weighed plate waste [56] . While Schwartz et al. found that individual-level consumption of milk was lower using weighed plate waste in two K-8 schools after two years of exposure to a policy that removed chocolate milk, importantly this study found that school-wide per-student consumption significantly increased from 2.1oz to 2.5oz due to the large increases in selection (from 51.5% to 72%) [131] . Interestingly, this study also found that milk consumption was adversely impacted by offering juice on the lunch menu.

Another practice included in local policies is serving breakfast in the classroom. Farris et al. examined breakfast in the classroom within seven elementary schools in Virginia and found that the policy was associated with decreases in food waste from 43% to 38.5% of overall meals discarded [132] . Another study mentioned earlier assessed the impact of a wellness policy that included provisions for more fruit and vegetable choices, as well as an increase in contact hours for health and physical education. This was associated with increases in fruit and vegetable consumption [47] . Lastly, a study conducted by Canterberry and associates in seven elementary schools in New Orleans assessed a local policy that required an increase in fresh and less-processed ingredients, as well as a restriction on deep-fried foods. There were three food service providers and the findings were mixed: some lower intakes were observed in the schools with one of the food service providers, but there were no associations with consumption in the schools with the two other food service providers [133] . (1) Choices: The number of vegetable and whole grain options available was increased (2) Nutrition Education: 10 food and environment lessons and/or 10 cooking lessons were provided (3)

Other: Students took field trips to a local community garden. There was parent outreach (newsletter, recipes, workshops)

Selection: Not measured Consumption: The intervention arm with cooking lessons was associated with increased consumption of vegetables and whole grains among younger students (p < 0.01). No association was observed among older children exposed to the cooking intervention. The nutrition education (food environment) intervention was not associated with consumption. Selection: Pre-slicing oranges was associated with increased selection. No significant differences were observed with pre-sliced apples.

Consumption: A greater proportion of students consumed at least half of a fruit serving when pre-sliced (vs. whole) oranges were served. There were differences observed by grade level (i.e., higher consumption rates among younger students with pre-sliced oranges). The intervention was not associated with consumption for apples. Selection: There was a significant increase in apple selection in intervention schools when pre-sliced apples were offered (5% difference in sales between intervention and comparison schools; p < 0.001).

Consumption: Pre-slicing apples was associated with an increase in the percent of students who selected an apple and ate more than half (75% increase; p = 0.02). Weighed plate waste Selection: Selection was measured in only one of the participating schools. The odds of selecting a vegetable (i.e., broccoli) increased when students were exposed to taste tests. The odds of selecting fruit increased when all intervention components were implemented simultaneously.

Consumption: The intervention components yielded inconsistent, but generally positive consumption results across the schools, particularly for fruits. Table S1 ). (1) Choice Architecture: Vegetables were pre-plated (vs. optional), and were placed first in line (2) Pre-sliced Fruit (3)

Recess before lunch (4) Choices: multiple (2+) fruit and vegetable options were provided (5) Lunch Duration

Selection: Not measured Consumption: Pre-plating vegetables (vs. optional for the student to select a vegetable) was associated with a small increase in consumption (0.02 cups; p < 0.001).

Positioning vegetables first on the serving line was not associated with vegetable consumption. Among students who selected fruit, pre-sliced fruit was associated with greater consumption (0.23 cups more p = 0.02) than whole fruit. Recess before lunch was associated with a small increase in fruit consumption (0.08 cups; p < 0.001) and vegetable consumption (0.007 cups; p = 0.04). Multiple fruit options and attractive serving bowls were not associated with fruit consumption. Lunch duration was not associated with consumption (although less than 15% of measurements had lunch durations of ≥20 min). (1) Choice Architecture: Fruits were placed first on the lunch line and were in attractive bowls with descriptive names. Promotional materials (e.g., fruit facts) were posted in the cafeteria (2) Choices: Multiple (2) fruit and vegetable options were provided (3) Pre-sliced Fruit

Selection: The intervention was associated with a 36% increase in fruit selection (p < 0.001), but no significant changes in vegetable or milk selection. Consumption: The intervention was associated with a 23% increase in fruit consumption (p < 0.001). There was no association with vegetable or milk consumption.

Gustafson et al. 2017 [87] Kearney, N; 4 elementary schools (1614 trays from students in grades K-5) Consumption: The intervention arm with both student participation in the poster design and the presence of marketing was associated with a significant increase in vegetable consumption compared with the control group. Table S1 ). (1) Nutrition Education: Five lessons that focused on whole grains were implemented by trained research assistants (2) Other: There was a family component (newsletters, supermarket and bakery tours, and an event at a milling museum). To increase the availability of whole grains in the cafeteria, there was culinary training on whole grains for cafeteria staff focused on menu planning, procurement, quality control and staff taste tests.

Selection: Not measured Consumption: In the intervention school, whole grain consumption increased by 1 serving (p < 0.0001) and refined grain consumption decreased by 1 serving (p < 0.001) compared with the control school. (1) Nutrition Education: Nutrition and agriculture content was integrated into classroom curriculum (2) Other: Schools incorporated farm to school activities, including a school garden, field trips to farms, cooking demonstrations and providing local items on the cafeteria menu Digital imagery Selection: Not measured Consumption: Students in the intervention schools consumed on average less fruit than students in control schools. There was no significant association with vegetable consumption. Selection: There was no change in selection among intervention schools but there was a decrease in the comparison schools, which resulted in a significant difference in selection.

Consumption: The intervention was associated with significant decreases in F/V waste at lunch (p < 0.001). (1) Nutrition Education: A farm to school gardening curriculum was available in some of the participating schools (2) Taste tests: Available in some of the participating schools (3)

Other: Farm to school activities were introduced (e.g., a school garden, field trips to farms, and local items on the menu) in some of the participating schools Digital imagery Selection: Not measured Consumption: The intervention was not associated with differences in F/V consumption, although the farm to school components were inconsistently implemented across the participating schools. (1) Choices: The number of vegetable and whole grain options available was increased (2) Nutrition Education: 10 food and environment lessons and/or 10 cooking lessons were provided (3) Other: Students took field trips to a local community garden. There was parent outreach (newsletter, recipes, workshops)

Selection: Not measured Consumption: The intervention arm with cooking lessons was associated with increased consumption of vegetables and whole grains among younger students (p < 0.01). No association was observed among older children exposed to the cooking intervention. The nutrition education (food environment) intervention was not associated with consumption. Table S1 ). Table S1 ). Consumption: Policies that place restrictions on the sales of competitive foods were associated with greater fruit consumption. Policies that restrict desserts were associated with greater vegetable consumption. Policies that limited French fries were associated with lower fruit consumption. Limiting whole and 2% milk was associated with greater fruit and vegetable consumption. Nutrition education and policies that increase the amount of fresh fruit and vegetable available daily at lunch policies were not associated with consumption of fruits or vegetables. Selection: Significantly fewer students selected milk when juice was available. There was approximately a 20 percentage point increase in milk selection in the second year of the policy compared with the in the first year.

Consumption: Among students who selected milk, milk consumption was lower in the second year of the policy compared with in the first year. On days when juice was offered, students consumed significantly less milk (at both time points).

Young et al. 2013 [47] N.S.; 1 middle school (3810 trays from students in grades 6-8)

Cross-Sectional 2011-12

Policy: A new wellness policy required schools to implement the practices below:

(1) Choices: Different fruits and vegetables were served each day of the week (2) Nutrition Education: There was an increase in student contact hours for health and physical education Visual estimation Selection: Not measured Consumption: After exposure to the wellness policy for over a semester, students consumed significantly more fruits and cooked vegetables. Table S1 ).

To our knowledge, this is the first systematic review of the literature that comprehensively examined initiatives, interventions, and policies associated with school meal consumption. The factors were at the school meal level (choices, food preparation, and taste tests); the cafeteria environment level (choice architecture, nutrition education, school lunch duration, and recess before lunch); and the policy level (local, state, and federal policies). The findings suggest that several practices are consistently associated with improved meal consumption. These include: (1) offering students choices within each meal component (particularly fruits and vegetables); (2) enhancing the palatability/cultural appropriateness of meals; (3) pre-slicing fruit; (4) incentivizing students to taste fruits and vegetables with rewards; (5) providing more time to eat lunch with longer lunch periods; (6) implementing recess before lunch; and (7) limiting access to competitive foods.

Many of these strategies can be implemented with minimal costs and/or additional labor. First, providing students with 30-min lunch periods can help to ensure that students have enough time to eat after accounting for time spent getting to the cafeteria and waiting in the lunch line. While some schools may be apprehensive to replace academic time with longer lunch durations, in prior research, teachers have reported that students are more attentive in class (and thus more efficient learners) with this policy and that this time in the cafeteria can be valuable for social and emotional learning [134] . Another concern that has been raised regarding longer lunch periods is disruptive student behavior in the cafeteria if they finish their meals quickly; however, implementing recess before lunch has been found to create a calmer lunchroom environment, suggesting that combining these two strategies would be beneficial for consumption and student behavior [134] .

Offering more fruit and vegetable choices (especially with pre-sliced fruits such as apples and oranges) was also found to be an effective method to increase consumption. Cost-effective strategies to increase the amount and variety of fruit available in schools include using USDA commodity foods; integrating foods from school gardens; and procuring locally grown produce. The USDA Department of Defense Fresh Fruit and Vegetable Program also provides fresh produce for schools. Restricting access to competitive foods can also be implemented in a cost-neutral manner; prior research has documented that when students have limited access to competitive foods, they instead participate more in the NSLP [135] . These increases in school meal sales have been found to offset revenue losses from the decreased sales of snacks and beverages [135] . While chef-based initiatives to enhance the palatability and cultural appropriateness of meals can be expensive, partnering with volunteer chefs from local restaurants can be a cost-effective solution. Additionally, schools can hire a chef when an existing cafeteria staff member retires, thus enabling this to be a more affordable strategy. Lastly, various free resources are available to schools through the USDA, state departments of educations, and non-profit organizations such as those that provide recipes and culinary workshops.

The findings from this review also highlight some strategies with mixed empirical support. Nutrition education was found to have mixed results when examining its impact on consumption in the cafeteria. This may have occurred due to several reasons. First, the nutrition education curricula across the studies varied in intensity, dose, and content. It is possible that greater time is necessary to dedicate to nutrition education, as well as more staff development and training to have improved effectiveness [136] . Second, it is possible that nutrition education is necessary but not sufficient on its own to reliably influence consumption, especially if there are other obstacles, such as insufficient time to eat. Although it may be possible to improve student intake in the short term without incorporating nutrition education, understanding how our food choices affect our health should still be considered an integral component of the Whole School, Whole Child, Whole Community model and may have several other benefits outside of the lunchroom and over the long term as students become more independent and make their own food choices [136, 137] .

The Smarter Lunchroom movement has gained traction in the United States with over 29,000 schools implementing these techniques to nudge students towards the healthier options available. Somewhat surprisingly, the results of this systematic review found that some of the commonly cited Smarter Lunchroom techniques had little to no impact on school meal intake [89] . One exception was when students were rewarded with small prizes such as stickers, classroom parties, or t-shirts for eating school meal components. While there may be additional concerns that providing rewards for a preferred behavior may potentially decrease the motivation and/or create negative associations with performing that activity, parenting research more broadly has found inconsistent results [138] . However, this strategy may be burdensome and expensive to implement and maintain. Additionally, the results were mixed when examining the sustained impact on consumption after the interventions concluded. Importantly, the more traditional techniques (e.g., attractive bowls, verbal prompts, creative names, etc.) have generally been found to be effective at improving the selection of certain meal components, especially fruits and vegetables, in prior systematic reviews [29] [30] [31] [32] . Therefore, schools should not abandon these practices, but also not rely only on these approaches to improve students' school meal intakes. Future studies should examine the combined effect of choice architecture techniques with other potentially successful strategies, such as longer lunch periods, recess before lunch, enhanced palatability/cultural appropriateness of meals, and limited access to competitive foods.

This study has several limitations. First, many studies had a high or very high risk of bias based on NOS scores. Conducting studies with sufficiently large sample sizes, longer periods of observation, obtaining informed consent from students for repeated data collection, and using validated dietary assessment methods can be expensive, highlighting the need for additional resources and grant funding for researchers. It is also noteworthy that when examining the studies with a low risk of bias, the conclusions did not change. Second, publication bias may have been an issue; however, a substantial number of studies examined found no significant associations with school meal intake. Third, the measures of school meal consumption varied between the studies, and some methods may not have been sensitive enough to detect the levels of change often observed in school-based interventions. However, multiple strategies were still found to improve intakes using the various measures. Additionally, while many studies of breakfast in the classroom have been conducted, most were excluded as they were evaluating universal school meals and not the independent impact of breakfast in the classroom. Therefore, future studies should examine this policy to determine whether it has additional benefits beyond universal free breakfast in the cafeteria. Lastly, the studies included in this systematic review were conducted only in the United States. While the findings are likely generalizable to other economically developed countries with strong nutrition standards, future studies should examine these strategies in other locations. A strength of this systematic review was the large number of studies evaluated. In addition, this review began with the conceptual framework outlined in Figure 1 and had strict consumption measurement criteria (described in Table 1 ).

Overall, this review suggests that many strategies have the potential to improve school meal consumption. The majority of studies in the current review found improvements in school meal intake when students were provided with multiple choices on the lunch line; pre-sliced fruits; recipes that focused on improving the palatability and cultural appropriateness of the foods offered; longer lunch period; recess before lunch; limiting access to competitive foods; and providing incentives for students to taste the fruits and vegetables offered at lunch. However, commonly used Smarter Lunchroom techniques were not found to be an effective strategy to increase intake of school lunch. While findings were mixed regarding nutrition education's impact on meal consumption, research has found other benefits of nutrition education to students. Further research is needed regarding school wellness policies and other district-level policies, including limited access to chocolate milk. Importantly, the results suggest that weakening the HHFKA would not be an effective strategy to reduce school meal waste. Instead, school districts and policy makers should consider the multiple strategies found to improve school meal consumption.

Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/nu13103520/s1, Table S1 : Quality Assessment for Cross-Sectional, Cohort, and Quasiexperimental Studies based on the Newcastle Ottawa Quality Assessment Form. 

No new data were created or analyzed in this study. Data sharing is not applicable to this article.

Facts: National School Lunch Program

School food environments and practices affect dietary behaviors of US public school children

National School Lunch Program: Total Participation

United States Department of Agriculture. National School Lunch Program and School Breakfast Program. In Nutrition Standards for All Foods Sold in School as Required by the Healthy

Local School Wellness Policy Implementation Under the Healthy, Hunger-Free Kids Act of 2010; Final Rule

Younger elementary school students waste more school lunch foods than older elementary school students

Examining vegetable plate waste in elementary schools by diversity and grade. Health Behav

The impact of school socioeconomic status on student lunch consumption after implementation of the Texas Public School Nutrition Policy

School lunch waste among middle school students: Nutrients consumed and costs

Food group preferences of elementary school children participating in the National School Lunch Program

A study of school lunch food choice and consumption among elementary school students

The importance of exposure for healthy eating in childhood: A review

Fourth-grade children's consumption of fruit and vegetable items available as part of school lunches is closely related to preferences

Modifying children's food preferences: The effects of exposure and reward on acceptance of an unfamiliar vegetable

Choice architecture as a means to change eating behaviour in self-service settings: A systematic review

Amount of time to eat lunch is associated with children's selection and consumption of school meal entrée, fruits, vegetables, and milk

Nutrition Education in US Schools

Elementary school children's recess schedule and dietary intake at lunch: A community-based participatory research partnership pilot study

Evaluation of the offer vs. serve option within self-serve, choice menu lunch program at the elementary school level

Peer reviewed: Factors related to fruit and vegetable consumption at lunch among elementary students: A scoping review

Impact of the new US Department of Agriculture school meal standards on food selection, consumption, and waste

New school meal regulations increase fruit consumption and do not increase total plate waste

Changes in foods selected and consumed after implementation of the new National School Lunch Program meal patterns in southeast Texas

Student participation, satisfaction, plate waste, and dietary intakes

Which healthy eating nudges work best? A meta-analysis of field experiments

Healthier choices in school cafeterias: A systematic review of cafeteria interventions

Behavioral economic approaches to influencing children's dietary decision making at school. Policy Insights Behav

Using nudges to promote healthy food choices in the school dining room: A systematic review of previous investigations

A systematic review of school meal nudge interventions to improve youth food behaviors

The impact the smarter lunchroom movement strategies have on school children's healthy food selection and consumption: A systematic review

Behavioral economics in the school lunchroom: Can it affect food supplier decisions? A systematic review

Reducing food waste and increasing fruit and vegetable consumption in schools. Health Behav

Salad bars in schools and fruit and vegetable selection and consumption: A review of recent research

Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement

Universal school meals and associations with student participation, attendance, academic performance, diet quality, food security, and body mass index: A systematic review

The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses

Newcastle-Ottawa Quality Assessment Scale Cohort Studies

Assessing the quality of studies in meta-analyses: Advantages and limitations of the Newcastle Ottawa Scale

Comparing reviewers' to authors' assessments

Differential improvements in student fruit and vegetable selection and consumption in response to the new National School Lunch Program regulations: A pilot study

The role of variety in increasing the consumption of fruits and vegetables among children

Increasing consumption of fruits and vegetables in the school cafeteria: The influence of active choice

Consumption of fruits and vegetables in middle school students following the implementation of a school district wellness policy

A randomized school trial of environmental strategies to encourage fruit and vegetable consumption among children

Fruit-promoting smarter lunchrooms interventions: Results from a cluster RCT

The cookshop program: Outcome evaluation of a nutrition education program linking lunchroom food experiences with classroom cooking experiences

School lunch environmental factors impacting fruit and vegetable consumption

Impact of a multicomponent, school-based nutrition intervention on students' lunchtime fruit and vegetable availability and intake: A pilot study evaluating the Shaping Healthy Choices Program

Salad bars increased selection and decreased consumption of fruits and vegetables 1 month after installation in Title I elementary schools: A plate waste study

Lunch salad bars in New Orleans' middle and high schools: Student intake of fruit and vegetables

Salad bars and fruit and vegetable consumption in elementary schools: A plate waste study

Long-term impact of a chef on school lunch consumption: Findings from a 2-year pilot study in Boston middle schools

Effects of choice architecture and chef-enhanced meals on the selection and consumption of healthier school foods: A randomized clinical trial

Impact of the Updated USDA school meal standards, chef-enhanced meals, and the removal of flavored milk on school meal selection and consumption

Eat your veggies: A chef-prepared, family style school lunch increases vegetable liking and consumption in elementary school students

Chefs move to schools. A pilot examination of how chef-created dishes can increase school lunch participation and fruit and vegetable intake

The addition of spices and herbs to vegetables in the National School Lunch Program increased vegetable intake at an urban, economically-underserved, and predominantly African-American high school

Using herbs and spices to increase vegetable intake among rural adolescents

Implementation of a multi-component school lunch environmental change intervention to improve child fruit and vegetable intake: A mixed-methods study

Impact of fruit smoothies on adolescent fruit consumption at school. Health Educ

Offering Self-serve sliced apples in bulk increases students' consumption and decreases apple waste in elementary school cafeterias

Modification of the form in which fresh fruit is served as a possible means of increasing the consumption of fruit offered to elementary and middle school students

Pre-sliced fruit in school cafeterias: Children's selection and intake

Promoting consumption of fruit in elementary school cafeterias. The effects of slicing apples and oranges

The impact of multiple strategies to encourage fruit and vegetable consumption during school lunch

Peer reviewed: Effects of a behavioral economics intervention on food choice and food consumption in middle-school and high-school cafeterias

A pilot intervention targeting dietary intake in school cafeterias

A randomized controlled trial of the Food Dudes program: Tangible rewards are more effective than social rewards for increasing short-and long-term fruit and vegetable consumption

Project FIT: A school, community and social marketing intervention improves healthy eating among low-income elementary school children

The Effect of Nutrition Education on Third Graders' School Lunch Consumption in a School Offering Food Pyramid Choice Menus

Nutrition education in the Little Rock school lunch program

Increasing the fruit and vegetable consumption of fourth-graders: Results from the high 5 project

Farm to elementary school programming increases access to fruits and vegetables and increases their consumption among those with low intake

Factors affecting fruit and vegetable school lunch waste in Wisconsin elementary schools participating in Farm to School programmes

Kids Choice" school lunch program increases children's fruit and vegetable acceptance

The FIT Game: Preliminary evaluation of a gamification approach to increasing fruit and vegetable consumption in school

Incentivizing children's fruit and vegetable consumption: Results of a United States pilot study of the Food Dudes Program

Promoting school lunch fruit and vegetable intake through role modeling: A pilot study

Longitudinal behavioral effects of a school-based fruit and vegetable promotion program

Decaying behavioral effects in a randomized, multi-year fruit and vegetable intake intervention

Promoting healthy food consumption among young children: Evaluation of a multi-component nutrition education program

Small prizes improve food selection in a school cafeteria without increasing waste

Impact of schoolchildren's involvement in the design process on the effectiveness of healthy food promotion materials

Location of school lunch salad bars and fruit and vegetable consumption in middle schools: A cross-sectional plate waste study

Smarter lunchrooms can address new school lunchroom guidelines and childhood obesity

Do environmental interventions impact elementary school students' lunchtime milk selection?

Healthy convenience: Nudging students toward healthier choices in the lunchroom

School transformation after redesign of 3 cafeterias (STARCafé). Health Behav

Factors influencing fluid milk waste in a Breakfast in the Classroom School Breakfast Program

The influence of a verbal prompt on school lunch fruit consumption: A pilot study

A plant to plate pilot: A cold-climate high school garden increased vegetable selection but also waste

Photographs in lunch tray compartments and vegetable consumption among children in elementary school cafeterias

Kindergarteners' entrée intake increases when served a larger entrée portion in school lunch: A quasi-experiment

Investigating the relationship between food pairings and plate waste from elementary school lunches; Food and Nutrition Sciences

Increasing portion sizes of fruits and vegetables in an elementary school lunch program can increase fruit and vegetable consumption

Serving vegetables first: A strategy to increase vegetable consumption in elementary school cafeterias

Serving first in isolation increases vegetable intake among elementary schoolchildren

Just dessert: Serving fruit as a separate "dessert" course increases vegetable consumption in a school lunch

Impact of a pilot school-based nutrition intervention on fruit and vegetable waste at school lunches

Outcomes from a school-based nutrition education program alternating special resource teachers and classroom teachers

Outcomes from a school-based nutrition education program using resource teachers and cross-disciplinary models

Healthy whole-grain choices for children and parents: A multi-component school-based pilot intervention

A nutrition education program at three grade levels

Healthy planet, healthy youth: A food systems education and promotion intervention to improve adolescent diet quality and reduce food waste

SC farm-to-school programs encourages children to consume vegetables

Children's consumption of fruits and vegetables: Do school environment and policies affect choices at school and away from school?

The effects of a nutrition education intervention on third-and fifth grade students' fruit and vegetable knowledge, preference and consumption

N" cool kids: The effects of character modeling and goal setting on children's physical activity and fruit and vegetable consumption

Effect of classroom intervention on student food selection and plate waste: Evidence from a randomized control trial

The relationship between the length of the lunch period and nutrient consumption in the elementary school lunch setting

Role of the elementary school cafeteria environment in fruit, vegetable, and whole-grain consumption by 6-to 8-year-old students

The relationship of meal and recess schedules to plate waste in elementary schools

Factors associated with school lunch consumption: Reverse recess and school "brunch

Food waste is reduced when elementary-school children have recess before lunch

Impacts of scheduling recess before lunch in elementary schools: A case study approach of plate waste and perceived behaviors

School lunch timing and children's physical activity during recess: An exploratory study

Modifying the recess before lunch program

recess and nutrition: Responding to time incentives in the cafeteria

Eat lunch first or play first? Inconsistent associations with fruit and vegetable consumption in elementary school

Impact of the National School Lunch Program on fruit and vegetable selection in northeastern elementary schoolchildren

Effect of a la carte and snack bar foods at school on children's lunchtime intake of fruits and vegetables

Fruits, vegetables, milk, and sweetened beverages consumption and access to a la carte/snack bar meals at school

Food type, food preparation, and competitive food purchases impact school lunch plate waste by sixth-grade students

Improvements in middle school student dietary intake after implementation of the Texas public school nutrition policy

Exploring changes in middle-school student lunch consumption after local school food service policy modifications

Chocolate milk consequences: A pilot study evaluating the consequences of banning chocolate milk in school cafeterias

Student acceptance of plain milk increases significantly 2 years after flavored milk is removed from school cafeterias: An observational study

Impact of breakfast in the classroom on participation and food waste

School lunch consumption among 3 food service providers in New Orleans

A mixed-methods evaluation of a school wellness initiative: An examination of longer lunch periods and more physical activity opportunities

Healthier standards for school meals and snacks: Impact on school food revenues and lunch participation rates

Schools and obesity prevention: Creating school environments and policies to promote healthy eating and physical activity

A systematic review and meta-analysis of school-based interventions with health education to reduce body mass index in adolescents aged 10 to 19 years

The influence of parental practices on child promotive and preventive food consumption behaviors: A systematic review and meta-analysis

The authors would like to thank the Healthy Eating Research staff at Duke University for their support.

The authors declare no conflict of interest.