CC BY-NC-ND
0ARTICLE IN PRESS
ELSEVIER
YPMED-03889; No. of pages: 7; 4C:
Preventive Medicine xxx (2014) xxx-xxx
Nutrient content of school meals before and after implementation of nutrition recommendations in five school districts across two U.S. counties
Patricia L. Cummings a,b'*, Sarah B. Welchc, Maryann Masonc,d, Lindsey Burbage b, Soyang Kwonc, Tony Kuo b,e
a Department of Epidemiology, Jonathan and Karin Fielding School of Public Health, University of California Los Angeles, 71-254 CHS, Los Angeles, CA 90024-1772, USA b Division of Chronic Disease and Injury Prevention, Los Angeles County Department of Public Health, 3530 Wilshire Blvd., 8th Floor, Los Angeles, CA 90010, USA
c Mary Ann and J. Milburn Smith Child Health Research Program, Ann and Robert H. Lurie Children's Hospital of Chicago Research Center, Consortium to Lower Obesity in Chicago Children, 225 E. Chicago Avenue, Box 205, Chicago, IL 60611, USA
d Department of Pediatrics and Preventive Medicine, Northwestern University Feinberg School of Medicine, 680 N. Lake Shore Drive, Suite 157, Chicago, IL 60611, USA e Department of Family Medicine, David Geffen School of Medicine, University of California Los Angeles, 10880 Wilshire Blvd., Suite 1800, Los Angeles, CA 90024-4142, USA
ARTICLE INFO
ABSTRACT
Available online xxxx
Keywords: Obesity Child nutrition Nutrition policy Nutrition assessment
Objective. To compare changes in nutrient levels of school meals before and after implementation of nutrition interventions at five school districts in two, large U.S. counties. School menu changes were compared against national school meal recommendations.
Methods. A large urban school district in Los Angeles County (LAC), California and four school districts in suburban Cook County (SCC), Illinois implemented school meal nutrition interventions. Nutrition analyses were conducted for school breakfast and lunch before and after changes were made to the meal programs. Means, % change, and net calories (kilocalories or kcal) offered as a result of the nutrition interventions were calculated.
Results. School districts in both counties made district-wide changes in their school breakfast and lunch menus. Menu changes resulted in a net reduction of calories, sugar, and sodium content offered in the meals. Net fewer calories offered as a result of the nutrition interventions were estimated to be about 64,075 kcal per student per year for LAC and 22,887 kcal per student per year for SCC.
Conclusions. Nutrition interventions can have broad reach through changes in menu offerings to school-aged children and adolescents. However, further research is needed to examine how these changes affect student food selection and consumption.
© 2014 Elsevier Inc. All rights reserved.
Introduction
The prevalence of childhood obesity in the United States (U.S.)1 has doubled for children and tripled for adolescents in the past 30 years. This is approximately 17% (12.5 million) of all children and adolescents ages 2-19 who are now obese (NCHS, 2012; Ogden and Carroll, 2010). Combating childhood obesity has been challenging due in part to the many and complex factors that are involved — the food environment being the most important of these factors (Drewnowski, 2004). For children and adolescents, school nutrition programs are a major component
* Corresponding author at: Los Angeles County Department of Public Health, 3530 Wilshire Blvd., 8th Floor, Los Angeles, CA 90010, USA. Fax: +1 213 351 2713.
E-mail addresses: pcummings@ph.lacounty.gov (P.L. Cummings), swelch@luriechildrens.org (S.B. Welch), mmason@luriechildrens.org (M. Mason), lburbage@ph.lacounty.gov (L Burbage), skwon@luriechildrens.org (S. Kwon), tkuo@ph.lacounty.gov (T. Kuo).
1 U.S. = United States.
of the food environment. Recognizing the central role that school nutrition can play in protecting health, a number of recent federal initiatives have invested substantively in school-based nutrition interventions aimed at improving the quality of foods served in school breakfast and lunch programs (Briefell et al., 2009; Bunnell et al., 2012; USDA, 2010). Improving the nutritional quality of food through the establishment of nutrient limits and other healthy food procurement practices in schools has emerged as a viable strategy for assuring a balanced diet and reducing childhood obesity in the U.S. (Briefell et al., 2009; Robles et al., 2013). National agencies, such as the Institute of Medicine (IOM)2 and the Alliance for a Healthier Generation, are supportive and have recommended this strategy as a way to lower caloric content in school meals, while preserving or improving their nutritional value (Alliance for a Healthier Generation, 2011; IOM, 2009).
2 IOM = Institute of Medicine.
http://dx.doi.org/mi 016/j.ypmed.2014.03.004 0091-7435/© 2014 Elsevier Inc. All rights reserved.
ARTICLE IN PRESS
P.L. Cummings et al. / Preventive Medicine xxx (2014) xxx-xxx
Although studies of school-based nutrition interventions are abundant in the literature (Doak et al., 2006; Katz et al., 2008; Roseman et al., 2011), few have described the core elements of design or the process by which these approaches can be implemented successfully in practice. To date, there are limited comparisons of nutrient changes in school menus after the implementation of school meal standards consistent with the Institute of Medicine, Alliance for a Healthier Generation, or the U.S. Department of Agriculture (USDA)3, especially for communities with a high prevalence of child obesity.
In 2011, a large, urban school district in Los Angeles County (LAC)4, California incorporated 1OM recommendations in their menu planning of school meals for the school year (SY)5 2011-12. Four school districts in suburban Cook County (SCC)6, Illinois implemented similar changes in their school meal programs; these changes aligned with the Alliance for a Healthier Generation school meal recommendations. In both counties, the nutrition interventions were implemented in advance of the USDA Final Rule for the National School Breakfast and Lunch Programs (NSBP/NSLP)7 (USDA, 2012). Both counties were also awardees of the Centers for Disease Control and Prevention's (CDC's)8 Communities Putting Prevention to Work (CPPW)9 program during 2010-2012 (Bunnell et al., 2012). Because the reach and impact of these nutrition strategies are often not well characterized in the literature, we described key meal program changes by nutrient categories for the five school districts that modified their SY2011-12 menus to meet nutrition standards recommended by the 1OM and the Alliance.
Methods
Nutrition interventions, school year 2010-11 to 2011-12
In addition to following the 1OM and Alliance recommendations, LAC and SCC included other strategies as part of their nutrition interventions (Table 1). These nutrition interventions were developed and implemented using food-based menu planning and aligned closely with anticipated changes to the USDA nutrition standards for school meals (USDA, 2012). For this comparison, LAC and SCC were selected for the following reasons: 1) school districts in both counties have parallel missions and similar operational scope; 2) LAC is one of, and SCC is located within one of, the largest counties in the nation and both have the most diverse student populations in the U.S. (Table 2); 3) they implemented comparable district-wide nutrition interventions that utilized healthy food procurement strategies (Table 1); 4) they periodically evaluated their school meal programs using nutrient analysis to monitor food quality; and 5) they were awardees of the national CPPW program during 2010-2012.
Nutrient analysis in Los Angeles County
In order to ensure adherence with the USDA nutrition standards, nutrient analyses of meal program menus are routinely performed by participants of the NSBP and NSLP. Through a data-sharing agreement with the Los Angeles Unified School District (LAUSD)10 Food Services Branch (FSB)",the Los Angeles County Department of Public Health (DPH)12 gained access to the nutrient analysis data for the months of October 2010 and October 2011, corresponding to the pre- and post-menu changes that took place as part of the school-based nutrition interventions implemented in LAC. The nutritional analysis was performed using the OneSource Point-of-Service software (Horizon Software International, Duluth, Georgia). OneSource uses the USDA food nutrient database to analyze recipes of food items on the menu; the database is continually updated to align with the NSBP and NSLP requirements. LAC analyzed the
3 USDA = U.S. Department of Agriculture.
4 LAC = Los Angeles County.
5 SY= School Year.
6 SCC = suburban Cook County.
7 NSBP/NSLP = National School Breakfast Program/National School Lunch Program.
8 CDC = Centers for Disease Control and Prevention.
9 CPPW = Communities Putting Prevention to Work
10 LAUSD = Los Angeles Unified School District.
11 FSB = Food Services Branch.
12 DPH = Los Angeles County Department of Public Health.
following nutrients: total fat, saturated fat, trans-fat, food energy (kilocalories or "kcal"), sugar, carbohydrates, cholesterol, dietary fiber, protein, iron, calcium, sodium, and vitamins A and C. In this article, we present nutrient data only for those collected by both LAC and SCC — i.e., trans-fat, carbohydrates, cholesterol, iron, and calcium were not included in the comparison analysis.
Data for the month of October were used for both school years because they: 1) allowed for assessments at two time points spaced apart by a 12-month interval, and 2) accounted for a 4-6 week start-up window, during which time the new menu underwent selected adjustments. The 900 + schools ( grades kindergarten [K]-12) of the LAUSD were included in the analysis for LAC. Detailed methods for the analysis methods have been described elsewhere (Cummings et al., 2014). Briefly, the analysis examined mean levels, 95% confidence intervals (CIs), and changes in nutrient content for student meals served during SY 2010-11 (n = 931 schools) and SY2011-12(n = 947 schools). The difference in number of schools included reflects the availability of the data for the two different academic years.
In SY 2010-11, four different meal categories were offered by the FSB: elementary breakfast, elementary lunch, secondary breakfast, and secondary lunch. Elementary grades include K-5 and secondary grades include 6-12. FSB served the same breakfast offerings for elementary and secondary grades in SY 2011-12; thus, these categories were combined for this school year. Each meal in each category (e.g., elementary lunch, secondary lunch) was offered to students as an assortment of entrées, at least one side option, milk, and condiments. Using estimation methods published previously by Cummings et al. (2014), nutritional content of the entrées, milk, and condiments were averaged and all sides were added into the total. These daily estimates were averaged for the entire month. For secondary school meals, the three lunch entrée options were averaged and for elementary school meals the two lunch entrée options were averaged. All analytic calculations were performed using the SAS statistical software package, version 9.3 (SAS Institute, Cary, North Carolina, USA). The LAC protocol was reviewed and approved by the Los Angeles County Department of Public Health Institutional Review Board (IRB).13 Since nutrient analysis data contained no individual identifying information, they were considered "exempt" by the IRB.
Nutrient analysis in suburban Cook County
Four school districts (n = 42 schools, grades prekindergarten [PK]-8) were randomly selected from a sample of seven eligible school districts in SCC to participate in SCC's CPPW Model Communities' Program. To be eligible, districts had to include elementary schools; as a result, the four participating districts in the program were strictly elementary school districts with a grade range of PK through 8. Each school district in SCC was required to post-menus and nutritional content online or make the information available to the public upon request. Menus for each of the four participating districts for the time periods May-June 2011 and March-May 2012 were collected and verified for adherence through observational audits during mealtime, randomly sampling approximately 25% of the schools, yielding 10 schools from the four districts. Utilizing similar nutritional analysis software as LAC, the main dish entrée, any side dishes listed on the menu, and the lowest calorie milk option for school meal nutrients were estimated as part of the daily totals. In cases where a range of side dishes were offered, only one of each was used in the calculation (e.g., for schools where students may choose up to 2 fruits or vegetables and up to 2 bread options, only 1 piece of fruit and 1 piece of bread was included in the calculation). This is based on the assumption that most students, on average, will take one of each side offered. Daily nutrient averages for each week were estimated by summing the daily total for each school and dividing by the total number of school days with menu data for that specific week. These weekly estimates were averaged for the three weeks. Unlike LAC, the selected school districts in SCC are small and preferred not to be identified by name. Thus, in the analysis they are labeled as District A, B, C, and D. The SCC protocol was reviewed and approved by the Ann and Robert H. Lurie Children's Hospital of Chicago Research Center Institutional Review Board.
Comparison
All LAUSD schools in LAC and all schools in the four selected school districts in SCC were included in the comparison described for the school years (SY) 2010-11 to 2011-2012. To compare the changes in nutrient levels after
13 IRB = Institutional Review Board.
ARTICLE IN PRESS
P.L. Cummings et al. / Preventive Medicine xxx (2014) xxx-xxx
Table 1
Communities Putting Prevention to Work nutrition interventions and food procurement strategies implemented in Los Angeles County, California and suburban Cook County, Illinois, 2010-2012.
Los Angeles County, California
Implement limits for food energy (kilocalories), sugar, sodium, and other nutrients for school breakfast and lunch programs district-wide, as recommended by the Institute of Medicine.a
Adopt an institutional nutrition policy that aligns menus with Institute of Medicine recommendations.
Assist with cafeteria improvements through healthy eating messaging and signage in school cafeterias (e.g., "I'm IN" campaign).
Work with executive chef and food vendors on new recipe development.
Conduct student and parent taste tests before rolling out the new menu (e.g., performed 30,000+ taste tests with students).
Implement a new food procurement model for the school district (e.g., converting to food-based menu planning, using categorical food partners, integrating marketing as a focus area in vendor contracts, purchasing from local sources — within a 200-mile radius).
Increase purchase and promotion of fresh fruits and vegetables (e.g., utilize approaches in behavioral economics to help increase selection).
Suburban Cook County, Illinois
Initiate school breakfast programs (if not already present) and promote these programs to parents and students.
Alter school breakfast and lunch programs district-wide to align with the Alliance for a Healthier Generation Healthy School Program nutritional standards.b
Assist with cafeteria improvements through healthy eating messaging and signage in school cafeterias.
Conduct healthy meal planning and budgeting workshops with student families.
Integrate fruits and vegetables into the school food environment through summer snack programs and school gardens.
a Institute of Medicine, 2009 recommendations for all students (tier 1 foods): <35% of total kilocalories from fat; < 10% total kilocalories from saturated fat; 0 g trans-fat; <35% of kilocalories from total sugars; <200 mg sodium per portion; water, low-fat or fat-free milk; and 100% fruit juice.
b Alliance for a Healthier Generation Healthy School Program nutritional standards: 1%, 1/2% or fat-free milk; use of whole grains in bread, pasta and pizza crust; non-fried foods; no-added-sugar fruits and vegetables; low-fat lunch entrée with no more than 12 g of total fat, 4.5 g saturated fat, 0 g trans-fat, and 480 mg sodium; use of unsaturated zero trans-fat oils; protein products with less than 10 g fat, 4.5 g or less saturated fat, and less than 95 mg cholesterol per serving and per 100 g; non-fried fish; and salad dressings with no more than 80 kilocalories per serving.
implementation of the nutrition interventions in both counties, we used the October 2010 school breakfast and lunch menus for elementary and secondary schools in LAUSD and compared them to the October 2011 menus. For SCC, we
used the May-June 2011 (three consecutive weeks) school breakfast and lunch menus for elementary schools and compared them to the March-May 2012 (three consecutive weeks) menus. These comparison time points were chosen based on the timeline of intervention implementation in each county, accounting for lag time between the two locales, but preserving the pre-and post-intervention interval at approximately 12 months apart. The post intervention results were then examined to see if they aligned with the IOM (for LAUSD) and Alliance for a Healthier Generation (for SCC) school meal recommendations.
Both counties had data for the following nutrients: food energy (kcal), protein (grams "g"), fiber (g), total fat (g), saturated fat (g), sugar (g), and sodium (milligrams "mg"). Means, 95% CIs, and percent change of nutrient levels pre- and post-intervention were compared for all LAUSD schools and all schools in the four districts in SCC. T-tests were performed to determine if nutrient changes were significant; where appropriate, log transformations were employed. Participation frequency (i.e., the number of students participating in school breakfast and lunch), average change in kilocalories per meal for breakfast and lunch, and the number of serving days per year were calculated and used to estimate net calories (kcal) offered annually for full-time (5 days per week) meal program participants (per student per year).
Results
Nutrition interventions implemented by LAUSD, which were based on IOM recommendations for healthy school meals (IOM, 2009), resulted in significant reductions in mean caloric and mean sugar content of breakfast and lunch school meals (Table 3). Similarly, for most meal categories, mean sodium content dropped. The most dramatic reductions were observed in the breakfast category for mean sugar, mean total fat, and mean sodium content. Although protein increased in the lunch meal category for elementary schools, the nutrient decreased in all other meal categories. Dietary fiber also decreased in all meal categories.
In SCC, where interventions were aligned with recommendations from the Alliance for a Healthier Generation, changes in nutrient content of meals were inconsistent (Tables 4A and 4B). Districts A and D, for instance, were able to significantly reduce mean sugar content in their lunch meals, whereas District C's mean sugar content for the same meal category slightly increased (Tables 4A and 4B). Aside from a slight increase in protein, District D did not improve on most of the nutrients for breakfast and District A's breakfast data were incomplete. District B baseline data for fiber, sugar, and sodium breakfast nutrients
Table 2
Characteristics of Los Angeles Unified School District (LAUSD) and the selected suburban Cook County school districts (SCC SD), 2010-2012.
Characteristics LAUSDa n (%) SCCSDAbn(%) SCCSDBbn(%) SCCSDCbn(%) SCCSDDbn(%)
Total 662,140 5040 5253 13,333 2431
Grades
Kindergarten-5 312,434 (47.2) 3325 (66.0)c 3626 (69.0)c 9200 (69.0)c 1696 (69.8)c
6-12 349,706 (52.8) 1715 (34.0)d 1627 (31.0)d 4133 (31.0)d 735 (30.2)d
Male 339,142 (51.2) 2535 (50.3) 2730 (52.0) 6768 (50.8) 1212 (49.9)
Female 322,998 (48.8) 2505 (49.7) 2523 (48.0) 6565 (49.2) 1219 (50.1)
Race/ethnicity
Black 63,714 (9.6) 43 (0.8) 1977(37.6) 241(1.8) 1880 (77.3)
Hispanic 478,943 (72.3) 257 (5.1) 3094 (58.9) 12,537 (94.0) 485 (20.0)
White 66,833 (10.1) 4126 (81.9) 101 (1.9) 432 (3.2) 9 (0.4)
Asian/Pacific Islandere 43,251 (6.6) 352 (7.0) 26 (0.5) 17(0.1) 34 (1.4)
Other1 9399 (1.4) 262 (5.2) 55 (1.1) 106 (0.8) 23 (0.9)
Obesity
All grades 174,143 (26.3)g 77 (9.1)lhi 341 (28.4)h 661 (32.3)h 189 (25.7)h
a Source: California Department ofEducation, school-year 2011-12. b Source: Illinois State Department ofEducation, school-year 2011-12. c Includes prekindergarten students. d Includes only grades 6-8. e Includes Filipino.
f Includes American Indian/Alaskan Native, two or more races (not Hispanic), and not reported. g Source: California Physical Fitness Testing Program (FITTNESGRAM©), grades 5, 7, and 9.
h Source: Cook County Department of Public Health. 2013. Unpublished, sample includes students from kindergarten and 6th grade. i Includes only kindergarten students.
ARTICLE IN PRESS
4 P.L. Cummings et al. / Preventive Medicine xxx (2014) xxx-xxx
(A) Breakfast menu: mean nutrient content for elementary and secondary schools in Los Angeles Unified School District, October 2010 and October 2011, grades kindergarten-12.
(B) Lunch menu: mean nutrient content for elementary and secondary schools in Los Angeles Unified School District, October 2010 and October 2011, grades kindergarten-12.
Key nutrients October 2010 mean (95% CI)a October 2011 mean (95% CI)a % change October 2010 mean (95% CI)a October 2011 mean (95%CI)a % change
Elementary (grades k-5) Secondary (grades 6-12)
(A) Food energy (kcal) Protein (g) Dietary fiber (g) Total fat (g) Saturated fat (g) Sugar(g) Sodium (mg) 603.2(564.7,641.7) 20.8(19.0, 22.6) 5.5 (4.8 6.2) 13.6(11.3,15.8)d 3.7 (3.3,4.1) 35.7(29.5,41.9) 746.8(677.4,816.1) 388.3 (354.7,421.8) 17.6 (16.6,18.6) 3.2 (2.5,3.9) 7.9 (6.0,9.7)d 2.3 (1.9,2.6) 14.3 (9.6,18.9) 494.3 (424.0, 564.6) — 35.6bc -15.5c -40.9c —42.0bc — 39.5c — 60.1bc — 34.4bc 834.6 (781.1,888.2) 27.7 (26.3,29.1) 7.7 (7.4, 8.0) 19.5 (16.8,22.1)d 5.6 (5.2, 5.9) 39.0 (32.4,45.6) 1064.3 (1003.5,1125.1) 388.3 (354.7,421.8) 17.6 (16.6,18.6) 3.2 (2.5, 3.9) 7.9 (6.0, 9.7)d 2.3 (1.9, 2.6) 14.3 (9.6,18.9) 494.3 (424.0, 564.6) - 53.5b,c - 36.5c -58.1c - 59.6b,c - 59.6b,c - 67.6b,c - 53.5b,c
(B) Food energy (kcal) Protein (g) Dietary fiber (g) Total fat (g) Saturated fat (g) Sugar(g) Sodium (mg) 714.9 (688.9, 740.8) 30.3 (28.8, 31.8)d 9.6 (9.0,10.3) 21.4(20.2, 22.7)d 5.2 (4.9, 5.5)d 16.7(11.1,22.3) 1000.6 (941.0,1072.3) 681.1 (627.1,735.1) 37.8 (31.1,44.5)d 8.3 (7.5,9.0) 24.3 (18.2,30.3)d 5.6 (4.1, 7.0)d 13.8 (10.6,17.0) 1327.0 (1093.3,1560.7) - 4.7b 24.6bc - 14.3c 13.1 7.1b - 29.9b 31.8c 790.7 (750.9, 830.5) 33.4(31.9,34.9) 10.9 (10.2,11.7) 22.9 (21.4,24.4) 5.4 (5.1, 5.8) 17.9 (12.5,23.3) 1139.0 (1054.1,1223.9) 634.8 (598.4,671.3) 30.7 (28.4,33.1) 7.6 (6.6, 8.7) 21.1 (18.9,23.3) 4.7 (4.3, 5.2) 12.9 (10.8,14.9) 1089.8 (962.1,1217.6) — 19.7b,c - 8.0c — 30.2c — 8.0b — 13.4b,c — 38.3b,c — 4.3b
Other abbreviations: kcal = kilocalories; g = grams. a CI = confidence interval. b Change in desired direction. c Significant at p value < .05. d log transformation performed.
were missing, thus percent changes were not calculated for these nutrients. For the school lunch programs, Districts A, C and D were able to achieve more substantive improvements (Tables 4A and 4B). District A reduced mean calories by 15.7%, mean sugar by 32.4%, and mean sodium by 21.6% for its lunches. District D was able to achieve similar results, while District B reduced mean calories by only 2.9% and did not possess baseline data to assess for changes in fiber, sugar, or sodium nutrient content. Although District C increased overall calories, fat, saturated fat, and sugar, it was able to reduce sodium and increase dietary fiber and protein in their lunch offerings.
Collectively, the estimated number of children and adolescents reached by the school-based nutrition interventions in both counties was estimated to be 688,197 students for the SY 2011-12 (Table 2). Net fewer calories (kcal) offered as a result of the nutrition interventions was estimated to be about 64,075 kcal per student per year for LAC and 22,887 kcal per student per year for SCC.
Discussion
Overall, reductions in calories, sugar and sodium content of student meals offered by LAC and SCC schools were achieved in the five school districts that modified their SY 2011-12 menus. These results, however, reflect only average nutrient changes by meal categories; they do not correspond to other salient factors that may also influence student nutrition — e.g., food presentation and appeal; taste of the new items; perceptions of freshness and food quality; density, composition or quality of the individual offerings including the number and type (variety) of entrées or sides prepared or available to choose from; and student food selection and actual consumption (or waste). In LAC and SCC, for example, the entrée or side variety changed from SY 2010-11 to SY 2011-12, reflecting the school districts' emphasis on not only meeting nutrient limits, but also addressing the context leading to food selection and consumption — i.e., using a food-based menu planning approach. In
Table 4A
Breakfast menu: mean nutrient content for school Districts B, C, and D in suburban Cook County, Illinois, May-June 2011 and March-May 2012, grades prekindergarten-8.
Key nutrients 2010-11 mean 2011-12 mean % 2010-11 mean 2011-12 mean % 2010-11 mean 2011-12 mean %
(95% CI)a (95% CI)a change (95% CI)a (95% CI)a change (95% CI)a (95% CI)a change
District B District C District D
Food energy 561.4 (525.6, 546.4(515.6,577.1) —2.7b 531.2 (468.0,594.4) 534.8 (476.3,593.2) 0.7 348.0 (303.7, 392.3) 406.6 (305.4, 507.7) 16.8
(kcal) 597.2)
Protein (g) 15.0(13.8,16.1)c 14.1 (13.1,15.0)c —6.0 14.1 (12.7,15.4) 15.3 (13.2,17.5) 8.5b 11.2(10.2,12.2)e 12.2(11.0,13.4)e 8.9b
Fiber (g) -d 3.4 (3.0, 3.9) -d 2.8 (2.0, 3.6)e 4.3 (1.9, 6.7)e 53.6b 1.9 (1.5,2.3) 1.3 (0.7,2.0) —31.6
Total fat (g) 14.3 (12.4,16.2) 13.2 (11.2,15.2) —7.7b 12.3 (9.3,15.2) 12.9 (9.5,16.3) 4.9 6.9 (5.7,8.1) 9.6 (5.4,13.7) 39.1
Saturated fat (g) 4.3 (3.8,4.8) 3.9 (3.4,4.5) —9.3b 3.5 (2.9,4.2) 4.0 (2.9, 5.0) 14.3 3.3 (2.7, 3.8) 3.7 (2.2, 5.2) 12.1
Sugar(g) -d 42.2 (38.7,45.7) -d 44.3 (37.3, 51.4) 41.8 (36.0,47.7) — 5.6b 36.2 (32.0,40.4) 41.8 (33.6, 50.0) 15.5
Sodium (mg) -d 515.5 (460.4,570.6) -d 543.7(455.1,632.3) 571.5 (460.1,682.9) 5.1 315.5 (285.0, 346.0) 339.1 (266.4,411.8) 7.5
Other abbreviations: kcal = kilocalories; g = grams.
Note: District A is not shown in the table because it did not offer school breakfast. a CI = confidence interval. b Change in the desired direction. c Sine transformation performed. d "-" indicates missing data. e Cosine transformation performed.
ARTICLE IN PRESS
P.L. Cummings et al / Preventive Medicine xxx (2014) xxx-xxx
^ on (N ("O
- tN ^ tN
, 00 cn. 00
m m co ^
<*! en
tN Cï
m Cn W> 00
en $ ^ ™ 'n
o t— m t r* '—• oo m m
^ oo ^ rn ^ "Î 1
tN tN tO t^
rn m ^ m O O CN
CN CN t^ o, m °° CO CO r-
m cd CN CN m co
M "O .
& (U ni g O
'O T3 '
LAC, the 2010-11 lunch menu had items such as beef chalupa, pepper-oni pizza, and Italian calzone with turkey pepperoni; whereas, the new 2011-12 lunch menu included black eyed pea salad, vegetable curry, Ancho chili chicken with yakisoba, and quinoa and veggie salads. Likewise, in SCC District A, the 2010 -11 lunch menu had items such as nachos, chicken nuggets, hot dogs, and cheese pizza; whereas, the new 2011-12 lunch menu offered turkey burgers, black bean burgers, tuna salad subs, and stuffed salmon. Clearly, taken together, more can be learned from the experiences in LAC and SCC. Further research using methods such as dietary pattern scores is needed and could provide additional insights on the impacts of these food-based offerings or strategies on student eating behaviors.
A multicomponent approach
The LAUSD experience in LAC suggests that a multicomponent approach was beneficial for introducing, integrating, and supporting healthy food modifications to the SY 2011-12 menus. The "I'm IN" public education campaign, for example, augmented the student and parent taste testing by LAUSD by helping to prepare students for the new menu items that were introduced (Table 1). Age-appropriate portion sizes for some of the meal categories also enabled reductions in key nutrients without significant modifications to food composition or taste. However, this latter action did contribute to unintended effects — e.g., the lowering of desirable nutrients such as protein and fiber. In addition, these complementary strategies do not necessarily improve nutrition for everyone. For instance, for those children whose energy intake is appropriate, simply reducing portion size does not alter the food selection or the composition of their diet, which may still be poor. Children can also compensate for lost energy intake by consuming undesirable foods from other sources.
School districts in the U.S. that are contemplating similar menu changes to their student meal program may find food-based menu planning more logistically feasible and in line with the USDA Final Rule (USDA, 2012). Protein, fiber, and other healthful nutrients are vital for ensuring proper nutrient intake among students and should be taken into account when making menu changes. Another factor to consider is children and adolescents who are not receiving adequate nutrient intake (i.e., poor diet composition with excess energy intake). This can occur even among children who are obese, not just for those who are underweight. Moderately active children, ages 4 - 8, for example, need 1400-1600 kcal per day; those, ages 9-13, need 1800-2200 kcal per day. Sedentary children and adolescents require the lower end of this range (USDA, 2010). In LAC and SCC, the average school meal caloric ranges were between 380 and 830 kcal per meal.
Recognizing the influential role that taste can play in food selection, the LAUSD (in LAC) conducted 30,000+ taste tests prior to finalizing the menu for SY 2011 -12 (Table 1). SCC took similar actions to improve the appeal of their new menu items to increase student receptivity (Mason et al., 2012). SCC school districts, for example, made changes to the formula of the school meals while concurrently providing public education to parents and students about the benefits of healthy eating (Table 1). The collective experiences of the four SCC districts point to the importance of having staff role models and stakeholder buy-ins during the menu planning stage. In one district, union regulations stalled the implementation of breakfast in the classroom.
It should be noted that there were key differences between the two counties. The sheer size of LAUSD translated to greater purchasing power and easier negotiations for better pricing from food suppliers, which in turn probably contributed to the district's capacity to offer a wider range of healthy food options (Robles et al., 2013). In SCC, each school district conceptualized and implemented different interventions based on their unique needs, assets and operating capacity. Differences in these factors likely contributed to the differences seen in the nutrient changes in the different school districts during SY 2010 -11 to 2011-12.
ARTICLE IN PRESS
P.L. Cummings et al. / Preventive Medicine xxx (2014) xxx-xxx
Overlapping strategies in all five districts made this evaluation salient and interesting, as they point to alternative lessons learned about effective ways to improve school nutrition. SCC schools customized their food procurement strategies based on district and school-level capacity, leading to more targeted changes that are specific to individual school cafeterias; whereas LAUSD's interventions were standardized and incremental but had broad reach due to the district's sheer size and centralized infrastructure.
Limitations
The present analysis is subject to a number of limitations. First, using nutrient analysis as an approach for program evaluation provides an incomplete picture of student nutrition in the school setting. On the other hand, examining nutrient changes by meal categories using standard nutrient-estimation protocols represents a practical approach for comparing institutional improvements in food offerings across different schools. Second, the nutrient analysis records from LAUSD and from the four school districts in SCC were compiled using nutritional software that analyzed information from menu recipes. While this is generally considered an acceptable alternative to laboratory nutrient analysis (gold standard), user errors can occur (Drake, 1992). Third, the nutrient analysis in this evaluation provides only a cross-sectional snapshot of the mean change per meal for each nutrient; it does not provide longitudinal confirmation of intervention effectiveness nor sustainability, since only one month during each school year was analyzed. Changes in certain nutrients, such as total fat, for example, may not equate to actual improvements in food offerings.
Although the strength of the analysis is its pre- and post-intervention design, factors such as student food selection pattern, taste, meal appeal, and receptivity to the menu changes all can attenuate the magnitude of the observed effects. For instance, in a prior analysis of LAUSD data, Cummings et al. (2014) demonstrated that changes to mean sodium content were not as substantial once student food selection patterns were accounted for. Other methods, such as plate waste studies represent potentially better measures of student food selection and consumption. In 2011, a plate waste study was conducted in LAC using a random sample of four middle schools; preliminary results suggest a relatively large amount of fruit and vegetable wastage after introduction of these plant-based food options.
Finally, the lack of homogeneity in the school-based nutrition interventions likely led to bias in the results. Given the diversity of the intervention components (from food service staff training to incorporation of new contract language), it is difficult to disentangle the contributions of each component. For example, LAC used a categorical food partner model to work with vendors on developing new recipes that included more fresh fruits and vegetables on the menu, while also utilizing behavioral economics approaches to promote fruit and vegetable selection (e.g., putting fruits in an attractive basket near check-out stands). These strategies likely worked synergistically to increase selection of these items by students.
Conclusions
Collectively, school-based nutrition interventions in LAC and SCC appeared to have contributed favorably to changes in the school cafeteria environment, including improvements to the overall nutrient base of school meals served. This suggests that federal as well as local initiatives in obesity prevention and in cardiovascular health promotion should continue to invest in these kinds of system and environmental changes aimed at creating healthier food environments for children and adolescents in the U.S.
Conflict of interest statement
The authors report no financial disclosures or conflicts of interest.
Acknowledgments
The authors would like to thank the Board of Education, the Office of the Superintendent, and the Food Services Branch in the Los Angeles Unified School District, and the Cook County Department of Public Health as well as the four participating school districts for their support and contributions to this project. The authors would also like to thank Janice H. Vick and Kathleen Whitten from ICF International for their careful review of this manuscript prior to submission. The project was supported in part by cooperative agreements from the Centers for Disease Control and Prevention (Communities Putting Prevention to Work #3U58DP002485-01S1, #1U58DP00263-01S1, and Sodium Reduction in Communities Program # 1U58DP003061-01). The findings and conclusions in the article are those of the authors and do not necessarily represent the views or the official position(s) of the Consortium to Lower Obesity in Chicago Children, the Los Angeles County Department of Public Health, the Cook County Department of Public Health, the Centers for Disease Control and Prevention, the Ann and Robert H. Lurie Children's Hospital of Chicago or any other organization mentioned in the text. In accordance with U.S. law, no Federal funds provided by CDC were permitted to be used by community grantees for lobbying or to influence, directly or indirectly, specific pieces of pending or proposed legislation at the federal, state, or local levels. As it relates to the CDC-sponsored supplement, staff training and reviews by scientific writers were provided as technical assistance to the authors, through a contract with ICF International (Contract No. 200-2007-22643-003). CDC staff has reviewed the project's evaluation design and data collection methodology and the article for scientific accuracy. All authors have read and approved the final version.
References
Alliance for a Healthier Generation, 2011. Alliance school meals criteria. (Available at) https://schools.healthiergeneration.org/_asset/gxljhk/11-2736_SMCriteria.pdf (Accessed on August 23, 2013). Briefell, R.R., Crepinsek, M.K., Cabili, C., Wilson, A., Gleason, P.M., 2009. School food environments and practices affect dietary behaviors of US public school children. J. Am. Diet. Assoc. 109 (2 Suppl.), S91-S107. Bunnell, R., O'Neil, D., Soler, R., et al., 2012. Communities Putting Prevention to Work Program Group. Fifty communities putting prevention to work: accelerating chronic disease prevention through policy, systems, and environmental change. J. Community Health 37, 1081-1090. http://dx.doi.org/10.1007/ s10900-012-9542-3 (epub). Cummings, P.L., Burbage, L., Wood, M., Butler, R., Kuo, T., 2014. Evaluating changes to sodium content in school meals at a large, urban school district in Los Angeles County, California. J. Public Health Manag. Pract. 20 (1 Suppl. 1), S43-S49. http://dx.doi.org/ 10.1097/PHH.0b013e31829f2e50. Doak C.M., Visscher, T.L.S., Renders, C.M., Seidell, J.C., 2006. The prevention of overweight and obesity in children and adolescents: a review of interventions and programmes. Obes. Rev. 7,111-130. Drake, M.A., 1992. Menu evaluation, nutrient intake of young children, and nutrition knowledge of menu planners in child care centers in Missouri. J. Nutr. Educ. 24, 145-148.
Drewnowski, A., 2004. Obesity and the food environment: dietary energy density and diet
costs. Am. J. Prev. Med. 27 (3 Suppl.), 154-162. Institute of Medicine (IOM) of the National Academies, 2009. School meals: building blocks for healthy children. (Available at) http://www.iom.edu/Reports/2009/School-Meals-Building-Blocks-for-Healthy-Children.aspx (Accessed on August 23, 2013). Katz, D.L., O'Connell, M., Njike, V.Y., Yeh, M.C., Nawaz, H., 2008. Strategies for the prevention and control of obesity in the school setting: systematic review and meta-analysis. Int. J. Obes. 32,1780-1789. Mason, M., Chavez, N., Kwon, S., Cedeno, A., 2012. Student school meal perception, participation, and consumption: implication for future school based nutritional interventions. Paper presented at: American Public Health Association. 140th Annual Meeting and Exposition. October 28-31, 2012; San Francisco, California.
National Center for Health Statistics (NCHS), 2012. Health, United States, 2011: With Special Features on Socioeconomic Status and Health. U.S. Department of Health and Human Services, Hyattsville, MD. Ogden, C., Carroll, M., 2010. Prevalence of Obesity Among Children and Adolescents: United States, Trends 1963-1965 through 2007-2008. National Center for Health Statistics (June 2010).
Robles, B., Wood, M., Kimmons, J., Kuo, T., 2013. Comparison of nutrition standards and other recommended procurement practices for improving institutional food offerings in Los Angeles County, 2010-2012. Adv. Nutr. 4 (2), 191-202. http://dx.doi.org/10.3945/an. 112.003285.
ARTICLE IN PRESS
P.L. Cummings et al. / Preventive Medicine xxx (2014) xxx-xxx 7
Roseman, M.G., Riddell, M.C., Haynes, J.N., 2011. A content analysis of kindergarten-12th grade school-based nutrition interventions: taking advantage of past learning. J. Nutr. Educ. Behav. 43 (1), 2-18. U.S. Department of Agriculture (USDA), 2010. Dietary Guidelines for Americans, 2010. U.S. Government Printing Office, Washington D.C.
U.S. Department of Agriculture (USDA), 2012. Food and Nutrition Service. 7 CFR Parts 210 220. Nutrition Standards in the National School Lunch and School Breakfast Programs; Final Rule. Fed. Regist. 77 (17), 4087-4167 (Retrieved at http://www.gpo.gov/fdsys/pkg/FR-2012-01-26/pdf/2012-1010.pdf. Last accessed: November 15, 2013).
