A Bin 1%. z... .3. .41“ A 3.1. . “9.44.1 {,3 Cl . :3. ‘- {Illi - 4; 11):: . . Hiram :g 2:. xi): f I? . ”figurvu . .nasr... “rig-3| ELEM...” . .iinhufixitahm? ‘ v (3‘ it... I an; S 1 ‘ .- '- fawn»... H‘ I". d‘e 5:... . u}. s 5 i . .. $97.5 l g 1, J3 sit-v 5...: u: is 1.31. 1. r!. 3‘1...“ 1» «En! .« .3. 3r THESIS 1:32 LIBRARY 3 . Michigan State I University This is to certify that the dissertation entitled EXAMINATION OF THE FACILITATORS, BARRIERS, AND RELATIONSHIPS AMONG SCHOOL NUTRITION POLICIES, SCHOOL NUTRITION ENVIRONMENTS AND PRACTICES, AND STUDENT DIETARY INTAKES IN LOW-INCOME MICHIGAN MIDDLE SCHOOLS presented by JENNIFER FAY MOSACK has been accepted towards fulfillment of the requirements for the Ph.D. degree in Human Nutrition KAlW > Major Professor’s §T§nature Date MSU is an Affirmative Action/Equal Opportunity Employer PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 5/08 K:IProj/Aoc&Pres/ClRC/Datoom.indd EXAMINATION OF THE FACILITATORS, BARRIERS, AND RELATIONSHIPS AMONG SCHOOL NUTRITION POLICIES, SCHOOL NUTRITION ENVIRONMENTS AND PRACTICES, AND STUDENT DIETARY INTAKES IN LOW-INCOME MICHIGAN MIDDLE SCHOOLS By Jennifer Fay Mosack A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Human Nutrition 2010 E) to c: nutri from Adi; “tilt make ABSTRACT EXAMINATION OF THE F ACILITATORS, BARRIERS, AND RELATIONSHIPS AMONG SCHOOL NUTRITION POLICIES, SCHOOL NUTRITION ENVIRONMENTS AND PRACTICES, AND STUDENT DIETARY INTAKES IN LOW-INCOME MICHIGAN MIDDLE SCHOOLS By Jennifer Fay Mosack Trends in adolescent obesity and current dietary intake are of concern due to their association with chronic diseases, disability, and reduced quality of life in adulthood. Schools have been identified as a setting in which health promotion efforts may help reverse these trends. This dissertation research aims to improve understanding of school nutrition to enhance the effectiveness of school health intervention efforts. Baseline data from 65 low-income Michigan middle schools participating in the School Nutrition Advances Kids (SNAK) research project were utilized. First, federally-mandated local wellness policies from 48 SNAK districts were examined. The primary determinant of wellness policy quality was the template used to create the policy. There was little agreement between written wellness policies and administrator or food service director (FSD) reported policies and practices. Next, this dissertation examined associations between the availability of competitive foods in schools and student dietary intake using data from 1544 students in 51 SNAK schools. Compared to schools with no competitive foods available: having both a la carte and vending in schools was associated with increased saturated fat intake; having a la carte only or vending only available was associated with an increased fruit intake; having only healthy beverages available in vending machines was associated with decreased energy, vegetable, fi'uit + vegetable intake; having mixed healthy and less healthy beverages available or mixed foods and beverages available in vending was associated with increased fat intake, and mixed beverages was associated with increased saturated fat intake. Availability of a la carte or vending individually revealed no significant associations. These results are likely due to differences in the nutrient content of foods available in these venues and/or limitations of the food frequency questionnaire used to assess dietary intake. Lastly, this dissertation identified barriers and accomplishments to promoting healthy eating and factors that facilitated change in schools using qualitative case studies of 8 SNAK schools. Administrators, F SDs, coordinated school health team members, and students at each school were interviewed. Barriers to promoting nutrition in these schools included budgetary constraints that led to low prioritization of health initiatives; the economic situation of the community that may lead to consumption of less healthy foods at home; quality of schools meals; widespread availability of unhealthy competitive foods; and perceptions that students would not eat healthy foods. Despite these challenges, many schools had made improvements to school meals and competitive foods and were increasing nutrition education efforts within and outside of the school setting. Support from school administrators, teamwork among staff members, and acknowledging student preferences helped to make positive changes in the food service program. Schools with a more health-promoting school culture (e.g., presence of a coordinated school health team, enforcement of nutrition policies, and a school health champion) made more changes to promote health and nutrition to students than other schools. These research results will inform fiiture intervention and policy efforts aimed at improving school nutrition environments and policies in order to improve adolescent dietary intake. Copyright by JENNIFER FAY MOSACK 2010 This work is dedicated to the children of the world, for they are the future and my inspiration. ACKNOWLEDGMENTS I am very proud of the work that I have put into this dissertation research; however, none of it would have been possible without support from a very long list of individuals who have assisted and influenced me over the years. First and foremost, I must thank the students, parents, and school staff members who participated in the School Nutrition Advances Kids (SNAK) project and allowed me and my colleagues into their schools. Additionally, I must also thank those that provided funding for this research and my education, including: the Robert Wood Johnson Foundation Healthy Eating Research program; USDA’s Supplemental Nutrition Assistance Program — Nutrition Education program by way of the Michigan Nutrition Network at Michigan State University Extension in partnership with the Michigan Fitness Foundation and supported in part by the Michigan Department of Human Services, under contract number ADMIN-10-99010; and the Michigan Department of Community Health who funded the School Nutrition Advances Kids (SNAK) project; Michigan State University (MSU) Department of Food Science and Human Nutrition, the John Harvey Kellogg Endowed Fellowship, and MSU Graduate School who helped to fund my education. I am grateful for the support of my Master's advisor Bill Saltarelli, who found me as an undergraduate student who didn't quite know what she wanted to be when she grew up, and saw my potential. If it were not for the opportunities and encouragement that Bill gave me, I likely would not have made it as far as I have today. I must also greatly acknowledge my dissertation advisor, Katherine Alaimo, who graciously allowed me to come into her lab, and work with her research projects, and gave me countless opportunities to develop my professional skills. Katherine took the time to work vi intensively with me to develop as both a researcher and a person. I must also thank the members of my guidance committee including Mike Hamm, Kim Chung, and Beth Olson, who were instrumental in helping me develop my dissertation from the proposal to the many changes that were necessary as my research progressed, and helped me to develop a quality end product. I also thank Ellen Velie, who I consider an honorary member of my dissertation committee, for the insight that she was able to provide me with regarding measuring dietary intake. A special thanks to project staff members Deanne Kelleher, Richard Miles, and Deb Bailey, friends and colleagues who always made time to discuss my latest thoughts about school nutrition and helped to shape my thinking about children's diets, and whose friendships and professional insights I value immensely. There are many other past and present members of the SNAK project planning team who must be acknowledged for their contributions to the planning and development of the SNAK project and guiding the project with their real-world knowledge of schools including: Elaine Belansky, Julie Marshall, Diane Golzynski, Shannon Carney Oleksyk, Nick Drzal, Ann Guyer, Dru Sczerba, Paul McConaughy, Connie Page, Hui "Cathy" Liu, Paul Baumgartner, Donna Hensey, Larry Merx, Whitney Vance, Robynn Corey, and Christian Hanna; the numerous research assistants that have assisted with all parts of the SNAK project including Caroline Martin, Leah Simpson, Meaghan Snowdin, Allison Krusky, Stephanie Gorte, Caitlin Fisher, Kati Garrison, Alexandra Strucel, Jessica Jenkins, and especially Ellen Mang, who worked very closely with me to code wellness policies and interview transcripts, and whose positive attitude and humor made this process a pleasure. vii Lastly, I must thank all of my family and friends, whose faith in me and support helped me to keep going when I wasn't sure if I had it in me! Thanks to my husband and best friend, Chris, who found it possible to love a doctoral student, and whose unwavering love and support will always be appreciated and reciprocated; to my much wiser sister, Jill, who has always been there for me throughout my life, and perhaps understands me more than anyone else; and to my parents, Jack and Nancy, who ingrained in me at an early age that I could do anything if I put my mind to it. Mom and Dad, you can finally tell everyone that your daughter is a doctor! viii LIST OF I LIST OF Fl CHAPTLR CHAPTER lite Trcr lnll \th St‘ii Sch lndi The Sch Lim ij CHAPTER ASSOC IN MIDDLE 3 Bar Mel Ana Rust CHAPTER I"0(le Ax MIDDLE g. Bacl Mcll TABLE OF CONTENTS LIST OF TABLES ............................................................................................................ xii LIST OF FIGURES ........................................................................................................ xiii CHAPTERS CHAPTER 1: INTRODUCTION ....................................................................................... l The importance of nutrition during childhood and adolescence .............................. 5 Trends in obesity and dietary intake ........................................................................ 7 Influences on adolescent dietary behaviors - an ecological model ........................ 9 Why study schools? Applying the ecological model to the school setting ........... 12 School policy ......................................................................................................... 15 School policies before child nutrition reauthorization ............................... 16 Support for wellness policy development .................................................. 18 School policies after child nutrition reauthorization .................................. 23 School nutrition environment ................................................................................ 26 School meals .............................................................................................. 26 Competitive Foods ..................................................................................... 30 Individual influences on adolescent dietary behaviors .......................................... 32 The role of administrators in nutrition promotion ................................................. 38 Schools as an intervention setting to improve adolescent dietary intake ............... 41 School nutrition education interventions .................................................. 41 School nutrition environment interventions .............................................. 42 Limitations of the school nutrition intervention literature ..................................... 47 Project justification ................................................................................................ 49 CHAPTER 2: THE QUALITY OF SCHOOL WELLNESS POLICIES AND ASSOCIATION WITH SCHOOL PRACTICES IN LOW-INCOME MICHIGAN MIDDLE SCHOOLS ........................................................................................................ 56 Background ............................................................................................................ 56 Methods .................................................................................................................. 58 School Nutrition Advances Kids project ................................................... 58 Study sample .............................................................................................. 59 Wellness policy evaluation ....................................................................... 59 School Environment and Policy Survey .................................................... 61 School characteristics ................................................................................. 62 Analysis ................................................................................................................. 63 Results .................................................................................................................... 65 CHAPTER 3: ASSOCIATION BETWEEN AVAILABILITY OF COMPETITIVE FOODS AND STUDENT DIETARY INTAKE IN LOW-INCOME MICHIGAN MIDDLE SCHOOLS ......................................................................................................... 81 Background ............................................................................................................ 81 . Methods .................................................................................................................. 84 ix School Nutrition Advances Kids (SNAK) project ..................................... 84 Study sample ............................................................................................. 84 Instruments and procedures ................................................................................... 85 Student dietary intake ................................................................................ 85 School nutrition environment .................................................................... 88 School characteristics ................................................................................. 89 Analysis .................................................................................................................. 90 Results .................................................................................................................... 92 Discussion .............................................................................................................. 99 CHAPTER 4: A QUALITATIVE EXPLORATION OF THE ACCOMPLISHMENTS AND CHALLENGES TO PROMOTING HEALTHY EATING IN LOW-INCOME MIDDLE SCHOOLS ....................................................................................................... 105 Background .......................................................................................................... 105 Methods ................................................................................................................ 108 Procedures ................................................................................................ 1 08 Instruments ............................................................................................... 1 10 Analysis ................................................................................................................ 111 Results .................................................................................................................. 112 Challenges to promoting healthy eating .................................................. 114 Financial challenges ..................................................................... l 14 Economic influences ................. ' ................................................... 1 14 Family influences ......................................................................... 115 School foods — meals ................................................................... l 16 School foods — competitive foods ................................................ 1 l7 Perceptions about students ........................................................... 118 Peer influences ............................................................................. 1 l9 Accomplishments to promoting healthy eating in middle schools .......... 120 School food improvements .......................................................... 120 Nutrition education ...................................................................... 121 Factors that facilitate change in schools .................................................. 122 School culture .......................................................................................... 124 Discussion ............................................................................................................ 125 CHAPTER 5: IMPLICATIONS AND CONCLUSIONS .............................................. 130 School wellness policies ...................................................................................... 130 Association between the school nutrition environment and student dietary intake .................................................................................................................. 133 Barriers and accomplishments to healthy eating, and factors that facilitate positive school nutrition changes ...................................................................................... 138 School culture ...................................................................................................... 139 A coordinated school health approach ................................................................. 141 Conclusion ........................................................................................................... 143 APPl BIBL APPENDICES Appendix A: Scoring for each question in the Wellness Policy Evaluation Tool .................................................................................................................. 145 Appendix B: Mean wellness policy comprehensiveness and strength scores by section and by template type used among SNAK school districts ..................... 150 Appendix C: Association between school-level characteristics and wellness policy template type used .................................................................................... 154 Appendix D: School-level influences on wellness policy total comprehensiveness and total strength scores ...................................................................................... 155 BIBLIOGRAPHY ............................................................................................................ 156 xi nnnnnnn Tail 317101 puht Tar? Milt“ Ifii‘l: SCilUi Tablt marl Idi‘lL lath Silldc Tal‘ic APPCI APPtr and i.“ ‘APPCr 13mph APP‘Cn 1013i 31 LIST OF TABLES Table 2-1: Description of variables from the School Environment and Policy Survey reported by administrators and food service directors. ...................................................... 63 Table 2-2: Mean wellness policy comprehensiveness and strength scores by section and template type used among SNAK school districts ............................................................. 69 Table 2-3. Differences in wellness policy mean comprehensiveness and strength scores based on how districts modified the MASB policy template among SNAK school districts ............................................................................................................................................ 70 Table 2-4. Percentage of school districts meeting federal wellness policy requirements among SNAK school districts. Table 2-5: Concordance between written wellness policies and administrator-reported school nutrition policies ............................................ 71 Table 2-5: Concordance between written wellness policies and administrator-reported school nutrition policies ..................................................................................................... 72 Table 2-6: Concordance between written wellness policies and administrator-reported school nutrition practices ................................................................................................... 73 Table 2-7: Concordance between written wellness policies and FSD-reported food service practices ............................................................................................................................. 74 Table 3-1: Differences in student dietary intake by race and gender ............................... 95 Table 3-2: Availability and type of competitive foods available and association with student dietary intake ......................................................................................................... 97 Table 4-1: Characteristics of SNAK case study schools and interview participants ....... 1 13 Appendix A: Scoring for each question in the Wellness Policy Evaluation Tool ........... 145 Appendix B: Mean wellness policy comprehensiveness and strength scores by section and by template type used among SNAK school districts ............................................. 150 Appendix C: Association between school-level characteristics and wellness policy template type used ........................................................................................................... 154 Appendix D: School-level influences on wellness policy total comprehensiveness and total strength scores ......................................................................................................... 155 xii LIST OF FIGURES Figure 1: Dissertation objectives ...................................................................................... 4 Figure 2: Ecological model of the influences and outcomes of adolescent dietary behaviors ......................................................................................................................... 10 Figure 3: Ecological model of the school-specific influences on adolescent dietary behaviors ......................................................................................................................... 13 xiii groml‘ includi Srfm'xz pm tilt decide indicit mestin nudcn: Discus. 12-19 3 higher 31- 3011 fOI nun promm dlSeagc COmml CHAPTER 1 : INTRODUCTION Healthy eating during childhood and adolescence is critical to ensuring proper growth, development, and functioning, as well as to prevent many chronic diseases including cardiovascular disease, diabetes, obesity, and osteoporosis (Berenson, Srinivasan et a1. 1998; Weaver 2000; Prentice, Schoenmakers et al. 2006). The prevalence of overweight in US. adolescents has more than tripled over the last several decades from 5.0% to 17.4%, (Ogden, Flegal et a1. 2002; Ogden, Carroll et al. 2006), indicating that children are not receiving optimal nutrition. Many US. children are not meeting important nutrition recommendations. For example, in Michigan only 17% of students consume 5 or more servings of fruits and vegetables per day (Centers for Disease Control and Prevention. 2005). Nationally, the mean intake of fat in adolescents 12-19 years old is 32% of total calories, and saturated fat intake is 11% of total calories, higher than the recommended levels of <30% and <10%, respectively (Wright, Wang et al. 2003). Many organizations and researchers have identified schools as an important site for nutrition promotion and interventions to reduce the prevalence of childhood obesity, promote overall health and wellbeing, and prevent adult chronic diseases such as heart disease and diabetes (Michigan Department of Education, Michigan Department of Community Health et al. 2001; American Dietetic Association 2006; Story, Kaphingst et a1. 2006). The goal of this dissertation research is to provide insight into the associations among school nutrition policies, environments, practices, and student dietary behaviors and to advance knowledge of the school-specific factors associated with healthy eating in kl“ - 113K! amo cm i Iht‘Sl furl lire 6513.: Stilt). 50hr). 31111 I Fink} prel‘g 300p} A530i low—income Michigan middle schools. A better understanding of these school-specific factors will inform legislative and intervention efforts to promote health and well-being among children. School nutrition interventions commonly include changes to the nutrition environment, nutrition education, and/or implementation of nutrition policies. Results of these interventions have been mixed and lack consistency, emphasizing the need for further research to understand how to effectively promote healthy eating within schools (French, Story et a1. 1997; Baranowski, Davis et al. 2000; French, Jeffery et a1. 2001; Sallis, McKenzie et a1. 2003; F ulkerson, French et al. 2004; Lytle, Murray et al. 2004; Engels, Gretebeck et al. 2005; Cullen, Watson et a1. 2006; Lytle, Kubik et a1. 2006; Wojcicki and Heyman 2006; Slusser, Cumberland et al. 2007). The US. Federal government also acknowledges the important role of schools in promoting lifelong health and well-being to children. The Child Nutrition and W1C Reauthorization Act of 2004 (Section 204 of Public Law 108—265 June 3, 2004) required all local education agencies (school districts) receiving fimding for school meals to establish a local wellness policy by July 15‘, 2006. Several nation-wide studies examined school nutrition policies prior to this federal mandate, at which time less than half of all school districts had adopted wellness policies or other policies to promote healthy eating and physical activity (Greves and Rivara 2006; O'Toole, Anderson et al. 2007; Finkelstein, Hill et al. 2008). Studies of individual states and national wellness policy prevalence afier the federal mandate took effect indicate that the majority of schools have adopted a policy and are in compliance with federal regulations (School Nutrition Association 2006; Metos and Nanney 2007; Moag-Stahlberg, Howley et a1. 2008; \\ iii sch dirt n': Probart, McDonnell et al. 2008; Belansky, Cutforth et al. 2009; Longley and Sneed 2009). However, most of these researchers note that wellness policy language is often ’9 6‘ weak and suggestive, using phrases such as “shall strive to, when possible,” or “will attempt to,” (Metos and Nanney 2007; Probart, McDonnell et al. 2008; Belansky, Cutforth et a1. 2009). School wellness policies may hold promise for improving school nutrition and health practices; however, it is important to determine whether these policies are being translated into school practices, and how to assist schools in creating effective wellness policies. Figure 1 illustrates the three primary objectives of this dissertation research. The first objective is to examine the associations between written school wellness policies and school nutrition policies and practices reported by school administrators and food service directors (F SDs). The second objective is to examine the associations between the school nutrition environment and student dietary intake. The third objective is to qualitatively examine the facilitators and barriers to promoting healthy eating in low-income Michigan middle schools. This qualitative component adds an in-depth understanding of the challenges and accomplishments schools experience in promoting healthy eating. liter; Cum the \ 35 “I Figure 1: Dissertation objectives School nutrition policy IObjective 3 I IObjective 1 I . Facilitators & Bamer S School nutrition environment \ | Objective 2 | Student dietary intake The remainder of this chapter is devoted to reviewing the relevant school nutrition literature. First, the importance of a healthy diet during childhood and adolescence, and current dietary and health trends are reviewed. Next, an ecological model that describes the various levels of influence on adolescent dietary behaviors both in a general context, as well as those influences specific to the school context is discussed. The school- specific influences on adolescent dietary intake directly related to this dissertation are reviewed in detail including: the federal wellness policy mandate, related resources, and current research; environmental influences on student dietary intake including school meals (breakfast and lunch) and competitive foods (e.g. a la carte, vending machines); and interpersonal (e. g. peer influence, role modeling) and intrapersonal (e.g. knowledge) dietary influences associated with the school setting. Additionally, the importance of dexeh Ph}8h SCth] nuakc dhprnj childhu dheasc. 19%; \l SChOenn iOWET n'S school administrators and other leaders in influencing these school-related factors is discussed. Next, school nutrition intervention studies are reviewed. As hundreds of school nutrition interventions have been published to date, an exhaustive review of these interventions is not necessary. Selected intervention studies that represent the diversity of the current literature are reviewed to provide the reader with an appreciation of: l) the variety of interventions that have been implemented, 2) the inconsistency in the results of these studies, and 3) the limitations of the school nutrition literature. Lastly, the justification for the research presented in the remaining chapters of this dissertation is described. THE IMPORTANCE OF NUTRITION DURING CHILDHOOD AND ADOLESCENCE Dietary intake can have both short and long-term impacts on grth and development, cognitive, emotional, and behavioral functioning, as well as academic and physical performance (Murphy, Pagano et al. 1998; Weaver 2000; Taras 2005; Prentice, Schoenmakers et al. 2006; Stevenson 2006; F anjiang and Kleinman 2007). Poor dietary intake during childhood and adolescence can have long-term health consequences, which disproportionately affect low-income and minority populations. Proper nutrition during childhood and adolescence can help to prevent chronic diseases such as cardiovascular disease, diabetes, obesity, osteoporosis, and certain cancers (Berenson, Srinivasan et al. 1998; Weaver 2000; Joint FAO/WHO expert consultation on diet 2003; Prentice, Schoenmakers et al. 2006). Intake of fruits and vegetables has been associated with a lower risk of obesity, type 2 diabetes, certain cancers, stroke, and cardiovascular disease 3111 has IN Suu &S n Sign 3Clh (US. Department of Health and Human Services and US. Department of Agriculture 2004). Low calcium intake may result in a lower peak bone mass, which has been correlated with an increased risk of hip fracture later in life (Heaney, Abrams et al. 2000). Low-income children have been shown to have lower intakes of calcium and fruits and vegetables (Neumark-Sztainer, Story et al. 1996; Fox and Cole 2004). Consumption of foods high in added sugars has been associated with dental caries, which are more common in low-income children and certain ethnic groups including African Americans, Hispanics, and American Indians (Lingstrom, Holm et al. 2003; DHHS 2004). The impact of poor nutrition goes beyond physiological consequences related to growth, development, and disease relationships. Malnutrition and food insecurity have been shown to have negative effects on psychosocial and cognitive development and behavior (Murphy, Wehler et al. 1998; Alaimo, Olson et al. 2001; Taras 2005; F anjiang and Kleinman 2007). Iron deficiency has been consistently associated with cognitive, behavioral, and learning difficulties in children (Pollitt 1997). Academic performance has been associated with dietary patterns such as eating dairy products, nutrient dense foods, and low-quality foods (F u, Cheng et a1. 2007). These findings are especially relevant to school professionals, because it is clear that in order to be able to learn, a child must be properly nourished. With the No Child Left Behind Act (Bush 2001) placing an emphasis on academic achievement, and fimding sources that are tied to standardized test scores, schools often prioritize core subjects such as math and reading over health and physical education, which do not appear on standardized tests. Oftentimes, the schools' limited financial resources are reserved for activities directly related to core subjects, which could negatively impact the amount of \ en. that of ti prm from dun Zita: mini nutr achi TRE 0\‘er al. 3( risk 1‘ El al. 2004; funding available to other programs, such as school meals. Compounding the problem, some schools rely on food service sales of competitive foods through a la carte and vending machines to provide additional revenue for the school. However, it is apparent that academic achievement depends, at least partially, on the health and nutritional status of the students. Many schools are acutely aware of this relationship, as indicated by provision of breakfast and snacks to students during standardized testing periods. Results from one study indicated that some schools provide more nutritious lunches to students during testing periods, which resulted in improved performance (F iglio and Winicki 2005). However, unless this effort is continued throughout the school year, it will have a minimal impact on overall student learning. Consistent prioritization of health and nutrition programs may be a more effective means of improving student academic achievement and test scores. TRENDS IN OBESITY AND DIETARY INTAKE Health and dietary trends indicate that adolescents are not receiving optimal nutrition, and that low-income and minority groups may be at increased risk. The prevalence of obesity (those having a body mass index (BMI) above the 95th percentile for age and gender) in adolescents aged 12-19 years old in the US. has more than tripled over the last three decades from 5.0% in 1976-80 to 17.4% in 2003-04 (Ogden, Flegal et al. 2002; Ogden, Carroll et al. 2006). Another 16.9% of adolescents are considered at risk for obesity (those having a BMI between the 85th and 95th percentile) (Ogden, Carroll et al. 2006). Minorities consistently have a higher prevalence of obesity (Sorof, Lai et al. 2004; Jago, Harrell et al. 2006; Ogden, Carroll et al. 2006). The relationship between 50 an ‘1'} 0?? no: \u‘ exam; Willie . less m] OfSChu income and big] socioeconomic status or food security and obesity rates is unclear and varies by age, sex, and ethnicity (Chang and Lauderdale 2005; Wang and Zhang 2006; Dinour, Bergen et al. 2007) Trends in adolescent dietary intake also indicate adolescents are not receiving optimal nutrition. Our youth typically eat foods that are high in energy-density but low in nutrient-density (Subar, Krebs-Smith et al. 1998; Kant 2003). National Health and Nutrition Examination Survey (NHANES) data from 1999-2000 and 2001-2002 indicate dietary intake of fi'uits, vegetables, calcium, magnesium, potassium, phosphorous, vitamin A, vitamin E, vitamin C, and fiber are below the recommended levels in both boys and girls, and dietary intake of iron, zinc, and folate are lower in female adolescents than in males (Wright, Wang et al. 2003; Ervin, Wright et al. 2004; Moshfegh, Goldman et a1. 2005; Institute of Medicine 2006). At 32.0% and 11.3% of total energy intake, total fat and saturated fat intakes were higher than the recommended levels of <30% and <10%, respectively (Wright, Wang et al. 2003). Dietary trends in adolescents show that sodium intake has increased by approximately 50% while calcium intake has decreased (Briefel and Johnson 2004). Socio-demographic characteristics have been associated with dietary intake. For example, Caucasian children typically consume more added sugars, soft drinks, and milk while African American and Hispanic children consume more fat and saturated fat and less milk (Institute of Medicine 2007). Using 1988-1994 NHANES data, nutrient intakes of school-age children and adolescents (5-18 years) were compared based on household income level: lowest income (930% of poverty); low income (131-185% of poverty); and higher income (2185% poverty) (Fox and Cole 2004). The lowest income group had furl Slit significantly lower intake of iron and calcium, significantly higher intakes of fat, saturated fat, cholesterol, and fiber than the low and higher income group did (Fox and Cole 2004). Other studies have also shown that low-income youth are less likely to have a healthy diet, and consume more fat and saturated fat, and fewer fruits and vegetables, further emphasizing disparities in dietary intake by income group (Neumark-Sztainer, Story et al. 1996). INFLUENCES ON ADOLESCENT DIETARY BEHAVIORS — AN ECOLOGICAL MODEL Individually-focused intervention efforts to improve dietary intake and prevent overweight and obesity in adolescents have had varied success; some researchers believe this is due to the fact that they typically don’t address environmental or societal influences on eating and physical activity (Sallis and Owen 2002). Interventions that address multiple levels of factors that influence adolescent dietary intake are beginning to emerge. Story and colleagues (Story, Neumark-Sztainer et al. 2002) developed a conceptual model of the various levels of influence on adolescent dietary behaviors based on ecological and social cultural models of health behavior theory. These models emphasize not only the characteristics of the individual, but also the interaction between an individual and their environment (Sallis and Owen 2002). In Story’s model, there are four primary levels of influence: individual (intrapersonal); social environmental (interpersonal); physical environmental (community settings); and macro-system (societal) (Story, Neumark-Sztainer et al. 2002). .l.‘:’11rl.l . lhrultlf\~F\l\u~ . at as.---HN moon Gucci—968 ._m n scanmiaesoz £28820 2:: J83 233:4 . 5on Sec coin—E $539.02 3.5.80 EwEB .mEotma :85 was 325 Reaganom . 8582.0 Amuse Sweat 32%—2m . $85535 12832-5 6me $0350.23 ago—392 :auegaaom Mtg—on £8393 Eoomocoamm . 90555820 aesseé 85330 Egan—2‘95:— omEocao< A $25.5 / 3.3.: >335 QUEQEHOHHQA— HEO—vam / note—.32 Aliv can 5.39.5 .3???— .fi 25o: 3208a #8an 300m - .EoEooSuEB mam—€22 . E8 was . museum . 8338:: . 353m . 8532.0 5.3: $850 53> mcouofififiv 3:932:85 £22 wcaaonm . moon—9.83 . mecca oocoEoEoQbooEO . 0:8: . Begum . uoofonzw_o:\b_§EEoU . 3.8% £5588“ as $5253 iv Eofiaegam .333.— Ar seas .855 480: $52 was 86:8 . mEBmxm 355$me can 552682 boom . mace: 3.523 28 38m . wfifltgwflflcoz . $08265 .8288 illiv £0532an £0323 benefit E08203 mo 888:5 use 8053?: 05 mo 3on _mo_wo_oom ”N oSwE 10 sod J. 3w At each level, many factors influence dietary behavior, as shown in Figure 2 adapted from Story (Story, Neumark-Sztainer et al. 2002). At the individual level, psychosocial (e.g., food preferences, knowledge, self-efficacy), biological (e.g., hunger, gender), behavioral (e.g., meal and snacking patterns, weight control practices), and lifestyle factors (e. g., convenience, cost, meal patterns) influence adolescents’ decisions of what to consume. Social environmental influences include family characteristics (e. g., demographics, family meal patterns, and availability of foods) and peer influences (e.g., social norms). Physical environmental influences include all of the settings in which adolescents can obtain foods and beverages or are exposed to food and nutrition messages. These include, for example, homes, schools, fast-food restaurants, vending machines, convenience stores, and worksites. Macro-system (societal) influences on adolescent eating behaviors include media and advertising (e.g., television commercials, billboards), cultural and societal norms, the food production and distribution systems, and policies related to foods and beverages (at local, state, and national levels). One key concept in ecological and social cultural theoretical perspectives is reciprocal determinism, meaning that all levels have the potential to influence one another, represented by the double-headed arrows between each level of influence in Figure 2 (Sallis and Owen 2002; Story, Neumark-Sztainer et al. 2002). Although ecological models hold promise for improving adolescent dietary intake by addressing the multiple levels of influence on dietary behaviors, little evidence currently exists examining the efficacy of these models (Sallis and Owen 2002). Additionally, because multiple intervention activities are taking place simultaneously, it is difficult to determine the effectiveness of individual components within a multi-level intervention. 11 ado CW8 ado] inter \R' H SCH The research in this dissertation examines multiple levels of influence on adolescent dietary behaviors (policies, practices, and the nutrition environment) using cross-sectional data. Results will help identify school-related factors associated with adolescent dietary behaviors, which may assist researchers and practitioners in designing intervention activities most likely to result in dietary improvements. WHY STUDY SCHOOLS? APPLYING THE ECOLOGICAL MODEL TO THE SCHOOL SETTING Schools have been identified as an important setting for nutrition promotion and interventions to reduce the prevalence of childhood obesity, promote overall health and wellbeing, and prevent chronic disease (Michigan Department of Education, Michigan Department of Community Health et al. 2001; American Dietetic Association 2006; Story, Kaphingst et al. 2006). Schools reach over 95% of US. children, and provide a cost-effective opportunity to reach low-income and minority children (who are at a higher risk of nutritional inadequacy) that might not be reached through traditional means, such as doctors’ visits (Story 1999; Story, Kaphingst et al. 2006). The nation’s Healthy People 2010 goals also recognize the importance of schools in promoting health by calling for an increase in “the proportion of children and adolescents aged 6 to 19 years whose intake of meals and snacks at schools contributes proportionally to good overall dietary quality” (US. Department of Health and Human Services 2000). 12 _ 83 ._m 3 bES~m¢tw5=oZ .me 58.. 3584c. 3052558 £58505 05: .6on 036mm: . meson 58:3 58 588.? $52222. 6558 Emma? .5533 20:30:58 «£80.30. xomcm 58 .355 Ecomifiom . 523585 820595.56? $8580.35 mvoom 0253500. .893 Sumo—bouzom 63230.2 £32 Begum. £023 .355va Efiomosoxmm . 03233 mowfigon Ea mwoom. $858835 3:232: 388 he baammoooa Ea bzsazgfi 2523.525 88 £8: .aueflonahu— 5055.5»5— 5393.— anum 3:255 : 3583 e Exam Begum 53522 855500 5.8: 3:35— \ TI 508on outflow woo,”— Efiofi— floaab££5c< Begum Sufism / I 9322.335 .3. L. 1 59555535 . mania .00an 3.550 . 58:03 50505356. :85 92.5 3035.52”— . «€80wa . 8:605:50 . 50:085.. 350585qu 505: Z NCSUI‘J menu 5pm l‘ dill‘cr' the C pract ucllr to fit \ariu bell» idea SCht‘ Lil?)- em: wellness policies. The content and focus of the wellness policy templates varies, presumably based on the priorities and intent of the sponsoring organizations. Several of the most common wellness policy templates and resources, as well as those that are specific to Michigan schools, are reviewed briefly below. The depth and quality of these policy templates varies widely, and may be due to differences in the organizations and individuals that created the document. For example, the Center for Ecoliteracy template policy emphasizes local and sustainable agricultural practices, while most other policies do not mention this. Most organizations with wellness policy templates encourage school districts to modify the template as necessary to fit their unique strengths, limitations, and needs to be most effective. Given the variation in templates, school districts should examine multiple wellness policy templates before adopting a policy. Ideally, districts would create their own policy that combines ideas from various templates, along with their own unique ideas, in order to best fit their school. United States Department of A griculture (USDA) The USDA Local Wellness Policy website can guide school districts through the entire wellness policy process, from creating a wellness team, performing a baseline - school wellness assessment, drafting and adopting a policy, to implementing and monitoring the policy (www.fns.usdagov/tn/Healtfl/wellncsspolicv.html). This website also provides sample statements that districts can use for their wellness policies, links to sample policies from states and other organizations, as well as additional resources and implementation tools. 19 Nani/n. nation; to dcx. school MT\CI Teani (I)( IN) and POI (Te L’r National Alliance for Nutrition and Activity WANA) NANA (www.nanacoalition.org) convened a working group of more than 50 national and state experts in the areas of nutrition, physical activity, health, and education to develop a model wellness policy (wwwschoolwcllnesspolicies.org). NANA suggests schools complete a baseline self-assessment, such as the Center for Disease Control and Prevention’s (CDC) School Health Index, (https://apps.nccd.cdcgov/shi/dcfaultaspx), Team Nutrition’s Changing the Scene (http://www.fns.usda.gov/tn/Resourccs/changinghtml), or the National Association for Sport and Physical Education’s (NASPE) Opportunity to Learn Standards for Elementary, Middle, and High School Physical Education (http://wwwaahperd.org/Naspe/pdftfiles/Opportunity%20to%201earn%20final%20%20 Middle%2OSchool.pdf ) to identify and prioritize goals for their wellness policy. The NANA wellness policy website also contains a list of resources that can help districts develop, implement, and evaluate their wellness policy. NANA also emphasizes use of CDC’s Coordinated School Health Model (Centers for Disease Control and Prevention 2009) to address other aspects of wellness that are not required in the federal mandate and are not included in model wellness policies, such as mental health and food safety policies. C enter for Ecoliteracy Inspired by the work of the Child Nutrition Advisory Council of the Berkeley Unified School District, the first school district in the nation to create and adopt a wellness policy in 1999, the Center for Ecoliteracy, Slow Food USA, and the Chez 20 Pani: 0}: than mm .‘iti bra! n1 Panisse Foundation created a Model Wellness Policy Guide (www.ccoliteracyorg/programs/wellness policyhtml). This is more of a guide rather than a template policy, and provides instructions that districts are encouraged to follow as they create their wellness policy. This guide emphasizes collaboration between the school and community, and creating a school culture that supports health and wellness. This guide focuses on local and sustainable agricultural and environmental practices more than other template policies. Action for Healthy Kids The Action for Healthy Kids coalition developed a website that guides school districts through a set of eight steps to create their local wellness policy (wwwactionforhealthykidsorg/wellnesstool/indexphp ). These steps include: gathering relevant information; developing a wellness team; conducting a needs assessment; drafting a policy; building awareness and support; adopting the policy; implementing the policy; and maintaining, measuring, and evaluating the policy. For each of these steps, the website provides links to relevant resources and to a “Virtual Wellness Policy Team” that provides answers to common questions that districts have. Michigan Department of Education The Michigan Department of Education, in collaboration with other state and local organizations, agencies, and citizens, developed a model wellness policy that was adopted by the Michigan State Board of Education (Michigan Action for Healthy Kids Fall 2007). This template policy focuses on creating a school environment that provides 21 mud: hcdt poiic “Hca 02$: Kl'dS hcahl Pulllt‘j b)3l3 thatu COnun Staten “EHng O’hc’r ”Utrim Genen That lin and mi “’hllef‘ students with consistent messages and opportunities to practice what they learn about healthy eating and physical activity. Also to assist schools in adopting a local wellness policy, the Michigan Action for Healthy Kids and partnering organizations created the “Healthy School Toolkit: Your Guide to Action” (www.tn.fcs.msue.msu.edu/HealthchhoolToolkit.html) (Michigan Action for Healthy Kids Fall 2007). This comprehensive guide provides schools with resources to support healthy eating and physical activity. Policy Company Another wellness policy template is available to schools in several states served by a company that provides school districts with template school board policies to ensure that they are in compliance with all local, state, and federal mandates. This template contains introductory paragraphs in each section, followed by a checklist of optional statements. Districts check the boxes of statements that they want included in their local wellness policy. No other resources are provided with this template policy. Other school health and nutrition policies Several organizations also provide model policies for specific topics, such as nutrition standards for competitive foods in schools. The Alliance for a Healthier Generation partnered with beverage manufacturers to create school beverage guidelines that limit the sales of certain drinks to students during the instructional day. Elementary and middle schools are restricted to water and 8 oz servings of milk and 100% juice, while high schools are able to sell other low—calorie beverages. The Alliance provides 22 nutrit lu\\\ C0n\t C01??- SNUFI (onlj proxi Water that l. the “Beverage Guide Implementation Kit” to assist schools in implementing a healthy beverage policy. The Alliance also partnered with major food manufacturers to create nutrition guidelines for competitive foods available in schools (www.healthiergencration.org). The Institute of Medicine (IOM) along with the CDC, as directed by Congress, convened a team of national experts in order to set nutrition recommendations for competitive foods available in schools. These guidelines set limits on total calories, fat, saturated fat, trans fat, sugar, sodium, beverages containing non-nutritive sweeteners (only allowed in high schools after the instructional day), and caffeine. Also included are provisions prohibiting use of food as a reward, ensuring access to free potable drinking water; and sports drinks are only available to students participating in sports activities that last one hour or more (Institute of Medicine 2007). School policies after child nutrition reauthorization Evidence indicates that the federal wellness policy mandate has increased health- promoting policies in school districts across the nation. Longley and Sneed (Longley and Sneed 2009) examined the extent of wellness policy components before and after the federal mandate took effect in a national sample of school districts. Prior to the mandate, 363 food service directors reported one-third of wellness policy components were in place. Afier the wellness policy mandate, nearly three-quarters of these components had been implemented (Longley and Sneed 2009). Several studies have examined the extent and content of local wellness policies in national samples of school districts after the mandate took effect. The first study by the 23 School Nutrition Association examined wellness policies from a national sample of 140 school districts in late 2006, just afier the mandate took effect (School Nutrition Association 2006). Nearly all district wellness policies mandated nutrition standards for school meals (99%) and competitive food standards (89% addressed a la carte, 87% addressed vending machines); addressed nutrition education (85%) and physical activity (94%); and had a plan for wellness policy implementation and evaluation (89%) (School Nutrition Association 2006). Another study examined 256 wellness policies from 49 states (Moag-Stahlberg, Howley et al. 2008). Results indicate that the majority of policies (68%) addressed all of the federal requirements (Moag-Stahlberg, Howley et al. 2008). More specifically, 81% included goals for nutrition education, 79% included goals for physical education, 88% addressed other school-based activities, 81% involved the community and/or families, and 78% addressed school meal standards. Lastly, while 73% of district wellness policies addressed policy implementation, oftentimes they included little detail regarding the manner in which the policy was to be implemented, or how implementation would be tracked. Without a specific plan for implementation and evaluation, wellness policies may lose their momentum and may not have the intended impact of creating a healthier school environment. Several studies have also examined wellness policies in individual states. Wellness policies fiom 75% of the public school districts in Utah were evaluated to determine if they met federal policy requirements (Metos and Nanney 2007). Analysis results showed that 78% of district policies met all of the federal requirements. Probart and colleagues examined local wellness policies in all Pennsylvania public school districts in early 2007, shortly after the wellness policy mandate went into effect (Probart, 24 Nlcl)« mesh nutrn based \L i100 185.6i 2009. POHC} itcnis SC i100 €\alua Uhoug Cuflbr McDonnell et a1. 2008). All public school districts had submitted a wellness policy to the state. The majority of districts met each of the federal requirements including nutrition education goals (100%), physical activity goals (99.8%), goals for other school- based activities (100%), nutrition guidelines for all foods available (99.8%), ensuring school meals meet USDA standards (99.0%), and a plan for measuring implementation (85.6%) (Probart, McDonnell et a1. 2008). Belansky et al. (Belansky, Cutforth et a1. 2009) evaluated wellness policies from 32 rural, low-income school districts in Colorado. Policy evaluation indicated that wellness policies only addressed about half of the 96 items examined, however only 15% of these items had required and specific strategies. School districts scored highest in goals for nutrition education and a plan for policy evaluation, and lowest in standards for school meals and goals for physical education (though physical education goals are not included in the federal mandate) (Belansky, Cutforth et al. 2009). These studies indicate that the federal mandate has been successful at increasing the number of schools that have written school nutrition/wellness policies. The federal mandate did not give districts specific guidelines, which gave schools the freedom to create a policy that meets their specific needs. Unfortunately, this has also resulted in a high degree of variability in the quality of wellness policies, as well as the implementation and enforcement of such policies (Institute of Medicine 2007). Researchers from the studies in Utah, Pennsylvania, and Colorado all noted that wellness policy wording included weak statements that often included qualifiers such as “will 9, 6‘ strive to, when possible,” “is encouraged,” and “will attempt.” These ambiguous statements make goals difficult to implement and measure. These types of statements act 25 mOI’C n0n4 lane em in oppor hedhl under nUU1U mand. SCHl ) dietary School and re; mescl nutn'm LUnch bicakf; ll’On‘ V1 nulllen more like recommendations rather than requirements, and are difficult to enforce due to non-specific language that can result in differences in interpretation. School wellness policies hold promise for improving school nutrition; however, little research on the efficacy of these policies at influencing school nutrition environments and practices has been published to date. This provides an ideal opportunity for researchers to examine the ability of local wellness policies to encourage healthy school environments and practices. The research in this dissertation aims to understand the impact the degree that wellness policies are associated with school nutrition practices, to better understand the impact of the federal wellness policy mandate. SCHOOL NUTRITION ENVIRONMENT School Meals The next section reviews the school environmental influences on adolescent dietary behaviors, primarily the foods and beverages that students are exposed to in the school setting, beginning with school meals (breakfast and lunch). In order to improve and regulate the nutritional quality of foods served in school meals, the USDA enacted the School Meals Initiative for Healthy Children (SMI) in l995, setting minimum nutrition standards for foods served in the School Breakfast Program and National School Lunch Program (US. Department of Agriculture 2001). SMI guidelines require school breakfasts to provide at least of one-fourth the RDA for total calories, protein, calcium, iron, vitamin A, and vitamin C, and school lunches must provide one-third of these nutrients (US. Department of Agriculture 2001). School meals must also meet the 26 1‘ lat It S\ll tar-gs mu St and h dislim comer prcfcn consur Content Student leprcsc Glitter) three q mm] 6, Only 1 27.4“. '0 lunchc. Dietary Guidelines for American recommendations to limit total fat to 30% and saturated fat to less than 10% of total energy intake (US. Department of Agriculture 2001). The SMI guidelines suggest, but do not require, that school lunch provide one third the daily target for dietary fiber (varies based on age), and contain <100mg cholesterol and <800 mg sodium (US. Department of Agriculture 2001). Mean nutrient content of school meals can be represented in two ways: 1) foods and beverages offered to students, or 2) foods and beverages served to students. The distinction between these two representations is that oflering data averages the nutrient content of all of the items available, whiles serving data takes into account student preferences by weighting the data to represent what students are purchasing and consuming. In this way, the school menu might offer foods that average <30% fat content, but if students are only purchasing the higher fat foods, what is served to students may be >30% fat content. The SNDA-III study examined the nutrient content of school meals in a nationally representative sample of schools. In the 2004-05 school year, nearly all middle schools offered lunches that met requirements for protein, calcium, and vitamin C, approximately three quarters met requirements for iron and vitamin A, and 58% met requirements for total energy (US. Department of Agriculture, Food and Nutrition Service et al. 2007). Only 16.7% of middle schools offered lunches that met requirements for total fat, and 27.4% met requirements for saturated fat, with the average total fat content of school lunches offered at 34% and saturated fat 11% of calories (US. Department of Agriculture, Food and Nutrition Service et a1. 2007). Most middle school lunches offered met recommendations for cholesterol and dietary fiber, however almost none met 27 smhmn d.200' SChUtll.‘ math: nhddk \hhnh huhnh mghcr 3FPI0\ H;8.l indkaz lkehm be!“ c( metg Rhmfi mosrd 2007). fa16$ [1 61211.2 10m] d pl’m-id sodium recommendations (US. Department of Agriculture, Food and Nutrition Service et al. 2007). When taking into account what students actually choose to consume, fewer schools met the SMI standards. Nearly all middle schools served lunches to students that met the protein requirements; however, only 38.5%, 42.8%, 66.1%, 83.4%, and 55.2% of middle schools served lunches that met SMI requirements for total energy, vitamin A, vitamin C, calcium, and iron, respectively (US. Department of Agriculture, Food and Nutrition Service et al. 2007). Middle school lunches served to students were slightly higher in % energy from total and saturated fat than those offered to students, with approximately the same proportion of schools meeting SMI requirements for fat content (US. Department of Agriculture, Food and Nutrition Service et al. 2007). These data indicate that even though schools may offer healthier options, that students are more likely to choose the less-healthy options. These data also highlight the difference between the nutrient content of foods and beverages available within school meals and the USDA SMI requirements. According to the SNDA-lll study of a nationally representative sample of US. schools, 22.1% and 73.8% of students report consuming school breakfast and lunch on most days of the week (US. Department of Agriculture, Food and Nutrition Service et al. 2007). Students who are eligible for free and reduced-price meals participate at higher rates than those not eligible (US. Department of Agriculture, Food and Nutrition Service et al. 2007). Foods and beverages consumed at school are an important contributor to the total dietary intake of adolescents. School lunch (which most students consume) provides nearly a third of students’ total daily energy intake, and generally provides a 28 21'6le lunch .Agnc cenai the)‘a COHSU Sll'et’lt Frencl 20091 18.)“) SlalUS 1 fries in high Sc fOOds (‘ HSSOQja flesh fl aSSOQI'a Offering Creplns greater proportion of total vitamin and mineral intake in students that consume school lunch, compared with those that do not eat school-provided meals (US. Department of Agriculture, Food and Nutrition Service et al. 2007). Students that participated in school breakfast and lunch were less likely to have certain nutritional inadequacies, compared with children that did not participate, however they also had higher sodium intakes (Clark and Fox 2009). School meal participants consumed fewer energy-dense foods at school, consumed fewer calories from sugar- sweetened beverages, but had higher intakes of low-nutrient energy dense foods (e. g. French fried, baked goods) when compared to non-participants (Briefel, Wilson et al. 2009). School breakfast participation has been associated with lower body mass index (BMI) in a nationally representative cross-sectional sample (Gleason and Dodd 2009). It is clear that school meals can play an important role in providing important nutrients to students. Specific school meal practices can also impact the dietary quality and weight status of schoolchildren. Analysis from the SNDA-III study show that not serving French fries in school meals resulted in decreased consumption of sugar-sweetened beverages in high school students (41 calories) and reduced consumption of low-nutrient energy dense foods (43 calories) (Briefel, Crepinsek et al. 2009). School meal practices were also associated with positive dietary intake trends in elementary school students. Offering fresh fi'uit and raw vegetables daily, and not offering French fries in school meals were associated with increased intake of vegetables (Briefel, Crepinsek et al. 2009). Not offering desserts in school meals was associated with increased intake of fruit (Briefel, Crepinsek et al. 2009). In elementary schools, offering French fries or desserts more than 29 once l)0dc andh apnoi Com not dd define \dllr‘r IUUIL an}\\h (FNlN‘ thesch COmpc brougl once per week in school lunches was associated with a higher likelihood of obesity (F ox, Dodd et al. 2009). Clearly, school meals have a significant impact on the dietary intake and health outcomes of students. Thus, it is imperative that school meals provide healthy options and make a positive contribution to dietary intake. Competitive foods The SMI regulates only foods and beverages available in school meals, but does not address competitive foods sold or served within the school. Competitive foods are defined by the USDA as any foods or beverages available in schools outside of the National School Lunch and School Breakfast programs (U.S. Department of Agriculture 2001). These include foods available individually (not as part of a school meal) anywhere in the school anytime of the day, as well as foods of minimal nutritional value (F MNV), which cannot be sold in the foodservice area (but can be sold elsewhere) during the school meal periods (U.S. Department of Agriculture 2001). Examples of competitive foods include vending machines, a la carte offerings, fundraisers, treats brought into the classrooms by teachers or parents for celebrations or rewards, and concession stands at school events. Competitive foods and beverages are widely available in schools. Several national studies in 2005 estimate that approximately 73-83%, 97%, and 99-100% of elementary, middle, and high schools had any type of competitive foods and beverages available to students (U.S. Government Accountability Office 2005; US. Department of Agriculture, Food and Nutrition Service et a1. 2007; Fox, Gordon et al. 2009). More specifically, 64%, 89%, and 92% of elementary, middle, and high schools had a la carte 30 food nndd Andi beta; deny “ihl 2003 Zinn) assoc COmp rniddl nndd] fruit a (cunt Pracri SChOC foods (Fox, Gordon et a1. 2009); and 17-26%, 62-87%, and 86-98% of elementary, middle, and high schools had vending machines available on school grounds (O'Toole, Anderson et al. 2007; F inkelstein, Hill et al. 2008; Fox, Gordon et al. 2009). The high prevalence of competitive foods available in schools is concerning because competitive foods and beverages sold in schools are often low in nutrient density, and high in energy, fat, sodium, and added sugars (Hamack, Snyder et a1. 2000; Wildey, Pampalone et al. 2000; US. Department of Agriculture 2001; French, Story et al. 2003; Wiecha, Finkelstein et al. 2006; O'Toole, Anderson et a1. 2007; Fox, Gordon et al. 2009). Additionally, competitive foods have consistently been associated with poor dietary habits in students. Purchase of competitive foods in middle schools has been associated with a higher intake of sugar-sweetened beverages (Wiecha, Finkelstein et al. 2006); a higher intake of calories, total and saturated fat, and lower intakes of protein, vitamins A and C, and calcium (Templeton, Marlette et al. 2005). A la carte availability has been negatively associated with fruit and vegetable consumption and positively associated with intake of total and saturated fat (Kubik, Lytle et al. 2003). When comparing students from elementary schools with no competitive foods available to middle schools that sell competitive foods, or those transitioning from elementary to middle school, availability of competitive foods was associated with decreased intake of fruit and fruit juices, vegetables, and milk and increased intake of sweetened beverages (Cullen, Eagan et al. 2000; Cullen and Zakeri 2004). Results from the SNDA-III study indicate that school nutrition environments and practices are associated with student dietary intake (Briefel, Crepinsek et al. 2009). School-level characteristics associated with a decreased consumption of energy from 31 sugar-sweetened beverages included not having a contract with a beverage company, not having a store or snack bar, and not having a la carte foods available (Briefel, Crepinsek et al. 2009). Availability of vending machines in or near the cafeteria that contain low— nutrient energy-dense foods was associated with a higher BMI z-score in middle school children (Fox, Dodd et al. 2009). However, having these foods available in a la carte was associated with a lower BMI z-score (Fox, Dodd et al. 2009). Despite this contradictory finding, the majority of research has found that competitive foods are associated with unhealthy dietary behaviors in adolescents. Furthermore, the USDA recognizes that competitive foods in schools may directly undermine nutrition and health education that students may be receiving in the classroom, as well as compete with, stigmatize participation in, and compromise the financial viability of the National School Lunch and School Breakfast program (U.S. Department of Agriculture 2001). In fact, sales of competitive foods are inversely associated with sales of school lunch (Fox, Crepinsek et al. 2001). Efforts to reduce the availability of competitive foods in schools and to enhance the nutritional quality of school meals may help to improve adolescent dietary behaviors and health outcomes. INDIVIDUAL INFLUENCES ON ADOLESCENT DIETARY BEHAVIORS Many studies have examined intrapersonal and interpersonal factors that influence adolescent dietary behavior. Story (Story, Neumark-Sztainer et al. 2002) and Baranowski (Baranowski, Cullen et al. 1999) provide excellent reviews of the current knowledge in the field. One of the factors most relevant to the school setting is provision of nutrition education; however, there appears to be no consistent relationship between nutrition 32 know progr; beha\ rein f0 Pi} ch \aluc Furthe this git import: uwdh) ()l)cat Austral using 2 SZIJlnc] Tina UHdcrsu “halsu “WORN USlng if nan,“ Off00ds aged 11. knowledge and dietary intake. A review of nutrition education interventions showed that programs that focused solely on providing information had limited successes, but behaviorally-focused interventions and those that included environmental changes to reinforce knowledge had more positive results (Contento, Balch et al. 1995). Psychosocial correlates (such as self-efficacy) have been shown to have low predictive value in dietary intake of fat, fruits and vegetables (Baranowski, Cullen et al. 1999). Further understanding of how adolescents make dietary choices is imperative in helping this group choose healthy diets. Qualitative exploration of influences on adolescent dietary behaviors is an important method to better understand adolescent dietary behaviors. Several studies have used focus groups and interviews to elucidate factors that are important to adolescents. O’Dea et al. (O'Dea 2003) conducted 38 focus groups with 213 students in grades 2-11 in Australia. Croll et al. examined adolescent’s perceptions of healthy eating in Minnesota using 25 focus groups with 203 junior and senior high-school students (Croll, Neumark- Sztainer et a1. 2001). A study by Neumark-Sztainer et al. used 21 focus groups with 141 7th and 10th grade students in St. Paul, Minnesota (Neumark-Sztainer, Story et al. 1999) to understand why adolescents eat what they eat, what makes healthy eating difficult, and what suggestions adolescents have for making it easier to eat healthy. Cullen et al. explored the social and environmental influences on fruit, juice, and vegetable intake using 16 focus group discussions with 180 students from six low-income parochial school districts (Cullen, Baranowski et al. 2000). Chapman and Maclean examined the meaning of foods and social context surrounding eating occasions in a group of female adolescents aged 11-18 years old (Chapman and Maclean 1993). The Minnesota Youth Poll explored 33 adolescents’ opinions of nutrition-related topics through small group discussions with 900 high school students (Story and Resnick 1986). Major findings of these studies are compared below. Children and adolescents were consistently able to identify healthy foods (e.g. fruits and vegetables) and their characteristics (e. g. low-fat, low-cholesterol) (Chapman and Maclean 1993; Neumark-Sztainer, Story et al. 1999; O'Dea 2003). Croll noted that adolescents rarely mentioned milk as a healthy item, and that adolescents were able to list many more unhealthy foods (e.g. candy, chips, soda pop) than healthy items (Croll, Neumark-Sztainer et al. 2001). Adolescents were able to identify many short-term benefits of healthy eating including enhanced cognitive functioning, physical performance, and appearance (Croll, Neumark-Sztainer et al. 2001; O'Dea 2003); enhanced self-esteem and pride (O'Dea 2003), and higher self-control (Chapman and Maclean 1993); and references to increased energy and endurance (Croll, Neumark-Sztainer et a1. 2001; O'Dea 2003), and weight loss (Chapman and Maclean 1993). Few adolescents mentioned long-term health consequences (such as prevention of heart attack) (Croll, Neumark-Sztainer et al. 2001). Adolescents were also able to articulate adverse physical and psychological effects of eating unhealthy foods including: experiencing guilt (Chapman and Maclean 1993; O'Dea 2003), disgust, and not being in control of oneself (Chapman and Maclean 1993); being in a bad mood (Story and Resnick 1986); gaining weight (Story and Resnick 1986; Chapman and Maclean 1993); poor health and cavities (Story and Resnick 1986); a slowing down of the mind and body (O'Dea 2003). 34 acknou it the prime this age g with \\ hit Neumark Healthy t} and Resni appeared 1 home emi Neumark-I experience (Chapman Namath—S f0f not con 61 al. 2000) Adi Sp em Prepa healthy low Neumarkfi‘ also indicai, School and 11 (Story and R Despite adolescents' knowledge of healthy eating and its consequences, they also acknowledge that most youth do not have healthy diets (Story and Resnick 1986). One of the primary influences on adolescent dietary intake was taste and preferences. In general, this age group seems to prefer the taste of sweets and salty snack, as well as foods that with which they are familiar (Story and Resnick 1986; Chapman and Maclean 1993; Neumark-Sztainer, Story et al. 1999; Croll, Neumark-Sztainer et al. 2001; O'Dea 2003). Healthy foods were generally perceived as not looking appealing or tasting good (Story and Resnick 1986; Neumark-Sztainer, Story et al. 1999). The social context also appeared to influence what adolescents consume. Healthy foods were associated with the home environment, relatives, and family meals (Chapman and Maclean 1993; Croll, Neumark-Sztainer et al. 2001). Junk foods were associated with many positive experiences, such as parties, socializing, having money, being able to do what you want (Chapman and Maclean 1993), and with social events or hanging out with fi'iends (Croll, Neumark-Sztainer et al. 2001). Students in one study reported peer influence as a reason for not consuming F VJ , as well as advertising for less healthy foods (Cullen, Baranowski et a1. 2000). Adolescents in these studies also cited numerous barriers to healthy eating. Time spent preparing food was mentioned in several studies as a barrier, with the belief that healthy foods take longer to prepare and were not convenient (Story and Resnick 1986; Neumark-Sztainer, Story et al. 1999; Croll, Neumark-Sztainer et al. 2001). Adolescents also indicated that healthy foods are not readily available where they eat, such as at school and in their homes, but especially in vending machines and fast food restaurants (Story and Resnick 1986; Neumark-Sztainer, Story et al. 1999; Croll, Neumark-Sztainer 35 61313001 they may I prohibit at barrier to l consequen. al. 1999: C. identify lor themselt es The to delineate Ol‘healtht’ e. recognized t adolescents included 13,»! pretenr then- in ma barriers to he Snainfl. Stor Resnick 1986 education ab‘. easier [0 eat h term benefits ‘ Slim-ne’- Cl 31_ ‘ et al. 2001; O'Dea 2003). When healthy foods are available in places such as the home, they may not be in a form that is ready to eat (e.g. not peeled or cut up), which can prohibit adolescents from consuming them (Cullen, Baranowski et al. 2000). Another barrier to healthy eating in adolescents was a general lack of concern for the health consequences of an unhealthy diet (Story and Resnick 1986; Neumark-Sztainer, Story et a1. 1999; Croll, Neumark-Sztainer et al. 2001). Even though adolescents were able to identify long-term consequences of poor dietary behaviors, adolescents appear to view themselves as "invincible." The combined results of these studies provide evidence that adolescents are able to delineate healthy and less healthy foods. They were also able to articulate the benefits of healthy eating, and the consequences of unhealthy eating. However, they also recognized that most people their age did not consume healthy diets. Even though adolescents know about healthy eating, the perceived barriers to healthy eating, which included taste preferences, inconvenience, and a lack of prioritization of healthy eating, prevent them from translating this knowledge into healthy dietary behaviors. In many of these studies, adolescents were asked to identify strategies for these barriers to healthy eating. Responses included increased parental support (Neumark- Sztainer, Story et al. 1999; O'Dea 2003); advance planning and preparation (Story and Resnick 1986; O'Dea 2003); use of cognitive motivational strategies; and increasing education about and advertising for healthy foods (O'Dea 2003); methods to make it easier to eat healthy in social situations and to emphasize both the immediate and long- term benefits of healthy eating may be beneficial in this age group (Croll, Neumark- Sztainer et al. 2001); eating more meals with their families (Story and Resnick 1986); 36 making heu of these itei on macros} product pae Story et al. “in behat iors. it setting. Onl entironment context tBau 26 students 11 faculty and St Adolescents ; ”016d 1n ‘2ch 100d arailablt C0mpetititc 1‘, lunch, which . Manors, tea cOncerm UlldClStandl'ng research Under income Sch 00 ’5 making healthy food look and taste better, and increase the availability and accessibility of these items (Neumark-Sztainer, Story et al. 1999). In one study, adolescents focused on macrosystem influences by suggesting use of and using media, advertising, and product packaging to make healthy foods cool, or the “thing to do” (N eumark-Sztainer, Story et al. 1999). While these studies provide insight on the overall influences on adolescent dietary behaviors, it is important to explore the contextual influences specific to the school setting. Only one qualitative study was identified that explored the social and environmental factors that relate to healthy eating in adolescents specific to the school context (Bauer, Yang et al. 2004). In this study, seven focus groups were conducted with 26 students in two suburban public middle schools in Spring 2000. Additionally, 23 faculty and staff members participated in focus groups and individual interviews. Adolescents and adults identified a number of barriers to healthy eating similar to those noted in general studies of adolescent dietary behaviors. These included the types of food available in the cafeteria (described as greasy, high fat); availability of less healthy competitive foods in vending machines and snack carts; not having enough time to eat a lunch, which forces some students to choose competitive foods; and unhealthy dieting behaviors, teasing other students about weight and appearance, and other weight-related concerns. Additional qualitative studies exploring school nutrition will lead to a better understanding of the school-specific influences on adolescent dietary behaviors. The research undertaken in this dissertation examined school nutrition in a group of low- income schools, which may reveal barriers specific to this at-risk population. 37 THE Rl dietary school-l hate in Admini: commur can erect 01. lithe SCth] p impleme identitie. (Greenbt S“l‘i’tintt Banee 31 financial supmn‘ (Standard lasr [“0 l [raining (. famOIS [h THE ROLE OF ADMINISTRATORS IN NUTRITION PROMOTION Until this point, this review has focused on school factors that impact student dietary behaviors, but it is important that we take into account what influences these school-level factors. It is important to recognize the central role that school personnel have in shaping a school’s culture, and the degree to which health initiatives are valued. Administrators, school board members, teachers, food service personnel, parents, and community members can influence school nutrition promotion policies and programs and can create a health-focused school culture. Limited research to date has studied the role of these stakeholders in promoting health and nutrition to students. Several surveys have examined the perceived barriers to health initiatives of school personnel. In a survey of school administrators assessed the barriers to implementing CSHPs in Ohio (Greenberg, Cottrell et al. 2001). The top barriers identified were a lack of prioritization, funding, personnel, time, and leadership (Greenberg, Cottrell et a1. 2001). In a study of the top priorities and concerns of school superintendents, a lack of prioritization of health initiative was also evident (Winnail and Bartee 2002). Of the top ten concerns reported, half were directly or indirectly related to financial issues (funding, salaries, attraction and retention of quality teachers, teacher support, and declining student enrollment); three were related to academic priorities (standards and assessment, content improvement, and graduation requirements); and the last two included a lack of time, and provision of staff development and in-service training (Winnail and Bartee 2002). School board members in California listed similar factors that inhibit school nutrition policies including nutrition not being considered a 38 priority 1 Akinlohl lunch an IOU-ll. F funding 2 as primar identified and lack t stall t‘elt t messages nutrition c Se Stiltool nut members 2 (Shahid 3(. Hot transla ”Ulrition. 3 studies. a d] health or m 31633 lShak EVit are “ileum pnncipalS' r priority by school board and the impact of the food program on school budget (Brown, Akintobi et al. 2004). One study examined the barriers to implementing a quality school lunch and providing nutrition education in Massachusetts’s schools (Cho and Nadow 2004). Food service directors, administrators and other relevant staff identified lack of funding and time, academic requirements, and students’ preference for unhealthy foods as primary barriers (Cho and Nadow 2004). Food service directors and other staff identified lack of communication and leadership, lack of support materials and training, and lack of parental support as additional challenges (Cho and Nadow 2004). Also, other staff felt that the media focus on junk foods, and a lack of reinforcement of nutrition messages in the home and school (e.g. vending machines) were challenges to providing nutrition education (Cho and Nadow 2004). Several studies have explored the attitudes of school administrators regarding school nutrition issues. Two studies have noted that school administrators and board members are aware of the relationship between nutrition and academic performance (Shahid 2003; Brown, Akintobi et al. 2004). In one of these studies, these beliefs were not translated into school practices, as principals did not encourage teachers to promote nutrition, and often permitted competitive foods in schools (Shahid 2003). In both studies, administrators and board members expressed an interest in being involved in health or nutrition initiatives, and were interested in receiving additional training in these areas (Shahid 2003; Brown, Akintobi et al. 2004). Evidence suggests that administrator knowledge and attitudes towards nutrition are reflected in school practices. In a survey of school principals in Minnesota, principals' positive attitudes toward the school nutrition environment were positively 39 related it 2002 l. 1 that 31 lot get these relationsl comparet promote l 3003 ). C0flC€fnSt strategies Concerns ( School sys be more 61 edUL‘ation ; means ofp Thi loo-.inCOmI the Ihln‘ss t. This “Ill pr do [0 aSSist Wlll ldenn‘fi related to the total number of school nutrition practices reported (French, Story et al. 2002). In another study, motivating factors for nutritional decisions differed in schools that allowed vending machines (need for revenue, belief that kids will find other ways to get these types of foods, other schools have vending machines, and there is not a relationship between consumption of these foods and academic performance), when compared with those that do not allow vending machines (they are not necessary, do not promote learning, create trash in the school, and the district does not allow them) (Shahid 2003) It is important that health and nutrition practitioners recognize the priorities and concerns of school administrators and other personnel, and design health promotion strategies that do not exacerbate these issues. Programs that can enhance the top concerns of school administrators would have the best chance of being welcomed into school systems. Identifying administrators' primary areas of concern can help researchers be more effective when working with school districts. Additionally, efforts to provide education and training to school administrators and other personnel may be an effective means of prioritizing health initiatives in schools. This study examines not only the barriers to promoting health and nutrition in low-income middle schools, but also the accomplishments that schools have made, and the things that have helped schools to make improvements that promote student health. This will provide valuable information about things that practitioners and researchers can do to assist schools in prioritizing health initiatives. The focus on low-income schools will identify any additional barriers experienced by this population. 40 SCHUOl DlETAR G many res health an including em ironm nutrition 1 are not dii intert emit Hott‘et er, Cm lfttmm the multip making ch COmbined Th “Ologicai inten'entit Cm irOnmc (Commit). SCHOOLS AS AN INTERVENTION SETTING TO IMPROVE ADOLESCENT DIETARY INTAKE Given the many school-specific factors that influence adolescent dietary intake, many researchers and practitioners turn to schools as intervention sites for promoting health and nutrition. School nutrition interventions include a variety of approaches including nutrition education curriculum to increase knowledge of healthy eating, environmental changes in the foods and beverages available in schools, and adoption of nutrition policies. Nutrition education interventions are reviewed briefly below, as they are not directly pertinent to the research in this dissertation. School environment interventions, as a primary focus of the current research, are reviewed in more depth. However, due to the large number of studies that have been published involving environmental interventions, select studies are reviewed to provide an understanding of the multiplicity of the literature. Additionally, most "policy" interventions involve making changes to the school nutrition environment; therefore, these studies are combined with the environmental interventions for this review. School nutrition education interventions The literature on school nutrition education interventions supports the use of ecological models of health behavior change. A review of nutrition education interventions showed that behaviorally-focused interventions and those that included environmental changes to reinforce knowledge had more positive results, while programs that focused solely on providing information and teaching skills were less successful (Contento, Balch et al. 1995). School nutrition education programs have been shown to 41 be etTectii Moschant foods. 1K: al. 1996; t'egetablt program consumr (James. 7 improtei l990)_ T: A Vallett- lfl‘leucd represent; perfonmc “Search 1 Well as Sch Mai fOOdg and b Pricing and . l be effective at increasing knowledge of healthy eating (Davis, Clay et al. 1999; Manios, Moschandreas et al. 1999; Moreno, Denk et al. 2004) enhancing preferences for healthy foods (Kelder, Perry et al. 1995), decreasing intake of high fat foods (Luepker, Perry et al. 1996; Gortmaker, Cheung et al. 1999), and increasing consumption of fruits and vegetables (Gortmaker, Cheung et al. 1999; Perry, Bishop et al. 2004). An educational program aimed at decreasing consumption of carbonated beverages resulted in decreased consumption, as well as a reduction in the mean % of overweight in intervention children (James, Thomas et a1. 2004). One study showed maintenance of some dietary improvements three years after a nutrition education intervention (Nader, Stone et a1. 1 999). School nutrition environment interventions The results of school nutrition environment interventions have been inconsistent. A variety of environmental intervention approaches have been employed, and are reviewed below. While this review is not comprehensive in nature, it provides representative examples of the variety of school nutrition interventions that have been performed. It also shows the diversity of outcome measures used in school intervention research, including individual psychosocial, behavioral, and physiological parameters, as well as school-level environmental variables. Many interventions focus on increasing the availability and marketing of healthy foods and beverages. The Changing Individuals' Purchase of Snacks (CHIPS) explored pricing and promotion strategies to encourage purchasing of healthy options in 12 secondary schools (French, Jeffery et al. 2001); a similar study was conducted in two 42 high schc strategies reading 2001). I teatime nochan; 2003). l M-SPA‘. Achieti' atailahi' (Cassadj atailahlt Changes. Saturated dt‘t‘rease. l Cal‘Clt’ria lOIt‘er.fat in The nice 2004'). Stt lonenfar 0 zmderchaS Howe‘fr. th high schools (French, Story et al. 1997). In these studies, pricing and promotion strategies were successful at increasing sales of low-fat snacks, carrots and fruit from vending machines and a la carte lines (French, Story et al. 1997; French, Jeffery et al. 2001). In the Middle-School Physical Activity and Nutrition (M-SPAN) study, a two- year intervention to provide and market healthy food choices in all middle school venues, no changes were observed in total fat or saturated fat intake (Sallis, McKenzie et al. 2003). It is possible that addition of pricing strategies to promote healthier options in the M-SPAN study could have resulted in decreases in fat intake. The Students Today Achieving Results for Tomorrow (START) after-school program increased the availability of fruits and vegetables during the snack period in 44 after-school programs (Cassady, Vogt et a1. 2006). No pricing efforts were necessary, as these snacks were available to students at no cost. These changes resulted in positive nutrition environment changes, including increased availability of fresh fruit and fi'uit juice and decreased saturated fat content of snacks; however, negative changes were also observed including decreased availability of milk, calcium, and vitamin A (Cassady, Vogt et al. 2006). Multiple levels of outcome variables were assessed in the Trying Alternative Cafeteria Options in Schools (TACOS) study, which aimed to increase availability of lower-fat foods in a la carte (French, Story et al. 2004). Sales data showed an increased in the mean percentage of lower-fat food sales in the second year (French, Story et al. 2004). Students reported improved perceptions about the school environment providing lower-fat options, social support for choosing lower-fat foods, and ease of identification and purchase of lower-fat foods in the school cafeteria (French, Story et al. 2004). However, there were no significant differences in intentions to buy lower—fat foods from 43 the cal‘et appears sales ot‘ improt t comper to pron Street d from th SCllOol I consum Vitamin decrease 2006), , tending schOOlS ( WHISOn 5 Ti School lUr rePOIted 1( hot” dli‘tar Wile): regu Carbonated . the cafeteria or self-reported choices of lower-fat items (French, Story et al. 2004). It appears that while the TACOS study was successful at enhancing the availability and sales of lower-fat a la carte options and increased student awareness, that these improvements were not translated into improved dietary behaviors. It is possible that compensation occurred in other venues, such as at vending machines or at home. Interventions have also tried removing less-healthy options available to students to promote healthy eating. An intervention study that removed snack chips, candies, sweet desserts, and sweetened beverages from snack bars and removed vending machines from three middle school cafeterias in Texas resulted in mixed changes at the student and school levels (Cullen, Watson et al. 2006). Positive changes included decreased consumption of sweetened beverages and increased consumption of milk, calcium, and vitamin A. Negative consequences included increased intake of saturated fat and sodium, decreased intake of vegetables, and increased sales of ice cream (Cullen, Watson et al. 2006). Additionally, following the removal of less-healthy items from the snack bar and vending machines from the cafeterias, the overall number of vending machines in the schools doubled, and sales of chips and candy from vending machines increased (Cullen, Watson et al. 2006), indicating that compensation was occurring. The Go for Health project improved the fat, saturated fat, and sodium content of school lunches (Simons-Morton, Parcel et al. 1991). Students from intervention schools reported lower intakes of fat, saturated fat, and sodium (significance not reported) in 24- hour dietary recalls (Simons-Morton, Parcel et al. 1991). In another study, a district-wide policy regulating the types of foods and beverages allowed in schools (e.g. plain or carbonated water or 100% juice with no added sweeteners, 1% or fat-free milks) as well 44 as the nutrient content of individual foods and beverages (e.g. 30% or less calories from fat, 10% or less calories from saturated fat, no more than 35% sugar by weight) was adopted in San Francisco (Wojcicki and Heyman 2006). Results included healthier food and beverage options in school meals and a la carte/snack bars, increased participation in and revenue from the school meals program, and decreased a la carte and snack bar sales (Wojcicki and Heyman 2006). It appears that in order to be effective, school environmental changes must be applied to all food venues within a school, or students will continue to seek unhealthy items from alternative sources. Several studies have used multiple intervention methods to promote healthy eating and physical activity to students in schools. The Teens Eating for Energy and Nutrition at Schools (TEENS) study in 16 middle schools in Minneapolis included classroom education, family newsletters and behavioral coupons, and school-wide environmental changes to promote lower-fat food service and a la carte offerings and increased fi'uits and vegetables (Lytle, Murray et a1. 2004). Results showed no differences in the fruits, vegetables, and salads offered in school meals (Lytle, Kubik et al. 2006). Intervention schools increased the proportion of healthier foods available in a la carte (p = .04) (Lytle, Kubik et al. 2006). Student dietary intake measured by 24-hour recalls and by a fruit and vegetable screener survey revealed no significant differences for the intervention group (Lytle, Murray et al. 2004). The only significant difference was seen in the student survey-reported usual food choice score, indicating students in intervention schools made lower fat choices (Lytle, Murray et al. 2004). The Child and Adolescent Trial for Cardiovascular Health (CATCH) study, considered to be one of the best school intervention studies to date, provides an excellent 45 example 0f ll [eyelinter\€' gud8n13.3nt study shott e honey er thL‘i 19961. P8.H increased stu arhlnfinhu or sell-ellica limited behat uhmflshadc inuke.butal: ngmficantdi Lyueetal 19 pressure, and Resuh tigeneral.tht bil'mges at a behavioral ant at'ailability of 10 see Changes 0Plionsoecrea to highlight hea example of the mixed outcomes seen in the majority of school interventions. This multi- level intervention included food service changes, nutrition education curriculum for students, and take-home lessons and activities for families. Results from the CATCH study showed some improvements in the fat content of school breakfasts and lunches, however there were increases in sodium content of these meals (Osganian, Ebzery et al. 1996). Psychosocial correlates of diet showed mixed changes. The intervention increased student-reported dietary intentions, usual food choice, nutrition knowledge and social reinforcement; however there were no differences in positive or negative support or self-efficacy for dietary behaviors (Edmundson, Parcel et al. 1996). There were limited behavioral and physiological changes in students. Students in the intervention schools had decreased intake of fat and saturated fat as a percentage of total caloric intake, but also had an increase in sodium intake (Lytle, Stone et al. 1996). No significant differences between groups were seen in intake of fruits or vegetables (Perry, Lytle et al. 1998), or in cardiovascular disease risk factors including obesity, blood pressure, and serum lipids (Webber, Osganian et al. 1996). Results from individual school nutrition environmental interventions indicate that, in general, these programs can be successful in creating positive changes in the foods and beverages available to students; however, these changes do not always result in positive behavioral and physiological results. It may also be necessary not only to increase the availability of healthy options, but also to reduce the availability of less healthy options to see changes in adolescent dietary intake. Combining increased availability of healthy options, decreased availability of less healthy options, and utilization of pricing strategies to highlight healthy items may be necessary to have the maximum impact on adolescent 46 dietary behaviors. Interventions that include strategies at multiple levels, such as nutrition education and family outreach could potentially increase the likelihood that behavioral changes will occur, however results from the TEENS and CATCH studies still reported mixed results, with some negative changes in dietary behavior. One difficulty with multi-level interventions is determining the individual factors that create positive and negative changes in student behaviors. Additionally, interventions that require numerous changes at multiple levels may be overwhelming to individual schools, resulting in decreased commitment and buy-in to the project. Therefore, it is necessary to identify which specific strategies seem to be most effective and tailor these activities to the specific needs of the schools. LIMITATIONS OF THE SCHOOL NUTRITION INTERVENTION LITERATURE Some of the inconsistency in results of school nutrition interventions may be attributable to the methods used in the field. When reviewing the school nutrition intervention literature, many limitations become apparent, several of which are outlined below. Many of the school intervention studies have small sample sizes, both in number of school and students, which decreases statistical power to detect differences between intervention groups. Many studies use a case-study approach with only one or a few schools, oftentimes with no comparison group. The largest school intervention thus far is the CATCH study involving 96 elementary schools in four regions of the United States (Luepker, Perry et al. 1996). Additionally, the outcome measures of interest vary among published studies. Many school nutrition interventions do not measure student-level indicators, such as the 47 CHIPS study (French, Jeffery et al. 2001), which collected sales data only at the school level. Student-level indicators commonly measured in intervention studies include both dietary intake, as well as psychosocial variables. Studies that measure dietary intake have used a wide variety of instruments. Methodologies used in several key studies include: a single 24-hour recall in the TEENS study (Lytle, Murray et al. 2004); one 24- hour recall supplemented with a qualitative diet record in the CATCH study (Lytle, Stone et al. 1996); seven-day food records as used in the Gimme 5 study (Baranowski, Davis et al. 2000); or a food-fiequency type survey including a checklist of low and high fat foods with frequency response options (not much, some, a lot) in the Cardiovascular Health in Children (CHIC) study (Harrell, Gansky et al. 1998). Lack of consistent methods of assessing dietary intake makes it difficult to quantify changes in dietary intake as a result of interventions, and limits the ability to compare results between studies. The choice of methods used to assess student dietary changes may have a significant impact on the results of the study, and are important when interpreting the results of school intervention studies. Results may vary when researchers measure total dietary intake (such as 24-hour recall) compared with a limited scope (such as plate waste or visual observation of a single meal). For example, the “Go for Health” school nutrition intervention uses two methods to assess dietary intake (Parcel, Simons-Morton et al. 1989). Direct observation of school lunch showed improvements in fat and sodium intake for the intervention group; however no significant differences were seen when analyzing 24-hour recall data (Parcel, Simons-Morton et al. 1989). This may reflect differences in measurement precision, or could also be an indication that students are compensating for changes in foods eaten in schools with foods eaten outside of the school 48 and at l measur beyerag machinu foods it only fot fruits an ”"0 fat. tt‘ validity 3000; Ly quanti lie Standard literature Uttderstar PROJEC' Tl demonstr; trend (0g: anOme an dlflary Qu; 3. $04; 1380 and at home. When interpreting study results it is necessary to take into consideration the measurement tools used. Measuring the school nutrition environment, including the types of foods and beverages available in school meals, as well as competitive foods such as vending machines and a la carte has also proven to be difficult. Some researchers categorize foods into healthy or less healthy options (using varying criteria to categorize foods), or only focus on a specific type of food or beverage (e.g., chips, carbonated beverages, or fruits and vegetables). Others report nutrient composition of foods and beverages (e.g., % fat, total calories). There are currently no instruments that have been tested for validity or reliability to assess the school nutrition environment (McGraw, Sellers et al. 2000; Lytle and F ulkerson 2002). Differences in the way environmental variables are quantified makes interpretation and comparison of results between studies difficult. Standardizing measurement and reporting methods used at all levels in the intervention literature, and measuring outcome variables at multiple levels would enhance understanding of the factors that influence adolescent dietary behaviors. PROJECT JUSTIFICATION The rapid increase in the prevalence of overweight and obese adolescents demonstrates the need for innovative and effective intervention strategies to reverse this trend (Ogden, Flegal et al. 2002; Ogden, Carroll et al. 2006). Evidence suggests that low- income and minority adolescents are at increased risk for overweight, obesity, and poor dietary quality (Neumark-Sztainer, Story et al. 1996; Fox and Cole 2004; Sorof, Lai et al. 2004; Jago, Harrell et al. 2006; Ogden, Carroll et al. 2006). 49 Schools have been identified by researchers and public policy makers as a logical setting to work with children because: a) they reach the large majority of children; b) students spend a large amount of their time at schools; and c) students typically consume at least one meal in the school setting each day (usually lunch, and sometimes breakfast and after-school snacks as well) (Story 1999; Michigan Department of Education, Michigan Department of Community Health et al. 2001; American Dietetic Association 2006; Story, Kaphingst et al. 2006). While school nutrition interventions involving education, environment, and policy changes (or a combination thereof) have been moderately successful at improving intrapersonal factors (such as knowledge, attitudes, and self-efficacy) and environmental factors (such as the nutrient contents of school meals), there have been limited and mixed changes in behavioral outcomes (such as dietary intake) or physiological disease markers (such as blood lipids and BMI) (Webber, Osganian et al. 1996; Atkinson and Nitzke 2001). These results clearly indicate a gap between the theory and practice of influencing dietary behaviors in adolescents. Lack of positive individual level results supports the need for multi-level interventions that target multiple factors that influence adolescent dietary intake. The lack of consistency and effectiveness of school nutrition interventions can partially be attributable to the many limitations of the literature mentioned above including: lack of valid and reliable assessment methods; variation in the measurement of outcome variables; inconsistency in reporting of results; use of small non-representative samples; and short-term follow-up periods. However, the most important problem could be the lack of clear theoretical basis in many intervention programs. Social ecological 50 models of health behaviors have been pr0posed for addressing adolescent dietary behaviors (Story, Neumark-Sztainer et al. 2002), however the efficacy of these models has not been proven (Sallis and Owen 2002). It is imperative that researchers strive to better understand the relationships between the various school-related influences on adolescent dietary behavior in order to design and evaluate more effective school nutrition interventions. There is very limited literature regarding the efficacy of school policies at creating change in school environment and practices, and ultimately resulting in changes in student dietary behaviors. The Child Nutrition Reauthorization Act of 2004 (Section 204 of Public Law 108-265 June 3, 2004) required all schools participating in the National School Lunch Program to establish a school wellness policy by fall of 2006. Given the novelty of the federal wellness policy mandate, now is the time to explore the effect that wellness policies have had on school nutrition and physical activity environments, policies, and practices. The goal of this dissertation research was to examine the associations between school nutrition policies, physical environments, and dietary behaviors to better understand what intervention strategies will be most effective at influencing adolescents’ dietary behaviors. Baseline data from an intervention study in low-income Michigan middle schools were used to examine these associations in a cross-sectional manner. The focus on low-income schools provides valuable information specific to this vulnerable population that is historically at a higher risk of obesity and nutritional deficiencies. This research helps to fill two major gaps in the school nutrition literature. First, it is the second known study to explore whether school wellness policies are associated 51 with school studies to ex dietary intai 31.3009; Fo shich utilize tool for expl. cross-sectior causational. be ideal to e Thus. result determine c identify- in adolescent \Vl deSlé’nl. sr mUlti-leye C0llectcd SChOOl art Students t number ( an UneqL agsumes with school nutrition environments and practices. Second, this research is among the first studies to examine the association between school nutrition environments and student dietary intake in a cross-sectional manner (Kubik, Lytle et al. 2003; Briefel, Crepinsek et al. 2009; Fox, Dodd et al. 2009; Gleason and Dodd 2009). Cross-sectional analysis, which utilizes observations made at a single point in time without intervention, is a useful tool for exploratory analysis in order to assess associations. However, due to the fact that cross-sectional data is collected at one point in time, it cannot be interpreted as causational. Longitudinal studies of a nationally representative group of schools would be ideal to establish and to observe the changes in environments and policies over time. Thus, results are best interpreted as providing direction for further intervention studies to determine causational relationships. Results from this dissertation research will help identify which school-level intervention activities are most likely to have an impact on adolescent dietary behaviors. When analyzing data involving students nested within schools (a cluster sample design), several methodological issues arise that must be addressed through appropriate multi-level statistical modeling techniques (which take into account the fact that data are collected at two different levels — students and schools). First, the students within one school are not independent of other students within that school; thus, the error terms from students within the same school will be correlated with each other. Furthermore, the number of students sampled within each school can impact the results of the analysis if an unequal number of students are sampled from schools but the statistical method assumes an equal number of students in each school. Results must be weighted so that 52 schools with more students responding are weighted more heavily in the analysis that schools with only a few students. The SNDA-III study data are cross-sectional in nature and were analyzed. for similar associations as examined in this dissertation (Briefel, Crepinsek et al. 2009; Fox, Dodd et al. 2009; Gleason and Dodd 2009). While the SNDA-III study contained a large number of schools (287), dietary data were only gathered from approximately 10 students in each school. Because of the large number of schools, multi-level modeling techniques were not used to accommodate the data structure of students nested within schools; however, the single-equation model accounted for the fact that the error terms from students within a single school would be correlated. Additionally, it was not necessary to weight observations from different schools, as there were approximately the same number of students sampled from each school. In contrast, in this dissertation research analysis, hierarchical linear modeling (HLM) software is used to fit the multi-level data structure of students nested within schools because of the smaller number of schools (65) and the variation in the number of students from each school sampled from 10 to 68 students (schools with less than 10 students participating were removed from analysis due to a lack of statistical power). The HLM program weights the results from each school based on the number of students sampled. The qualitative component of this dissertation utilized interviews and focus groups with school administrators, food service directors, coordinated school health team members, and students to assess the facilitators and barriers schools experience when promoting nutrition. Few studies have explored the influence of these key adults in promoting school health (Greenberg, Cottrell et al. 2001; Winnail and Bartee 2002; 53 Shahid 2003; Brown, Akintobi et al. 2004; Cho and Nadow 2004), with only one focusing on nutrition specifically (Shahid 2003). Similarly, while several studies have qualitatively explored how adolescents perceive healthy eating (Story and Resnick 1986; Chapman and Maclean 1993; Neumark-Sztainer, Story et al. 1999; Cullen, Baranowski et al. 2000; Croll, Neumark-Sztainer et al. 2001; O'Dea 2003; Bauer, Yang et al. 2004), only one has focused on the school setting (Bauer, Yang et al. 2004). Both the student and adult perspectives are necessary to understand the full context of nutrition promotion in schools, and to aid researchers in developing interventions and programs that recognize and address the challenges schools experience. Furthermore, it is important to explore barriers in low-income schools with both students and adults to determine unique challenges they may experience in promoting school nutrition that may not be apparent in the other populations that have been studied. Taken together, the results from this dissertation help define and justify use of social ecological theories in the school setting. Researchers and policymakers can use this information to design programs, policies, and resources to enhance the dietary behaviors of adolescents, especially those in low-income schools. The next three chapters are written as articles for publication. Chapter Two examines the associations between local wellness policies and healthy school environments and practices. Chapter Three examines the effect of school nutrition environments and practices on student dietary intake. Chapter Four uses a qualitative approach to examine the facilitators and barriers to healthy eating in the school setting from the perspective of food service directors, school administrators, coordinated school health team members, and from students themselves. The concluding chapter brings the 54 three articles together, further discusses the results and overall conclusions that can be drawn from this research, as well as the implications for school practice and future research steps. 55 CHAPTER TWO: THE QUALITY OF SCHOOL WELLNESS POLICIES AND ASSOCIATION WITH SCHOOL PRACTICES IN LOW-INCOME MICHIGAN MIDDLE SCHOOLS BACKGROUND Schools have been identified as an important setting for nutrition promotion and interventions to reduce the prevalence of childhood obesity, promote overall health and well-being, and prevent chronic diseases such as heart disease and diabetes (Michigan Department of Education, Michigan Department of Community Health et al. 2001; American Dietetic Association 2006; Story, Kaphingst et al. 2006). The Child Nutrition and WIC Reauthorization Act of 2004 (Section 204 of Public Law 108-265 June 3, 2004) required all local education agencies (school districts) receiving funding for school meals (National School Lunch program, School Breakfast Program, or Afier School Snack Program) to establish a local wellness policy by July I", 2006. These local wellness policies serve as a written document that outlines the actions schools will take to promote health to students and staff. Wellness policies must include: goals for nutrition education, physical education, and physical activity; nutrition guidelines for school meals that meet or exceed USDA requirements; nutrition guidelines for all other foods available on campus (i.e., competitive foods); 3 plan for measuring implementation of the wellness policy; and involvement of key stakeholders in development of the policy including Parents, students, school food authority, administration, school board, and the public (Section 204 of Public Law 108-265 June 3, 2004). Recognizing that each school district has unique strengths and challenges, no specific details were given for each of the six 56 areas, allowing districts to tailor their policy to their needs. This lack of structure has resulted in a high degree of variability in the quality, implementation, and enforcement of such policies (Institute of Medicine 2007; Story, Nanney et al. 2009). Several nation-wide studies examined the existence of school nutrition policies prior to the federal mandate, at which time less than half of all school districts studied had adopted a wellness policy or other policies to promote healthy eating and physical activity (Greves and Rivara 2006; O'Toole, Anderson et al. 2007; Finkelstein, Hill et al. 2008). After the federal mandate took effect, the majority of districts had adopted a policy (School Nutrition Association 2006; Metos and Nanney 2007; Moag-Stahlberg, Howley et al. 2008; Probart, McDonnell et al. 2008; Belansky, Cutforth et al. 2009; Longley and Sneed 2009); however, wellness policy language is often vague, making it difficult to implement and evaluate wellness policy effectiveness (Metos and Nanney 2007; Probart, McDonnell et al. 2008; Belansky, Cutforth et al. 2009). School wellness policies may hold promise for improving school nutrition; however, few studies have examined the association between written policy and actual school health practices; those that have show mixed results. A Connecticut report indicated that wellness policy strength was associated with fewer unhealthy competitive foods available (Friedman 2009). Another study found significant improvements in wellness practices following the wellness policy mandate (Longley and Sneed 2009). In contrast, a Colorado study found little change in physical activity provisions or school nutrition environments after the wellness policy mandate (Belansky, Cutforth et al. 2009; Belansky, Cutforth et al. 2009). Another study found little concordance between written 57 policies and school fundraising practices (Kubik, Lytle et al. 2009). It is unclear whether or not wellness policies are being translated into the intended healthier school practices. The current study describes the association between written wellness policies and school-reported nutrition policies and practices to better understand the impact of the federal wellness policy mandate. The goal of the current study to examine the relationships among written wellness policies and school nutrition policies and practices as reported by school administrators and food service directors. This study is among the first to evaluate the associations between written wellness policies and parallel school practices. METHODS School Nutrition Advances Kids project The current study utilized baseline data collected as part of the School Nutrition Advances Kids (SNAK) project. The SNAK project, fianded by the Robert Wood Johnson Foundation Healthy Eating Research program, Supplemental Nutrition Assistance Program Education (SNAP-ed), and the Michigan Department of Community Health, is a collaboration between researchers at Michigan State University (MSU), the Michigan Departments of Education and Community Health, and several'partnering organizations of the Michigan Action for Healthy Kids coalition. The SNAK project aims to improve school nutrition environments through Coordinated School Health, Michigan’s Healthy School Action Tool (HSAT), and implementation of the Michigan 58 State Board of Education nutrition policy. All study procedures and instruments were approved by the MSU Institutional Review Board. Study sample Schools were recruited to participate in the SNAK project through an application for small grant funding to collect data and implement a nutrition environment and policy intervention or act as a comparison school. Eligibility criteria included having 50% or more of students eligible for free or reduced price meals, and having 7th and 8th grades within the same building (for follow-up purposes). School recruitment methods included direct mailings, e-mails, and phone calls to eligible schools and a posting on the Michigan Team Nutrition website. The SNAK project is a two-year intervention study with an overlapping design including two cohorts. The first cohort included 32 schools in 30 districts participating from October 2007 — June 2009, and the second cohort included 33 schools in 20 districts participating from September 2008 — June 2010. Wellness policv evaluation Local wellness policies were collected for 48 of the 50 school districts participating in the SNAK project (one district had not created a policy, and one was not able to locate their policy). For districts with more than one school building participating in the SNAK project, one building was randomly selected to represent the district in the wellness policy analysis. 59 The quality of the local wellness policies from each district was quantified using the School Wellness Policy Evaluation Tool (Schwartz, Lund et al. 2009). This tool contains 96 items within seven sections that correspond with the federal wellness policy requirements (Table 2-2). Each item received zero points if the item is not addressed in the written policy; one point if the item is addressed, but the statement is weak or only suggestive (e.g., schools should provide an adequate amount of time for lunch); and two points if the statement is specific and required (e.g., schools will provide at least 20 minutes daily to eat lunch). Response options were condensed into "No" (0 points) and "Yes" (1 or 2 points) categories. Each section, and the assessment as a whole, received two scores: 1) the comprehensiveness score represents the percent of items within the section addressed at all in the written policy (those receiving one or two points); and 2) the strength score represents the percent of items within the section that had strong and required statements (those receiving two points). Evaluation of the tool indicates it has adequate internal consistency and inter-rater reliability (Schwartz, Lund et al. 2009). Two researchers independently scored each policy using the Wellness Policy Evaluation Tool, and discrepancies between coders were discussed and reconciled with a third researcher. A comprehensive set of decision rules was created based on these decisions. Each policy was then rescored by both researchers, and results were compared once again to ensure consistency of scoring. It became apparent during the wellness policy scoring that the majority of policies were based on two common wellness policy templates that were made available for districts to use and modify to fit their needs, and that wellness policy quality differed based upon the template used. The first template was the Michigan Association of 60 School Boards (MASB) recommended policy developed by the Michigan Department of Education in collaboration with other state and local organizations, agencies, and citizens. The second was from a company that schools can hire to provide template school board policies that ensure schools are in compliance with all federal and state mandates (designated "Policy Company"). In addition, two schools utilized a template from the National Alliance for Nutrition and Activity (NANA) (www.5choolwelInesspoliciesorg), and four schools did not follow a recognizable policy template. Policies were therefore categorized based on the template type used to create them (MASB, Policy Company, NANA, Other). The MASB policies were also further categorized based on how districts modified the policy - shortened, lefi as intended, or enhanced the template policy. School Environment and Poligy Survey The School Environment and Policy Survey (SEPS) was used to gather data regarding the nutrition and physical activity policies, practices, and environmental features as reported by school personnel. The SEPS was developed by Dr. Elaine Belansky and the Rocky Mountain Prevention Research Center for use in Colorado elementary schools, and preliminary validation findings suggest little reporting bias (Belansky, Cutforth et al. 2009). The SEPS contains 3 modules with unique questions for administrators, food service directors (F SDs), and physical education (PE) teachers, based on their areas of expertise. Each module takes approximately 30 minutes to complete. The SEPS was adapted for use in Michigan middle schools based on a literature TeVleW, best practice recommendations for schools, and experience in working with 61 middle schools. The adapted SEPS was reviewed by project team members from various areas of expertise relating to school nutrition and by several school food service and administration representatives to establish face and content validity. The first cohort of schools completed the SEPS between January and March 2008, and the second cohort between November 2008 and March 2009. Individuals were mailed a paper survey and also e-mailed with a link to the survey online. Follow-up phone calls, e-mails, and mailings encouraged survey completion, and a $25 gift card was used as an incentive. Response rates were 85% for administrators, 91% for F SDs, and 86% for PE teachers. Table 2-1 describes the school nutrition policy and practice variables from the SEPS survey as reported by administrators and FSDs. School characteristics Information regarding the following school characteristics was gathered in several ways. Schools were asked to indicate the number of 7th grade students and the total building enrollment on their application to participate in the SNAK project. The percent of students eligible for free or reduced price school meals at each school was obtained through the Michigan Department of Education. School setting (urban, rural, or suburban) was determined using US. Census data (2000) for each community. Presence of a coordinated school health team (CSHT) prior to joining the SNAK project, and public vs. charter were determined through interactions with each school. The Healthy School Action Tool (HSAT) website and information from the Michigan Department of Community Health were used to determine whether schools had completed the HSAT self-assessment prior to enrollment in the SNAK project. Questions from the SEPS 62 survey indicated the percentage of minority students, whether the food service director had a nutrition-related degree, and whether the school had participated in any extra nutrition or physical activity programming. ANALYSIS All analyses were performed using Stata statistical software (Stata Corporation, Release 10.0, College Station, Tex, 2008). Descriptive statistics were used to express the quality of written local wellness policies as assessed by the Wellness Policy Evaluation Tool (represented by the mean comprehensiveness and strength scores for each section, and the total policy score) and the number of districts meeting all federal wellness policy requirements. Analysis of variance was used to examine differences in wellness policy quality based on the policy template type and school characteristics, and estimates of proportions were used to determine associations between school characteristics and wellness policy template type used. Multivariate regression analysis was used to determine associations between school characteristics and wellness policy quality scores, while controlling for wellness policy template type. Fisher's exact test (one-sided) was used to evaluate degree of agreement between written wellness policies and the school nutrition policies and practices reported in the SEPS. 63 TIA fifl\._t.._., _ Table 2-1: Description of variables from the School Environment and Policy Survey reported by administrators and food service directors (FSD) Variable LDescription Administrator-reported school nutrition policies Prohibits use of food as a No (no written policy, written policy not reward or punishment enforced)/Yes (unwritten policy always enforced, Healthy foods in vending written policy sometimes or always enforced) machines Healthy foods in a la carte Healthy foods in fundraising Healthy foods in class parties FSD-reported food service practices Breakfast available Yes/No Serving low-fat options in Yes (everyday, 1-2 times/week, or 3-4 school meals times/week)/No @ever) Strategies to encourage Yes (any of the following: offering mini-servings of participation in school meals new healthy foods, taste tests, incentives for school lunch, announcing menu, discarding damaged produce, displaying foods in a way that is visually appealing, surveying students about foods, beverages, preferences, time to eat, or general opinion of food service)/No (none of the above) Adequate time to eat lunch Yes (an average of 15 minutes or more to eat after being servedjNo (less than 15 minutes) Provide training for food Yes (school provided training/education service opportunities to food service staff)/ No FSD Degree Yes (FSD has a nutrition-related degree)fNo Administrator-reported school nutrition practices Coordinates nutrition Yes (any of the following: posters encouraging education with the entire healthy eating can be found throughout the school, school bulletin boards feature healthy eating information, school announcements include messages about healthy eating)/No (none of the above) Teachers model healthy Yes (teachers model healthy eating behaviors to eating behaviors students)fNo Integrate nutrition education Yes (the classroom curriculum integrates healthy into classroom curriculum eating messages)/No Presence of a coordinated Yes (existence of coordinated school health team school health team prior to start of SNAK project)/No 64 RESULTS Schools enrolled in the SNAK project had an average of 134 seventh grade students (range: 23-431) and an average building enrollment of 490 students (range: 13 l- 1217). Two-thirds of schools were located in urban settings, 20% were rural, and 14% were suburban. The majority of schools were public (85%), and 57% had >50% minority population. Twenty-two percent of SNAK schools used the same food service management company. Schools had an average of 71% of students eligible for free or reduced-price school meals (range: 50-100%). Demographic characteristics were similar for school buildings selected for wellness policy analysis compared with the total sample (data not shown). The results for each item scored in the Wellness Policy Evaluation Tool are shown in full in Appendix A. The comprehensiveness (percentage of items receiving one or two points) and strength (percentage of items receiving two points) of wellness policies for each section, and the overall policy are shown in 2- 2. The total comprehensiveness score indicates that local wellness policies on average addressed 40% of the all items, and the total strength score indicates only 18% of items had specific and required strategies. Wellness policies scored highest in the nutrition education section (mean comprehensiveness: 62%; mean strength: 3 1%), and lowest in the nutrition standards for competitive food section (mean comprehensiveness: 33%; mean strength: 5%). Local wellness policies were categorized based on the policy template type used to create the policy. Of the 48 policies examined, most districts used either the MASB (n 65 = 21) or the Policy Company (n = 21) policy template, two schools used the NANA policy template, and four others had policies that did not clearly resemble any model policy template. The MASB policies were further categorized based on how districts modified the template policy. Of the 21 districts using the MASB template, 2 enhanced the policy by adding additional requirements, 11 adopted the template with minimal changes, and 8 adopted a shortened version that either included only the introductory pages of the template, or removed other sections of the policy. The wellness policy quality scores by section and for the overall policy are shown in full by policy type in Appendix B. In general, the NANA-based policies had higher than average scores in most sections, and the policies that did not use a recognizable template were shorter and scored lower than average in most sections (Appendix B). For further analysis, the policies were grouped together into three categories: MASB, Policy Company, and NANA + Other. Analysis of variance was used to compare mean strength and comprehensiveness scores by policy template type. MASB- based policies scored significantly higher than Policy Company-based policies in nutrition education comprehensiveness, competitive food standards comprehensiveness, physical education comprehensiveness and strength, communication and promotion comprehensiveness and strength (Table 2-2). The MASB-based policies mean strength scores for competitive food standards were higher than Policy Company-based policies, which all received zero points in this section. MASB-based policies also scored Significantly higher in the total comprehensiveness and total strength scores. NANA + Other policies scored significantly lower than MASB-based policies in physical 66 education comprehensiveness scores, but there were no differences in any other sections or for the total assessment. There was a high level of variation in wellness policy quality within districts using the same policy template type, based on how districts modified the template. Table 2-3 shows that both the enhanced MASB policies and as-intended MASB policies scored significantly higher in total comprehensiveness and total strength when compared to the shortened MASB policies. Less than half of the local wellness policies (46%) met all of the federal requirements (Table 2-4). Most wellness policies met the minimum requirements for nutrition education (96%), school meal standards (93%), and physical activity goals (67%), with the majority of policies having strong and required statements, receiving the maximum of two points. While most district wellness policies addressed competitive food standards (77%) and policy evaluation (81%), the large majority of policies received only 1 point for having statements that were vague or suggestive. Associations between school characteristics and policy template were determined by comparison of proportions (Appendix C). Having a high percent of students eligible for free or reduced-price school meals, completing the HSAT assessment, and having a CSHT were not found to be associated with schools selecting any particular policy template type. Schools implementing extra nutrition or physical activity programs were Significantly more likely to use the Policy Company policy template. Small, rural, and public schools, those with a high percentage of minority students, and those using the food service management company were significantly less likely to use other wellness p01icy templates. Analysis of variance and multivariate regression analysis were used to 67 examine the associations between school characteristics and wellness policy quality (total comprehensiveness and strength). No school characteristics were found to be significantly associated with wellness policy quality independently or after controlling for policy template type (Appendix D). Fisher’s exact test (one-sided) was used to explore the degree of concordance between written local wellness policies and: administrator-reported school nutrition policies (Table 2-5); administrator-reported school nutrition practices (Table 2-6); and FSD-reported school food service practices (Table 2-7). A concordant pair is when a practice has been reported in the SEPS survey, and that item is included in the written wellness policy; a discordant pair indicates that the SEPS-reported practice is not included in the wellness policy, or vice-versa). The overall concordance is the percentage of responses that were similarly classified in written wellness policies and the SEPS survey (Yes/Y es or No/No pairs). The only practice that showed significant concordance was having a policy regarding healthy foods in fundraising activities (71% similarly classified, p = 0.01) (Table 2-5). The percent of concordant policies and practices ranged from 9-71%. 68 25 + _m:o;o.anU 2.8 .m _ 4.9 3.9. 2.2 .229 8.2 2.9. 84.9 8.2 2 m. :. .229 8.2 i 282 328 2.9. .8. 3 2.2 2.2 .229 2.2 3.2 .869 2.2 8.2 .229 9.2 £9.28 - .5888 8:82 a 2 20533229800 E2... .29 2.2 2.8 .229 2.2 Go. 2 .889 2.2 8.8 .229 8.8 - 8882. 8282 cu 229 :82 no 229 :82 no 229 522 co 229 :82 880m 20:8 mesa? A c :2 n 5 . . . . , . . w H = oomozom , H a . l worm—on $50 +.5,000 kcal/day, n = 94) were removed fiom analysis. Next, surveys with the most extreme values for several error-checking variables (e. g., the number of solid foods consumed per day, the percentage of foods eaten every day or never in the previous week, the percentage of foods with the same portion size) were again visually examined for "playful" patterns; however, no clearly unreliable patterns were discovered, and none of these surveys were removed. Schools with fewer than 10 complete surveys were removed from analysis due to a lack of statistical power (1 1 schools, n = 62 students). Schools with no baseline data other than student surveys were also removed from analysis (2 schools, n= 21 students). In preliminary analyses, race was found to be significantly associated with many dietary intake outcome variables, thus surveys with missing race were removed (n = 50). The final sample for this dissertation consisted of 51 schools with 1544 students (the number of schools and students varies for each individual analysis due to missing school-level data). The mean response rate of surveys analyzed for all SNAK schools was 24% (range: 0% to 66%). 87 School nutrition environment Data collection forms were created to gather information about the foods and beverages available in school meals, a la carte, and vending machines and provided to the Food Service Director (F SD) twice a year as a packet. Baseline data collection occurred for a one-week time period during November/December 2007 for the first cohort and during November/December 2008 for the second cohort. During the data collection period, the FSD (or other food service personnel) at each school reported all foods and beverages available to students in meals (breakfast and lunch) and a la carte every day. F SDs were asked to write down all items available in vending machines on one day during the data collection period, as the items available in vending machines were not expected to change as frequently. As an incentive for data collection, school food service programs received $325 for each packet returned. Information gathered in these forms was used to identify availability of competitive foods in these schools. Availability of competitive foods was examined in several ways. First, schools were divided into 4 categories based on availability of both vending and a la carte. The four groups included no competitive foods available, only a la carte available, only vending available, and both a la carte and vending available. Next, each competitive food venue was characterized individually. A la carte available (yes/no) and vending available (yes/no) variables were created. Lastly, vending machines were further categorized into by the type of items available. Groups included no vending machines; healthy beverage only vending machines that contained only water, 100% fruit juice, and very low calorie sports drinks; mixed beverage only vending machines which contained a mix of healthy and less healthy beverages but no food items; 88 and mixed vending machines which contained a combination of healthy and less healthy beverages and food items. The availability of fruits and vegetables in school lunches was calculated from the food service data collection forms reporting all foods available in school lunch for a one week time period. The mean number of fresh fruits, vegetables, and entree salads available per day were calculated. Availability of a salad bar (yes/no) and a fruit bar (yes/no) was assessed for each school. School cmracteristics Information regarding school characteristics was gathered using several methods. The percent of students eligible for free or reduced price school meals at each school was obtained through the Michigan Department of Education. School setting (urban, rural, or suburban) was determined using 2000 US. Census data for each community. Presence of a coordinated school health team (CSHT) prior to joining the SNAK project, and public vs. charter were determined through interactions with each school. The Healthy School Action Tools (HSAT) website and information fiom the Michigan Department of Community Health were used to determine whether schools had completed the HSAT assessment prior to enrollment in the SNAK project. Eleven schools were from the same district, and a “district” variable was created to represent this grouping. Type of food service program was determined through interactions with each school and was characterized as traditional (with a full service kitchen, or a satellite kitchen in the district where foods were prepared on site) or other (a heat and serve kitchen without full cooking capacity or a vendor-based operation where fully cooked foods were delivered to 89 the school). Similarly, 12 schools used the same food service management company, and were grouped into a “management company” variable. ANALYSIS Stata statistical software (Stata Corporation, Release 10.0, College Station, Tex, 2008) was used for descriptive analysis of school-level variables. Due to the hierarchical nature of the data (students within schools), hierarchical linear modeling was used to examine the associations between school-level variables and student dietary intake. Linear regression analyses using restricted maximum likelihood ratio were performed using HLM version 6.08 software (Scientific Software International 2009). The dietary intake variables of interest included energy intake, percentage of energy intake from total fat, percentage of energy intake from saturated fat, servings of fi'uits, servings of vegetables, servings of fruits + vegetables, and fiber intake. Variables other than energy intake were energy-adjusted (intake/1,000 kcal/day) to account for potential under- and over-reporting (Willett 1998). In order to reduce skewness and enhance normality of distribution, variables other than total and saturated fat were log-transformed, and results are reported based on geometric means rather than absolute means. For descriptive analysis of dietary intake variables by race, models were constructed using individual nutrients as the outcome variable, and entering race categories (white as reference group, African American, Hispanic/Latino, and other) into the model for females and males separately. Next, individual models were created for each racial category to examine gender differences in dietary intake. To determine the association between school nutrition environmental features and student dietary intake of 90 individual nutrients, several models were used. In all models, student characteristics (gender and race) and school characteristics (presence of CSHT, completion of HSAT, setting, district, food service management company, % students eligible for free/reduced- price meals, other foodservice, and public vs. charter) were included in the regression model as covariates. Total energy intake was included as a student-level covariate in all models (except for those with energy intake as the outcome variable) to adjust for potential under- and over-reporting of dietary intake. Student race and gender were allowed to have random error terms when the variance was found to be significant in the full model (all student and school-level covariates) for each dietary intake variable. The following random effects were discovered: Hispanic/Latino had random effects for % energy from fat and vegetable intake; sex had random effects for fi'uit intake, and African American had random effects for mu + vegetable intake. Additional school-level covariates were included for specific nutrient outcome variables. With fruit intake as the outcome variable, the mean number of fruits available per day in school lunch and availability of a fruit bar were included as covariates. With vegetable intake as the outcome variable, the mean number of vegetables available per day in school lunch, mean number of entree salads available in lunch, and availability of a salad bar were included as covariates. With fruit + vegetables and fiber intake as outcome variables, availability of a salad bar and mean number of fruits, vegetables, and entree salads available per day were included as covariates. 91 RESULTS Schools enrolled in the SNAK project had an average of 134 7th grade students (range: 23-431) and an average building enrollment of 490 students (range: 131-1217). Two-thirds of schools were located in urban settings, 20% were rural, and 14% were suburban according to 2000 US. census data. The majority of schools were public (85%), and 57% of schools had >50% minority population. Twenty-two percent of SNAK schools used the same food service management company. Schools had an average of 71% of students eligible for free or reduced-price school meals (range: 50- 100%). Mean dietary intake was examined by gender and race (Table 3-1). All racial groups (except males in the other racial group) had a significantly higher intake of calories and lower percentage of energy intake fi'om saturated fat compared with white students in both males and females. Hispanic/Latino males had a lower percentage of energy intake from total fat compared with white males, and Hispanic/Latino females had a lower percentage of energy intake from total fat than all other racial groups. Among males, all other racial groups had a higher intake of fi'uits compared with white males; Hispanic/Latino students had a lower vegetable intake than white males; the African American and other racial groups had a higher intake of fi'uits + vegetables combined than white males; and the Hispanic/Latino and other males had a higher fiber intake than white males. Among females, the other racial group had a lower fi'uit intake than white females; the Hispanic/Latino females had a lower vegetable intake and fruit + vegetable intake compared with white females; and the Hispanic/Latino females had a higher fiber intake compared with white and other females. Hispanic/Latino females also had a lower 92 vegetable intake than African American females, while Hispanic/Latino males had a lower vegetable intake than males in the other racial group. When comparing gender differences in dietary intake, white females had significantly lower energy intake and percentage of energy intake from saturated fat, and a higher fruit, vegetable, fruit + vegetables, and fiber intake compared with white males. African American females had a significantly higher vegetable intake than African American males. Table 3-2 shows the associations between availability of competitive foods and student dietary intake. In the first analysis, availability of both vending and a la carte in a school was associated with a higher percentage of energy intake from saturated fat (0.43% of energy intake; p = 0.032), while availability of only a la carte or only vending were significantly associated with an increase in fruit intake (0.08 servings/1,000kcal/day; p = 0.042 and 0.15 servings/1,000kcal/day; p = 0.011, respectively) when compared with schools that have no competitive foods available. When examined individually, availability of vending and availability of a la carte were not significantly associated with student dietary intake. When examining the types of vending machines available in schools, many interesting associations were seen. Availability of vending machines that contained only healthy beverages (e.g., water, very low-calorie sports drinks) was associated with a significant decrease in energy intake (p = 0.009), and availability of vending machines with mixed beverages but no foods showed a trend for decreased energy intake (p = 0.063) compared with schools that did not have vending machines. Furthermore, having vending machines with mixed foods and beverages was associated with a significantly 93 higher energy intake than schools with only healthy beverages in their vending machines (p<0.05). Availability of mixed beverage and mixed food and beverage vending machines was associated with higher percentage of energy intake from fat (p = 0.032 and p = 0.040, respectively), and mixed beverage vending was associated with higher percentage of energy intake from saturated fat (p = 0.032) compared with schools with no vending. Lastly, availability of healthy beverage only vending machines was associated with lower vegetable, and mu + vegetable intake (p = 0.019 and p = 0.049, respectively). 94 33 25 was :3 one 2.: need 5 me: Q sec 83 25 8.2 83 who 3.8 83 a mm: em 8:3 $585 83 w; E: 23 N2 2.: :55 m2 :8 3 55:2 §o£< 8.: 3% $2 mom as: 823 as: m... 83 26 NE: 335 who 43:. 83 me we: ms ago 83 mg 3.2 See 85 3.8. 83 E on: w: 255 a 35%: 83 85 2.2 33 33 mmmm Sod 3 32 as $822 a snow—.34 3: a. _ m 8: am as: 333 < < 295... 02m>i Mm 582 eigi Mm :82 o=_m>-n_ mm :32 s 9e as 82.. mass .8 .5 E8: @8333 Bob $55 mo e\ev 835 Hum 855 38cm 835 Km 33.25am 83% was 83 F 83:: buyout E225 E 80:08th ”Tm £an modv a .masew 50253 oocoeotmc “Swami—mi n a modv a d:9w 38.. 553 3:2: 55 .532 3:805ch 838$ n < modvm dzew .568 ESE 3:2: 55 5:3; bEeofiEwa 8388 u ... 83. mg 3w :3 :5 a: seed 85 $5 Sod So :3 mm 550 See N2 3w 83 2d 2: 83 23 £3 83 So e3 em 8:3 page: £3 £6 a: was a; a: $3 85 So 83 86 3o 3 e855. 58:2 a; :1 as one mom as: e23 2e: 8% 3o sees 83 so mo. omwd 2:. NS 33 :3 $6 we 350 see two 8.x Sod 23 NE 83. 85 one a: 85 Ed w: 8:3 a a 358.5 Ed moo 8.x 32 25 w: 33 So e2 23 Ed e; 3 5825. a... 5351‘ m: m: Ed 23 am as: £55 ... a. ... * EMF—om 623A— mm :82 623A— mm :32 o=_m>-m mm 58.2 eigd mm :82 : aeeseeeoocgfl degessooége aeeseesooimwé 523858.:me 8.55 tees 39E assess + :5 £35 ezseme> £35 gas 3:53 ”I 22¢ 96 dad 2d 8.: Edd mmd Ram wdmd E \38 w: d 3333 25 Be 852 ded 2d 3.: mmdd 3d 3.2 Mddd S R: EM 2 swag; 352 ad ad m2 _ dmdd 3d a. _ m dddd mm NS: 3 v 33325 2:8: 35 8.. _ dd. _ m d 82 m2 8 co: mass; 02 wave? mo 09? a _ d ed 8.: Edd mg 3% add NV 82 3w 3 3> 2.: 2. _ m d 28 ME 8 cod 02 03.23% w:€:o> «dmd 2d 2.: Sad ad 8.2 :Nd 8 E2 32 mm 8> 2.: 3.3. d 32 d8 2 ea: oz oEw=w>M 230 E < mad dmd wt: 83 Rd 3% dddd G 22 § mm 9:8 2 a 28 wfivcg 50m Ed dmd 8.: and odd damn deed m: wda d2 v wads? 50 M _ d 2d 5: Edd mmd dd. _ m did dd $2 «3 2 0:8 2 a $5 8.: Q. _ m d $2 a: d cad 982 DEM—Haw mvoe 03:39:00 039» Mm .502 03?» m—m Gav—Z o=_m> mm Ewe—2 c IL man can N A355 mo Xv Ahmbco mo o\dv 28v: 353% flog—om 0x35 Em @8833 8.85 3”— 8—85 Que—Hm 83:: ESE“. “sowed 5;» 5.3588 98 28256 $58 035388 mo cab 28 D=Em=m>< ”N-m 2an a . . . . . 97 modv a 838% 5233 855:6 Emoficflm -\ :85 058.8% 05 3:82;?— EBE 05 8822.: 825885582 83 03353 «803.8 - m b6 .69 03283 35b 82... .«o “38:: :38 98 o E 822238 =m wows—2: EcoE - n :5 3.8 a .8 8:805 98 58 Ba End—8% mum—8 355 Mo Bass: :85 58 Ba End—82w 83.3w? .3 33:5: :88 05 wows—2: E58 - u :3 “St a no oocomoa 98 b8 8a 03288 93b 80¢ 80 398:: :38 05 82:2: BEE - a £083? 8‘86 EB .3388 EoEomewE 33.8 too,“ .832: 30:8 035 32623on no.“ o_£w=o 3598 ..\d $538 cargo .m> 232— 458888 Ho_ “528 Sm 383:8 £238 =a - a Sod «Nd dds Edd ddd d2 add 3d 8d add mdd 8d d; d 8855; as 88 852 8d 3d 8d 8d dd 2: ddvd odd 3d Mdmd 3d 8d d? 2 8233 852 Ed 8d NE add 8d 83 add 3d 8d 3d 3d Rd E v 88833 828; 3:0 d _ d NE 3d odd 8 mm cod wads: oz wfiucoio 09? add dmd NS Ed 8d d2 mad ddd 3d wed 8d 8d :8 8 3% d2 NE 3d odd 8 mm cad oz Baa—:26 @5285 M: _d dmd 8d Ed wdd 8; Ed ddd 8d E d 3d 2d 2: m 3> d2 a: 8d 8d 98 2 co: oz DEN—33w 3.30 a. < 38d Rd NE dmdd 2d *2 mad :d ddd Edd Vdd 8d g mm 3.8 2 a can wfivfis 50m Ed 2d is Ed 2d of 8d Ed who :dd ddd 2d d2 v 86:3 25 Ed 3d _ E Bmd 2d 84 mddd 2d ddd add 3d 8d d: 2 0:8 2 a 2:0 2 d 2: 3d 8d d: d cod 282 oEm=m>w 858 3:33:80 02$, Wm 53: 0:7; mm Gav—Z 03—?» mm 532 025’ Mm 532 Z. G I“ W In N Iflm w lam m n Eddsxdddiwv M32838. :85 uaédmoxdddémwé $853088..qu 9538 28:8 835 be» 8.95 3.55 0338; 9.95 :5 0338; + dam a 3:53 “N8 29¢ 98 DISCUSSION Many significant differences were seen when examining dietary intake by race. Previous research has also identified differences amongst various racial groups in dietary intake of energy, percentage of energy intake from fat and saturated fat, certain vegetables (dark—green leafy vegetables and other starchy vegetables) and fruit (U.S. Department of Agriculture, Food and Nutrition Service et al. 2001). The racial differences in dietary intake seen in the current study may be due to true differences in food intake patterns, but may also be a limitation of the food F FQ if cultural foods commonly consumed in minority groups are not included in the food list. Only two published studies to date have examined the reliability and validity of the Block F FQ, and both were conducted in primarily minority populations (Smith and F ila 2006; Cullen, Watson et al. 2008). The first study found that in a group of 61 Native American adolescents aged 9-13, the Block FF Q was not significantly different than a 24-hr recall for estimation of energy, protein, grams of fat and saturated fat intake, and some vitamins and minerals; other nutrients such as percentage of energy from fat and saturated fat, carbohydrate, and some vitamins and minerals were significantly different (Smith and F ila 2006). The second study found in a sample of 83 Hispanic/Latino, African American, and white adolescents aged 10-17 that there were not significant differences in dietary intake of percent of energy from fat, fi'uit, and fruit juice servings, but other food groups and nutrients did show significant differences including energy, percent of energy from protein and carbohydrate, vegetables, grains, milk products, and calcium (Cullen, Watson et al. 2008). Additional studies examining the validity of FFQ results against 99 multiple 24-hr recalls in ethnically diverse samples of adolescents would help to determine whether the FF Q is an adequate representation of adolescent dietary intake. Differences were seen when examining dietary intake by gender in each individual racial group. In white students, females had significantly lower energy intake, and significantly higher saturated fat, fi'uit, vegetables, fruit + vegetables, and fiber intake than white male students. African American female students had a significantly higher vegetable intake than African American male students. Previous research has also found differences in adolescent dietary intake of energy, grams of fat and saturated fat, percentage of energy intake from saturated fat based on gender (U.S. Department of Agriculture, Food and Nutrition Service et al. 2001; Wright, Wang et al. 2003; Ervin, Wright et al. 2004; Institute of Medicine 2007). Several competitive food variables were associated with an increase in fat or saturated fat intake. These results are likely due to the types of items available in competitive food venues; many are high in fat. For example, many schools have dairy products such as plain and flavored milks, string cheese, and yogurt available because then are deemed more nutritious than typical snack foods, but they contain fat and saturated fat as well. The results that indicated that having only vending or only a la carte being associated with a higher intake of fruits was in the unexpected direction, as fruits are not ofien available in these venues. After extensive examination of SNAK school-level characteristics, it was observed that over half of schools with either vending or a la carte offer fruit snacks in those venues (e.g., Welch's fruit snacks, Fruit Roll-Up, etc.) that contain very little real fruit juice and are not fruits, but high sugar candies. Upon further 100 examination of the FF Q used, it was discovered that one of the questions measuring fruit intake asks students to report intake of "Any other fruit, like grapes, peaches, watermelon, cantaloupe, fruit rollfls". It is likely that many students reported intake of these fruit snack items in this category, which could explain how having vending or a la carte available would be associated with a higher fruit intake. Another plausible explanation is the high prevalence of mu juice-like beverages available in schools today. Four questions in the Block F FQ ask about sugar-sweetened beverages including soda, "Slurpees, snow cones, popsicles," "Hawaiian Punch, Kool- Aid, Sunny Delight, Gatorade, ice tea, Snapple," "Hi-C, Tang, Tampico, Mr. Juicy, Ssips punch". There are a wide variety of these sugar-sweetened fruit-flavored beverages available (e. g., Capri Sun) that are can easily be confused as fruit juice, and may have been reported by students under the question where they are asked to indicate consumption of "Any other real juices like apple juice or grape juice. (Remember juice boxes)". Furthermore, there are no questions that address many of the new artificially sweetened beverages available such as fruit-flavored waters, which students may be reporting as fruit juice. Research examining how students report these types of items would add to the current field of dietary assessment in adolescents. Future studies utilizing the Block FF Q should clarify instructions to adolescents on these questions to more accurately assess dietary intake of fruits. The results for the type of vending machine available showed the lowest energy intake in schools with healthy beverage-only vending machines, slightly higher energy intake with mixed beverage vending, and the highest energy intake with mixed food and beverage vending. Three out of four schools in the only healthy beverage vending lOl category only had plain or flavored water available, and the fourth school had water and 100% juice. Schools in the mixed beverage category typically carried sports drinks, other fi'uit-flavored drinks and teas, and occasionally soda in addition to water and fruit juice. The differences in the calorie content in the beverages typically offered in healthy beverage only compared with the mixed healthy and less healthy beverage only vending machines likely accounts for the differences seen in energy intake between students in these schools. Schools with only healthy beverages available in vending had a lower energy intake most likely because they mostly only had water available, whereas the sports drinks and other beverages available in mixed beverage vending machines offered more calories than water. Furthermore, the mixed food and beverage vending machines offered a wide variety of foods and drinks with a high energy content. It may seem counterintuitive that the reference group (no vending available) had the third highest energy intake, however it is important to note that over half of these schools had a la carte available. Results from this study support previous evidence that competitive foods in schools are associated with student dietary intake. Purchasing competitive foods in middle schools has been associated with a higher intake of sugar-sweetened beverages (Wiecha, Finkelstein et al. 2006); a higher intake of calories, total and saturated fat, and lower intakes of protein, vitamins A and C, and calcium (Templeton, Marlette et al. 2005). Results from the SNDA-III study indicate that not having a store or snack bar, and not having a la carte foods available were associated with a decreased consumption of energy from sugar-sweetened beverages (Briefel, Crepinsek et al. 2009). Availability of vending machines in or near the cafeteria that contain low-nutrient energy-dense foods 102 was associated with a higher BMI z-score in middle school children (Fox, Dodd et al. 2009). In previous studies, availability of competitive foods has been negatively associated with fruit and vegetable consumption and positively associated with intake of total and saturated fat (Cullen, Eagan et al. 2000; Kubik, Lytle et al. 2003; Cullen and Zakeri 2004). Similarly, in the current study, having both vending and a la carte available was associated with an increased percentage of energy intake from saturated fat. The current study also found that having mixed healthy and less healthy beverages in vending machines was associated with increased energy intake from total and saturated fat, and mixed food and beverage vending machines was associated with increased percent of energy intake from saturated fat. Also, availability of vending machines with only healthy beverages was associated with a lower consumption of vegetable and fruit + vegetable intake. In contrast, results from the current study also indicated that availability of vending machines only or a la carte only was associated with an increased fruit intake, though these results were likely due to inadequacy of the F FQ used to assess dietary intake. The small sample size of schools and students in each group in some analyses may have reduced the statistical power to detect differences between groups. Future studies conducted with more schools and students might reveal statistically significant associations. The results of the current study were mixed in their association between student dietary intake and competitive foods. Some findings indicated that competitive foods may be associated with better student dietary intake (e.g., having only healthy beverages 103 available was associated with a decreased energy intake compared with schools that did not have vending, and with schools that had mixed food and beverage vending), while other findings indicated competitive foods are associated with poor dietary intake (e.g., availability of both vending an a la carte was associated with increased saturated fat intake). It is clear that there are associations between the school nutrition environment and student dietary intake. Therefore, improving the overall healthfulness of competitive foods available in schools by increasing healthy options such as water, fruit, and vegetables, and removing less healthy options may be an effective strategy for improving adolescents' dietary intake. 104 CHAPTER FOUR: A QUALITATIVE EXPLORATION OF THE ACCOMPLISHMENTS AND CHALLENGES TO PROMOTING HEALTHY EATING IN LOW-INCOME MIDDLE SCHOOLS BACKGROUND Healthy eating during childhood and adolescence is critical to ensuring proper growth, development, and fimctioning, as well as to prevent many chronic diseases including cardiovascular disease, diabetes, obesity, and osteoporosis (Berenson, Srinivasan et al. 1998; Weaver 2000; Prentice, Schoenmakers et al. 2006). The prevalence of overweight in US. adolescents has more than tripled over the last several decades from 5.0% to 17.4%, (Ogden, Flegal et al. 2002; Ogden, Carroll et al. 2006), one indication that children are not receiving optimal nutrition. Low-income children are less likely to have a healthy diet, and more likely to consume more fat and saturated fat, and lower amounts of fruits and vegetables (Neumark-Sztainer, Story et al. 1996). Foods and beverages consumed at school are an important contributor to the total dietary intake of adolescents. However, the school food environment does not always have a positive influence on adolescents' diets. Competitive foods (those available outside of school meals, including vending machines, a la carte, fundraisers, class parties, etc.) are widely available in schools (Fox, Gordon et al. 2009), and often include items that are low in nutrient density, and high in energy, fat, sodium, and added sugars (Hamack, Snyder et al. 2000; Wildey, Pampalone et al. 2000; US. Department of 105 Agriculture 2001; French, Story et al. 2003; Wiecha, F inkelstein et al. 2006; O'Toole, Anderson et al. 2007; Fox, Gordon et al. 2009). Additionally, competitive foods have consistently been associated with poor dietary habits in students (Cullen, Eagan et al. 2000; Kubik, Lytle et al. 2003; Cullen and Zakeri 2004; Templeton, Marlette et al. 2005; Wiecha, Finkelstein et al. 2006), and are inversely associated with sales of school lunch (Fox, Crepinsek et al. 2001). In order to provide appropriate educational, environmental, and policy supports to encourage healthy eating in schools, it is necessary to understand students' perceptions of healthy eating and of the school nutrition environment. Studies indicate that adolescents have sufficient knowledge about healthy eating, are able to identify healthy and unhealthy foods, and can identify short-term physical and psychological benefits of healthy eating (Story and Resnick 1986; Chapman and Maclean 1993; Croll, Neumark- Sztainer et al. 2001; O'Dea 2003). However, nutrition knowledge and psychosocial correlates have been shown to have low predictive value for dietary intake (Baranowski, Cullen et al. 1999; Story, Neumark-Sztainer et al. 2002). It is also important to understand the perspectives of those individuals that are directly involved in making decisions regarding nutrition education and the school nutrition environment, namely administrators, teachers, food service directors, coordinated school health teams, and the community. Studies exploring barriers to health and nutrition initiatives in schools have consistently found a lack of prioritization of health initiatives (sometimes due to a focus on academic performance) and a lack of funding often being cited by school personnel (Greenberg, Cottrell et al. 2001; Meyer, Conklin et al. 2001; Winnail and Bartee 2002; Brown, Akintobi et al. 2004). In several 106 qualitative studies, students and school personnel cited challenges to eating healthy in schools that included the widespread availability of unhealthy competitive foods, low- quality school meals, insufficient time to eat, peer pressure, weight-related concerns, media promotion of unhealthy foods, and a lack of support from parents and the community (Meyer, Conklin et al. 2001; Bauer, Yang et al. 2004; Cho and Nadow 2004). While it is important to understand the challenges schools experience in promoting healthy eating, it is also important to learn from the accomplishments schools make despite these barriers and listen to their advice regarding what they need to be able to further promote health to students. Few studies have focused on school successes. Several resources available to schools share case studies of successful health promotion efforts in schools, including "Making it Happen! School Nutrition Success Stories" (Food and Nutrition Service US. Department of Agriculture, Centers for Disease Control and Prevention et al. January 2005), and the Michigan Healthy School Success Story website (http://mihealthtools.org/schoolsuccess/). This study explored the topic of healthy eating in schools in a sample of low- income middle schools from both the student and staff perspective. We chose to study low-income schools, as the factors influencing these schools may differ from those found in wealthier school districts. The goals of this study were: 1) to describe challenges to promoting healthy eating experienced by low-income middle schools; 2) to illustrate accomplishments low-income schools have made that promote healthy eating; and 3) to understand factors that facilitate school change to promote healthy eating. 107 METHODS The qualitative data used in this study were collected as part of the School Nutrition Advances Kids (SNAK) project, which is funded by the Robert Wood Johnson Foundation Healthy Eating Research program, Supplemental Nutrition Assistance Program Education (SNAP-ed), and the Michigan Department of Community Health (MDCH). SNAK is a collaboration between researchers at Michigan State University (MSU), the MDCH, the Michigan Department of Education, and partnering organizations of the Michigan Action for Healthy Kids coalition. The SNAK project aims to improve school nutrition policies and environments through school self-assessment, action planning, and implementation, and/or adoption of a Michigan State Board of Education nutrition policy. All study procedures and instruments were approved by the MSU Institutional Review Board. Informed consent was obtained from all adult participants, and parental consent and student assent were obtained for all student participants. Procedures Eight schools (of 65 total schools enrolled in the SNAK project) were selected as case study schools and invited to participate in the qualitative component of the study. All schools were low-income Michigan middle schools (50% or more students eligible for free or reduced price school meals). Case study schools were selected based on demographic characteristics to explore the diversity of experiences in these schools (Table 4-1). These characteristics included setting (rural, suburban, or urban), public vs. charter school, type of food service program (traditional kitchen, food service management company, heat-and-serve only kitchen, or no outside vendors that deliver 108 ready-to-eat food to the school), size (based on the number of 7th grade students), and building type (middle grades only; elementary and middle grades; middle and high grades; or elementary, middle, and high grades). Schools were not selected to represent all schools with middle-level grades, but to explore the topic of healthy eating in this diverse group of schools, so that the themes can be further explored in a larger sample of schools. At each case study school, interviews were conducted with one school administrator, the food service director (FSD), and one member of the coordinated school health team (CSHT), for a total of 24 school personnel interviewed. Group or individual interviews were conducted with middle school students, with the number of students interviewed ranging from 1-5 students at each school, for a total of 23 students interviewed. School personnel received a $25 gift card as an incentive to participate, and schools received $50 towards student activities as an incentive for student participation. Interviews typically lasted 30-60 minutes. Interviews were digitally recorded when consent was given (20 school personnel, 7 student groups); otherwise, detailed notes were taken during the data collection and expanded immediately thereafter (4 school personnel, 1 student group). Recordings were transcribed verbatim using word processing software. Interviews were conducted in May-June 2008, at the end of the first school year that schools participated in the SNAK Project. Follow-up interviews with school personnel were conducted in May-June 2009, after completion of SNAK project. The interviews were conducted with two goals, to understand the barriers and facilitators to promoting 109 healthy eating in this group of low income Michigan schools, and to evaluate the SNAK program activities. This paper reports on the challenges and accomplishments for promoting healthy eating described by these schools during the May-June 2008 interviews only. Instruments Interview guides were developed by the research team and partnering organizations based on review of the existing school nutrition literature and the team’s experience working in low-income middle schools. Separate interview guides were created for school administrators, FSDs, and CSHT members based on their areas of expertise; however, a number of cross-cutting questions were asked of all school personnel, including the challenges their school faced and accomplishment they have made in promoting health and nutrition. The administrator interview guide included thirteen questions, with a focus on school nutrition policies and their enforcement. The FSD interview guide included twenty-six questions related to school food service operations and competitive foods. The CSHT member interview guide included thirty questions with a focus on CSHT characteristics and SNAK project intervention activities. A separate interview guide was created for students that contained fourteen questions. The student interviews were designed to describe students' experiences with food in the school setting, and to understand their perceptions of healthy eating. 110 ANALYSIS Transcripts and notes were entered into Atlas.ti (version 5.0.66, 2005), and thematic analysis was used to establish a comprehensive list of relevant ideas, or "codes". A codebook was created that contained the code name, definition, rules for use, and examples. Initially, codes corresponded directly to questions in the interview guides. Next, a sample of transcripts was reviewed to identify additional themes and to categorize responses to interview questions. The codebook was further refined during analysis to accommodate new codes, or clarify existing codes. Codes were then attached to relevant quotations in the transcripts. Each of the student group interview transcripts was independently coded by two researchers. Inter-coder reliability was >90% after the second transcript was coded by both researchers. These transcripts were then reviewed by a third researcher, who resolved any discrepancies in coding. Six of the school personnel interview transcripts were independently coded by two researchers. Coding was compared and refined until an inter-coder reliability of >75% was achieved. The remaining 18 transcripts were coded by one researcher and reviewed by a second researcher to maximize accuracy and comprehensiveness. Once transcript coding was complete, all quotations associated with each code were reviewed, and summary statements of each code were created for each school personnel and student group. Responses were compared across schools and across participant type (administrator, FSD, CSHT member, students) when appropriate. Many common ideas were identified across interview participants, thus results are presented by lll theme, and results from the various school personnel and students are presented together when it was suitable. RESULTS Table 4-1 describes the diversity of characteristics of the eight SNAK case study schools. Five schools were public and three were charter schools. Four types of food service operations were observed. The five public schools all used "traditional" kitchens where food was prepared at each school or at central kitchen located within the district. Three of these five public schools utilized a food service management company, which oversaw the food service operations, negotiated with vendors for pricing, and provided FSDs with recipes, nutrition information, marketing materials, and educational opportunities. The three charter schools all utilized "alternative" food service programs. Two charter schools had no physical kitchen and used an outside vendor that delivered ready-to-serve foods to the schools daily. One charter school had a heat-and-serve kitchen where they were only able to warm pre—cooked frozen foods. The percent of students eligible for free or reduced price meals ranged from 50- 97%. One school was classified as rural, three as suburban, and four as urban, based on US. census data. The number of 7th grade students in the school building ranged from 49 to 248 seventh grade students. Half of the schools had middle grades only, two had elementary and middle grades, one had middle and high grades, and one served elementary, middle, and high grades in their building. 112 an Gong—0E 5.50 8:83. 892% .m EofiowmcmE 2&8 Rafi—Em .N m6 NS 53:53 mm + 3:323; _ 225% N Qmm .2 335 .m 93:58 Hmmo 8224 “Gobi .m 22:2 8050 canoes/x N220 .N mo. mm :3“: 3 H83> _ .22WE _ Qmm ._ SEED N Na ABnEoE Emmo Heyman—um :28: .m “snowman?“ 3305.5 Hgflmmlx .N wé o2 535 cm + E5223; o:We _ Dmm ._ 233m .0 “space Emu 58>» 38m .m 22:2 “Sacha Begum .N NTN N» :35 Nb chum ES .3: N .2.WE N Qmm ._ beano .m a $582: Elmo “58..”— .m EoEowdcaE 2582 grain .N wé 9: 32.52% pm + 3:223; 2 .22: v Qmm ._ 25:; .v @8505 Ewe 5%» .28: .oonom BEE—B80 “2.55 .m 22: .3655 Esmmmmzw .N m4. wVN 53:55 on 35:23:. _ 22:2 N Own 2 235 .m 53:58 HImU 852 .oonom .m 22: fifiofib—bfifiaouctonzm .N mi v: 5&3 on have“; _ 22:2 N Dmm ._ .2525 .N @382: Himo 553,—. mm .m 325 .N NT» ow 35% cm 3:223; 225m N Qmm ._ 232m .— 238 83?: fiEvBm no out 32:5 02:» .2 .8 2&2; 02:3 383m .828 5 8920 u 38% mExum 8595 g boom we 09C. 832325 $58th 33% 332285 no 2:? Fl 2:322th 322%: 25 28:8 beam 88 M150) 13 39.34 6.00 17.31 3.29 Food service management company No 36 38.89 3.35 18.66 2.24 Yes 12 44.53 5.87 19.27 2.52 Food Service Director has a nutrition-related degree No 29 38.00 3.53 18.89 2.62 Yes 17 40.63 5.00 17.10 2.10 School setting Urban 26 42.27 4.46 20.23 2.79 Rural 13 35.74 4.25 17.31 3.05 Urban Cluster 9 41.20 6.25 16.90 2.76 School type Charter/Private 8 33.59 8.74 19.53 7.23 Public 40 41.64 3.02 18.67 1.65 155 BIBLIOGRAPHY Action for Healthy Kids (2007). Local Wellness Policies One Year Later: Showing Improvements in School Nutrition and Physical Activity. Action for Healthy Kids (Fall 2008). Progress or Promises? What's Working For and Against Healthy Schools. Available at: http://wwwactionforhealt_hykids.org/pdf/Progress%200r%2OPromises.pdf Alaimo, K., C. Olson, et al. (2001). "Food insufficiency and American school-aged children's cognitive, academic, and psychosocial development." Pediatrics 108(1): 44-53. Allensworth, D. D. and L. J. Kolbe (1987). "The comprehensive school health program: exploring an expanded concept." J Sch Health 57(10): 409-12. American Dietetic Association (2006). "Position of the American Dietetic Association: Local support for nutrition integrity in schools." J Am Diet Assoc 106(1): 122- 133. Atkinson, R. L. and S. A. Nitzke (2001). "School based programmes on obesity." BMJ 323(7320): 1018-1019. Baranowski, T., K. W. Cullen, et al. (1999). "Psychosocial correlates of dietary intake: Advancing dietary intervention." Annu Rev Nutr 19: 17-40. Baranowski, T., M. Davis, et a1. (2000). "Gimme 5 fruit, juice, and vegetables for fun and health: Outcome evaluation." Health Educ Behay 27(1): 96-111. Barratt, R. D., N. A. Cross, et a1. (2004). "School policies that promote healthy eating: A survey of foodservice directors in North Carolina public schools." The Journal of Child Nutrition & Margementfl). Bauer, K. W., Y. W. Yang, et al. (2004). ""How can we stay healthy when you're throwing all of this in front of us?" Findings from focus groups and interviews in middle schools on environmental influences on nutrition and physical activity." Health education & behavior 31(1): 34-46. 156 Belansky, E. S., N. Cutforth, et al. (2009). "Early impact of the federally mandated local wellness policy on school nutrition environments appears modest in Colorado's rural, low-income elementary schools." In submission. Belansky, E. S., N. Cutforth, et al. (2009). "Early impact of the federally mandated local wellness policy on physical activity in rural, low-income elementary schools in Colorado." J Public Health Pol 30(81): 8141-8160. Berenson, G. S., S. R. Srinivasan, et a1. (1998). "Atherosclerosis: A nutritional disease of childhood." Am J Cardiol 82(IOB): 22T-29T. Briefel, R. R., M. K. Crepinsek, et al. (2009). "School food environments and practices affect dietary behaviors of US public school children." J Am Diet Assoc 109(2 Suppl): S91-107. Briefel, R. R. and C. L. Johnson (2004). "Secular trends in dietary intake in the United States." Annu Rev Nutr 24: 401-431. Briefel, R. R., A. Wilson, et al. (2009). "Consumption of low-nutrient, energy-dense foods and beverages at school, home, and other locations among school lunch participants and nonparticipants." J Am Diet Assoc 109(2 Suppl): $79-90. Brown, K. M., T. H. Akintobi, et a1. (2004). "California school board members' perceptions of factors influencing school nutrition policy." J Sch Health 74(2): 52-8. Brownell, K. D., R. Kersh, et a1. (2010). "Personal responsibility and obesity: a constructive approach to a controversial issue." Health Affairs 29(3): 379-387. Bush, G. W. (2001). No Child Left Behind. Washington, DC: Office of the President. Cassady, D., R. Vogt, et al. (2006). "The power of policy: A case study of healthy eating among children." Am J Public Health 96(9): 1570-1571. Centers for Disease Control and Prevention (2009). Coordinated School Health Program. Available at: http://www.cdc.gov/Healtthouth/CSHP/. 157 Centers for Disease Control and Prevention. (2005). Youth Risk Behavior Survey, Available at: www.cdc.gov/yrbss. Chang, V. W. and D. S. Lauderdale (2005). "Income disparities in body mass index and obesity in the United States, 1971-2002." Arch Intern Med 165(18): 2122-2128. Chapman, G. and H. Maclean (1993). ""Junk F ood" and "Healthy Food": Meanings of food in adolescent women's culture." J Nutr Educ 25(3): 108-113. Cho, H. and M. Z. Nadow (2004). "Understanding barriers to implementing quality lunch and nutrition education." Journal of Community Health 29(5): 421-435. Clark, M. A. and M. K. Fox (2009). "Nutritional quality of the diets of US public school children and the role of the school meal programs." J Am Diet Assoc 109(2 Suppl): $44-56. Contento, I., G. Balch, et al. (1995). "The effectiveness of nutrition education and implications for nutrition education policy, programs, and research: A review of research." J Nutr Educ 26(6): 284-418. Croll, J ., D. Neumark-Sztainer, et al. (2001). "Healthy eating: What does it mean to adolescents?" J Nutr Educ 33(4): 193-198. Cullen, K. W., T. Baranowski, et al. (2000). "Social-environmental influences on children's diets: Results from focus groups with Afi'ican-, Euro- and Mexican- American children and their parents." Health Educ. Res. 15(5): 581-590. Cullen, K. W., J. Eagan, et al. (2000). "Effect of a la carte and snack bar foods at school on children's lunchtime intake of fruits and vegetables." J Am Diet Assoc 100(12): 1482-1486. Cullen, K. W., K. Watson, et al. (2008). "Relative Reliability and Validity of the Block Kids Questionnaire among Youth Aged 10 to 17 Years." J Am Diet Assoc 108(5): 862-866. Cullen, K. W., K. Watson, et al. (2006). "Exploring changes in middle-school student lunch consumption after local school food service policy modifications." Public Health Nutr 9(6): 814-820. 158 Cullen, K. W. and I. Zakeri (2004). "Fruits, vegetables, milk, and sweetened beverages consumption and access to a la carte/snack bar meals at school." Am J Public Health 94(3): 463-467. Davis, S. M., T. Clay, et al. (1999). "Pathways curriculum and family interventions to promote healthful eating and physical activity in American Indian schoolchildren." Prev Med 37(Supp1ement 1): 824-834. DHHS (2004). Child Health. Rockville, MI, US. DHHS. Dinour, L. M., D. Bergen, et al. (2007). "The food insecurity-obesity paradox: A review of the literature and the role food stamps may play." J Am Diet Assoc 107(11): 1952-1961. Economos, C. (2008). Plenary Session - Community Interventions: Shape Up Somerville. Healthy Foods, Health Moves: Delivefinghe Childhood Obesity Prevention Message in Schools and Communities. Chicago, IL. Edmundson, E., G. Parcel, et al. (1996). "The effect of the child and adolescent trial for cardiovascular health upon psychosocial determinants of diet and physical activity behavior." Prev Med 25: 442-454. Engels, H. J ., R. J. Gretebeck, et al. (2005). "Promoting healthful diets and exercise: efficacy of a 12-week after-school program in urban Afi'ican Americans." J Am Diet Assoc 105(3): 455-459. Ervin, R. B., J. D. Wright, et al. (2004). "Dietary intake of fats and fatty acids for the United States population: 1999-2000." Adv Data(348): 1-6. Ervin, R. B., J. D. Wright, et a1. (2004). "Dietary intake of selected vitamins for the United States population: 1999-2000." Adv Data(339): 1-4. F anjiang, G. and R. E. Kleinman (2007). "Nutrition and performance in children." Curr Qpin Clin Nutr 10(3): 342-7. Figlio, D. and J. Winicki (2005). "Food for thought: the effects of school accountability plans on school nutrition." Journal of Public Economics 89(2-3): 381-394. 159 F inkelstein, D. M., E. L. Hill, et al. (2008). "School food environments and policies in US public schools." Pediatrics 122(1): e251-259. Food and Nutrition Service US. Department of Agriculture, Centers for Disease Control and Prevention, et al. (January 2005). FNS-374, Making it Happen! School Nutrition Success Stories. Alexandria, VA. Fox, M. K. and N. Cole (2004). Nutrition and Health Characteristics of Low-Income Populations. Volume 3, School-Age Children, United States Department of Agriculture, Economic Research Service. Fox, M. K., P. Crepinsek, et al. (2001). School Nutrition Dietary Assessment Study-II: Summary of Findings. Alexandria, VA, US Department of Agriculture, Food and Nutrition Service, Office of Analysis, Nutrition and Evaluation. Fox, M. K., A. H. Dodd, et al. (2009). "Association between school food environment and practices and body mass index of US public school children." J Am Diet Assoc 109(2 Suppl): 8108-117. Fox, M. K., A. Gordon, et al. (2009). "Availability and consumption of competitive foods in US public schools." J Am Diet Assoc 109(2 Suppl): 857-66. French, 8., R. Jeffery, et a1. (2001). "Pricing and promotion effects on low-fat vending snack purchaseszThe CHIPS Study." Am J Public Health 91(1): 112-117. French, S. A., M. Story, et al. (2002). "School food policies and practices: A state-wide survey of secondary school principals." J Am Diet Assoc 102(12): 1785-1789. French, S. A., M. Story, et al. (2003). "Food environment in secondary schools: A la carte, vending machines, and food policies and practices." Am J Public Health 93(7): 1161-1167. French, S. A., M. Story, et al. (2004). "An environmental intervention to promote lower- fat food choices in secondary schools: Outcomes of the TACOS Study." Am J Public Health 94(9): 1507-1512. French, S. A., M. Story, et al. (1997). "Pricing strategy to promote fruit and vegetable purchase in high school cafeterias." J Am Diet Assoc 97(9): 1008-1010. 160 Friedman, R. R. (2009). School wellness policies: a policy brief, Rudd Center For Food Policy & Obesity, Yale University, New Haven, CT. .Fu, M.-L., L. Cheng, et a1. (2007). "Association between unhealthful eating patterns and unfavorable overall school performance in children." J Am Diet Assoc 107(11): 1935-1943. Fulkerson, J. A., S. A. French, et al. (2004). "Promotions to increase lower-fat food choices among students in secondary schools: Description and outcomes of TACOS (Trying Alternative Cafeteria Options in Schools)" Public Health Nutr 7(5): 665-674. Gemmill, E. and N. Cotugna (2005). "Vending machine policies and practices in Delaware." The Journal of school nursing: the official publication of the National Association of School Nurses 21(2): 94-99. Gleason, P. M. and A. H. Dodd (2009). "School breakfast program but not school lunch program participation is associated with lower body mass index." J Am Diet Assoc 109(2 Suppl): S118-128. Gortmaker, S. L., L. W. Y. Cheung, et a1. (1999). "Impact of a school-based interdisciplinary intervention on diet and physical activity among urban primary school children: Eat well and keep moving." Arch Pediatr Adolesc Med 153(9): 975-983. Greenberg, J ., R. Cottrell, et al. (2001). "Baseline data on coordinated school health programs in the state of Ohio." American Journal of Health Studies 17(1): 1-6. Greves, H. M. and F. Rivara (2006). "Report card on school snack food policies among the United States' largest school districts in 2004-2005: Room for improvement." International Journal of Behavioral Nutrition and Physical Activity 3(1): 1. Hamack, L., M. P. Snyder, et al. (2000). "Availability of A La Carte Food Items in Junior and Senior High Schools: A Needs Assessment." Journal of the American Dietetic Association 100(6): 701-703. Hamack, L., P. Snyder, et al. (2000). "Availability of a la carte food items in junior and senior high schools: A needs assessment." J Am Diet Assoc 100(6): 701-703. 161 Harrell, J. S., S. A. Gansky, et al. (1998). "School-based interventions improve heart health in children with multiple cardiovascular disease risk factors." Pediatrics 102(2 Pt 1): 371-380. Heaney, R. P., S. Abrams, et al. (2000). "Peak bone mass." Osteoporos Int 11(12): 985- 1009. Institute of Medicine (2006). Dietary Reference Ingrkes: The Essential Guide to Nutrient Requirements Washington, DC, National Academies Press. Institute of Medicine (2007). Nutrition Standards for Foods in Schools: Leadinflhe Way Toward Healthier Youth. Washington, DC, The National Academies Press. Jago, R., J. S. Harrell, et a1. (2006). "Prevalence of abnormal lipid and blood pressure values among an ethnically diverse population of eighth-grade adolescents and screening implications." Pediatrics 117(6): 2065-2073. James, J ., P. Thomas, et al. (2004). "Preventing childhood obesity by reducing consumption of carbonated drinks: cluster randomised controlled trial." BMJ 328(7450): 1237. Johnson, E. J. and D. Goldstein (2003). "Medicine: do defaults save lives?" Science 302(5649): 1338-1339. Joint FAO/WHO expert consultation on diet, n. a. t. p. o. c. d. (2003). "Diet, nutrition, and the prevention of chronic diseases." World Health Oigan Tech Rep Ser 916: 58-63. Kant, A. K. (2003). "Reported consumption of low-nutrient-density foods by American children and adolescents: Nutritional and health correlates, NHANES III, 1988 to 1994." Arch Pediatr Adolesc Med 157(8): 789-796. Kelder, S. H., C. L. Perry, et al. (1995). "Community-wide youth nutrition education: Long-term outcomes of the Minnesota Heart Health Program." Health Educ. Res. 10(2): 119-131. Kristal, A. R., U. Peters, et al. (2005). "Is it time to abandon the food frequency questionnaire?" Cancer Epidemiol Biomarkers Prev 14(12): 2826-2828. 162 Kristal, A. R. and J. D. Potter (2006). "Not the time to abandon the food frequency questionnaire: counterpoint." Cancer Epidemiol Biomarkers Prev 15(10): 1759- 1760. Kubik, M. Y., L. A. Lytle, et al. (2009). "Food use in middle and high school fundraising: does policy support healthful practice? Results from a survey of Minnesota school principals." J Am Diet Assoc 109(7): 1215-1219. Kubik, M. Y., L. A. Lytle, et al. (2003). "The association of the school food environment with dietary behaviors of young adolescents." Am J Public Health 93(7): 1168- 1173. Larson, N. and M. Story (2010). "Are 'competitive foods' sold at school making our children fat?" Health Affairs 29(3): 430-435. . Lingstrom, P., A. K. Holm, et al. (2003). "Dietary factors in the prevention of dental caries: A systematic review." Acta Odontol Scand 61(6): 331-340. Longley, C. H. and J. Sneed (2009). "Effects of federal legislation on wellness policy formation in school districts in the United States." J Am Diet Assoc 109(1): 95- 101. Luepker, R. V., C. L. Perry, et al. (1996). "Outcomes of a field trial to improve children's dietary patterns and physical activity. The Child and Adolescent Trial for Cardiovascular Health. CATCH collaborative group." JAMA 275(10): 768-776. Lytle, L. A. and J. A. F ulkerson (2002). "Assessing the dietary environment: Examples from school-based nutrition interventions." Public Health Nutr 5(6A): 893-899. Lytle, L. A., M. Y. Kubik, et al. (2006). "Influencing healthful food choices in school and home environments: Results from the TEENS study." Prev Med 43(1): 8-13. Lytle, L. A., D. M. Murray, et al. (2004). "School-based approaches to affect adolescents' diets: Results from the TEENS study." Health Educ Behav 31(2): 270-287. Lytle, L. A., E. J. Stone, et al. (1996). "Changes in nutrient intakes of elementary school children following a school-based intervention: Results from the CATCH Study." Prev Med 25(4): 465-477. 163 Manios, Y., J. Moschandreas, et al. (1999). "Evaluation of a health and nutrition education program in primary school children of crete over a three-year period." Prev Med 28(2): 149-159. McDonnell, E., C. Probart, et al. (2006). "School competitive food policies: Perceptions of Pennsylvania public high school foodservice directors and principals." J Am Diet Assoc 106(2): 271-276. McGraw, S. A., D. Sellers, et a1. (2000). "Measuring implementation of school programs and policies to promote healthy eating and physical activity among youth." Preventative Medicine 31(2): $86-$97. Metos, J. and M. S. Nanney (2007). "The strength of school wellness policies: One state's experience." J School Health 77(7): 367-372. Meyer, M. K., M. T. Conklin, et al. (2001). "Barriers to healthy nutrition environments in public school middle grades." Journal of Child Nutrition and Management 25(2): 66-71. Michigan Action for Healthy Kids (Fall 2007). Local Wellness Policy Implementation: Strategies and Successes in Michigan, Michigan Department of Education. Michigan Department of Education (October 2007). Local Wellness Policy Implementation Grant Data Report: Status of Local Wellness Policy Adoption and Implementation among Michigan Local Education Agencies. Michigan Department of Education, Michigan Department of Community Health, et al. (2001). The Role of Michigan Schools in Promoting Healthy Weight. Moag-Stahlberg, A., N. Howley, et al. (2008). "A national snapshot of local school wellness policies." J School Health 78(10): 562-570. Moreno, N. P., J. P. Denk, et al. (2004). "An approach to improving science knowledge about energy balance and nutrition among elementary- and middle-school students." Cell Biol Educ 3(2): 122-130. Moshfegh, A., J. D. Goldman, et al. (2005). What we eat in America, NHANES 2001- 2002: Usual nutrient intakes from food compared to Dietary Reference Intakes US. Department of Agriculture, Agricultural Research Service. 164 Murphy, J ., C. Wehler, et a1. (1998). "Relationship between hunger and psychosocial functioning in low-income American children." J Am Acad Child Adolesc Psychiatgg 37(2): 163-170. Murphy, J. M., M. E. Pagano, et al. (1998). "The relationship of school breakfast to psychosocial and academic functioning: cross-sectional and longitudinal observations in an inner-city school sample." Arch Pediat Adol Med 152(9): 899- 907. Murray, N. G., B. J. Low, et al. (2007). "Coordinated school health programs and academic achievement: a systematic review of the literature." J Sch Health 77(9): 589-600. Nader, P. R., E. J. Stone, et al. (1999). "Three-year maintenance of improved diet and physical activity: The CATCH cohort." Arch Pediat Adol Med 153(7): 695-704. Neumark-Sztainer, D., M. Story, et al. (1999). "Factors influencing food choices of adolescents: Findings from focus-group discussions with adolescents." J Am Diet Assoc 99(8): 929-37. Neumark-Sztainer, D., M. Story, et al. (1996). "Correlates of inadequate fruit and vegetable consumption among adolescents." Prev Med 25(5): 497-505. O'Dea, J. A. (2003). "Why do kids eat healthful food? Perceived benefits of and barriers to healthfirl eating and physical activity among children and adolescents." J Am Diet Assoc 103(4): 497-501. O'Toole, T. P., S. Anderson, et a1. (2007). "Nutrition services and foods and beverages available at school: Results from the school health policies and programs study 2006." J Sch Health 77(8): 500-521. Ogden, C. L., M. D. Carroll, et al. (2006). "Prevalence of overweight and obesity in the United States, 1999-2004." JAMA 295(13): 1549-1555. Ogden, C. L., K. M. Flegal, et al. (2002). "Prevalence and trends in overweight among US children and adolescents, 1999-2000." JAMA 288(14): 1728-1732. 165 Osganian, S. K., M. K. Ebzery, et al. (1996). "Changes in the nutrient content of school lunches: Results from the CATCH eat smart food service intervention." Prev Med 25(4): 400-412. Parcel, G., B. Simons-Morton, et al. (1989). "School promotion of healthful diet and physical activity: impact on learning outcomes and self-reported behavior." Health Educ Q 16: 181-99. Perry, C. L., D. B. Bishop, et a1. (2004). "A randomized school trial of environmental strategies to encourage fruit and vegetable consumption among children." Health Educ Behav 31(1): 65-76. Perry, C. L., L. A. Lytle, et al. (1998). "Effects of the Child and Adolescent Trial for Cardiovascular Health (CATCH) on fruit and vegetable intake." Journal of Nutrition Education 30(6): 354-360. Pliner, P. (1982). "The effects of mere exposure on liking for edible substances." Appetite 3: 283-90. Pollitt, E. (1997). "Iron deficiency and educational deficiency." Nutr Rev 55(4): 133-141. Prentice, A., I. Schoenmakers, et al. (2006). "Nutrition and bone growth and development." Proc Nutr Soc 65(4): 348-360. Probart, C., E. McDonnell, et al. (2008). "Statewide assessment of local wellness policies in Pennsylvania public school districts." J Am Diet Assoc 108(9): 1497-1502. Prochaska, J. O., C. A. Redding, et al. (2002). The transtheoretical model and stages of change. Health Behavior and Health Education. K. Glanz, B. K. Rimer and F. Lewis, Jossey-Bass: 99- 120. Rosas, S., J. Case, et al. (2009). "A retrospective examination of the relationship between implementation quality of the coordinated school health program model and school-level academic indicators over time." J Sch Health 79(3): 108-115. Sallis, J. F., T. L. McKenzie, et al. (2003). "Environmental interventions for eating and physical activity: A randomized controlled trial in middle schools." Am J Prev Med 24(3): 209-217. 166 Sallis, J. F. and N. Owen (2002). Ecological models of health behavior. Health Behavior and Health Education: TheoLv, Research and Practice. K. Glanz, B. K. Rimer and F. M. Lewis. San Francisco, CA, Josssey-Bass. School Nutrition Association (2006). A Foundation for the Future 11: Analysis of Local Wellness Policies from 140 School Districts in 49 States. Schwartz, M. B., A. E. Lund, et al. (2009). "A comprehensive coding system to measure the quality of school wellness policies." J Am Diet Assoc 109(7): 1256-1262. Scientific Software lntemational (2009). HLM 6.08. Lincolnwood, IL. Section 204 of Public Law 108-265 (June 3, 2004). Child Nutrition and WIC Reauthorization Act of 2004; www.usda.gov/tn/Heath/ 1 08-265. Shahid, B. (2003). "A study of school principals and the promotion of nutritional health in middle grade schools." Education 123(3): 552-563, 569. Simons-Morton, B. G., G. S. Parcel, et al. (1991). "Promoting physical activity and a healthful diet among children: Results of a school-based intervention study." _A_r_n_ J Public Health 81(8): 986-991. Slusser, W. M., W. G. Cumberland, et al. (2007). "A school salad bar increases frequency of fruit and vegetable consumption among children living in low-income households." Public Health Nutr 10(12): 1490-6. Smith, C. and S. F ila (2006). "Comparison of the Kid's Block Food Frequency Questionnaire to the 24-hour recall in urban Native American youth." American Journal of Human Biology 18(5): 706-709. Sorof, J. M., D. Lai, et al. (2004). "Overweight, ethnicity, and the prevalence of hypertension in school-aged children." Pediatrics 113(3 Pt 1): 475-482. Stevenson, J. (2006). "Dietary influences on cognitive development and behaviour in children." P Nutr Soc 65(04): 361-365. Story, M. (1999). "School-based approaches for preventing and treating obesity." International Journfial of Obesity 23(Suppl 2): S43-SS 1. I67 US. Department of Agriculture, Food and Nutrition Service, et a1. (2007). School Nutrition Dietary Assessment Study-III, Vol. II: Student Participation and Dietary Intakes, by Anne Gordon, et al. Project Officer: Patricia McKinney. Alexandria, VA. US. Department of Health and Human Services (2000). Healthy People 2010. 2nd ed. With Understanding and Irrmroving Health and Objectives for Improving Health. 2 vols. Washington, DC, US. Government Printing Office. US. Department of Health and Human Services and US. Department of Agriculture (2004). Report of the Dietary Guidelines Advisory Committee on the Dietary Guidelines for Americans, 2005. A Report to the Secretary of Health and Human Services and the Secretary of Agriculture. [Online]. Available: http://wwwhealthgov/dietmguidelines/dgaZOOS/report/ [accessed May 30, 2009} US. Government Accountability Office (2005). School Meal Programs: Competitive Foods Are Widely Available and Generate Substantial Revenues for Schools. Wang, Y. and Q. Zhang (2006). "Are American children and adolescents of low socioeconomic status at increased risk of obesity? Changes in the association between overweight and family income between 1971 and 2002." Am J Clin Nutr 84(4): 707-716. Wardle, J ., M.-L. Herrera, et al. (2003). "Modifying children's food preferences: the effects of exposure and reward on acceptance of an unfamiliar vegetable." European Journal of Clinical Nutrition 57(2): 341-348. Weaver, C. M. (2000). "The growing years and prevention of osteoporosis in later life." Proc Nutr Soc 59(2): 303-306. Webber, L. S., S. K. Osganian, et a1. (1996). "Cardiovascular risk factors among children after a 2 l/2-year intervention-the CATCH study." Prev Med 25(4): 432-441. Wharton, C. M., M. Long, et al. (2008). "Changing nutrition standards in schools: the emerging impact on school revenue." J Sch Health 78(5): 245-251. Wiecha, J. L., D. Finkelstein, et al. (2006). "School vending machine use and fast-food restaurant use are associated with sugar-sweetened beverage intake in youth." 1 Am Diet Assoc 106(10): 1624-1630. 169 Wildey, M. B., S. Z. Pampalone, et a1. (2000). "Fat and sugar levels are high in snacks purchased from student stores in middle schools." J Am Diet Assoc 100(3): 319- 322. Willett, W. (1998). Nutritional Epidemiology, 2nd ed. New York, NY, Oxford University Press. Willett, W. and F. B. Hu (2006). "Not the time to abandon the food frequency questionnaire: point." Cancer Epidemiol Biomarkers Prev 15(10): 1757-8. Winnail, S. D. and R. T. Bartee (2002). "How can primary concerns of school district superintendents guide school health efforts?" J Sch Health 72(10): 408-412. Wojcicki, J. M. and M. B. Heyman (2006). "Healthier choices and increased participation in a middle school lunch program: Effects of nutrition policy changes in San Francisco." Am J Public Health 96(9): 1542-1547. Wright, J. D., C. Y. Wang, et al. (2003). "Dietary intake of ten key nutrients for public health, United States: 1999-2000." Adv Data(334): 1-4. 170