Impact of competitive foods in public schools on child nutrition: effects on adolescent obesity in the United States an integrative systematic literature review

ABSTRACT Background: The United States (US) is currently facing a public health crisis due to the percentage of obesity in adolescents. The Center for Disease Control (CDC) stated the risks for children due to obesity are many. Adolescents obtain a large portion of their daily caloric intake at school; therefore, what foods/drinks they are consuming is so serious. Objective: To identify and analyze literature on the effects of competitive foods in public schools on adolescent weight, or Body Mass Index (BMI), and possible impacts they may have on adolescent obesity in the United States. Methods: An integrative systematic review of literature was conducted. The literature was collected in CINAHL, MEDLINE and EMBASE databases. Refined keyword search is further detailed in the report. Year restrictions were 2006–2017 from peer-reviewed journals and published in English, including adolescents 13–18 years old in the US. Criteria for inclusion targeted at least one of (1) sugar-sweetened beverages (SSB), (2) competitive foods, (3) commercial foods, (4) vending machines, (5) al a carte venues, and (6) school stores, examining their associations with weight measurements, using either weight or BMI, or caloric intake analysis. Results: A total of 164 articles were detected and assessed, for a final analysis of 34 full text articles. Twenty-six articles met the inclusion criteria. Common aspects of interest involved BMI/Obesity/Weight (73%), (58%) examined Calorie density or consumption, (77%) discussed the Availability of competitive foods in schools, (54%) included Analysis of competitive food, beverage and nutrition policies, and (69%) addressed Other effects. Conclusion: This review discovered substantial evidence that competitive foods are highly available in schools, however, lacking in robust evidence proving causality in increasing BMI or weight. There is strong corroboration in the research revealing that Other effects are factors worthy of studying further. Additional longitudinal and higher-quality research needs to be performed.

School health policy; child nutrition; Body Mass Index (BMI); junk food; foods of minimal nutritional value; (FMNV); commercial food; vending machines; sugar-sweetened beverages; al a carte Background Obesity is currently a public health epidemic in the US, with more than one-third of adolescents (aged [13][14][15][16][17][18] currently weighing in as overweight or even obese [1]. BMI (weight (kg)/height (m2)) is accepted as the standard index for defining overweight and obesity due to the specificity and sensitivity it delivers [2]. It is projected that as the population becomes older, both public health and economic burdens from obesity will grow along with it [3]. The CDC reported concerning adolescent obesity, the risks for children due to obesity are numerous, including the many complications which are linked to Type 2 diabetes, including stroke, diabetic neuropathy, gastroparesis, heart disease, blindness, and even depression [4]. It is because of these increasing public health concerns that we turn our sights to the youth of America to discover why obesity is growing so rapidly and what can be done to reduce it.
Adolescents receive more than 40% of their daily caloric intake while in school, therefore, the influence of what foods and beverages they are consuming is so critical [5]. However, even with initiatives such as The Healthy, Hunger-Free Kids Act of 2010, which sets standards on competitive foods that are allowed to be sold throughout the school day for students, and the National School Lunch Program (NSLP) [6], there is still an undeniable need for further research to determine why we are still seeing a concerning increase in weight and BMI in the youth across the country.
There are many contributors to the obesity problem, with a significant amount of research on the marketing of unhealthy food towards youth and children [7][8][9][10][11][12]. Other factors include increased portion sizes, and the multitudinous snack food options offered [13].Studies on commercial and competitive foods which, according to the US Department of Agriculture (USDA), are foods available to be purchased by students which are not part of the regular school lunch programs [1] have been ongoing [5,[14][15][16][17][18][19][20][21][22][23][24][25][26][27][28], with research thus far examining the fact that youth in schools which serve a la carte meals eat fewer healthy options, and in a school setting where soft drinks and junk foods are sold, are more likely to be obese [29].
Influences in an eating environment may directly affect food intake that could lead to overeating as well as increased risks of obesity [30]. The concern of the obesity pandemic is enough to warrant significant changes in how the US addresses issues such as 'population-wide education, training, and motivation concerning obesity' [31, p. 620]. Further, it is worthy to examine Foods of Minimal Nutritional Value (FMNV), which are the food choices most often available in venues of competitive food sources (a la carte, vending, and school stores) [32]. Current research also proposes that some types of carbohydrates, particularly fructose, might play a significant role in increased adiposity [33].
Hennessy et al. [26] concluded that competitive food and beverage laws within schools deserve to gain more attention as we look to addressing the obesity epidemic currently facing our nation. According to a review performed by Jaime & Lock [34, p. 52], 'there are currently few studies which have measured the impact of school food policies on BMI'. Although adolescent obesity has been researched and discussed extensively, to my knowledge, research examining the association between availability of competitive foods in school settings and adolescent weight (BMI) has not been systematically reviewed.
The purpose of this integrative systematic literature review is to make known the literature on research which has addressed commercial foods in schools and adolescent weight (BMI), and synthesize the data to explore recommendations for further research, increase the potential for stronger health policies, and hopefully provide insight into how to better protect our youth and combat the health crisis of obesity which America is currently facing. The main aim of this thesis is to summarize the research on the effects of competitive foods in public schools on adolescent weight (BMI), and possible impact it may have on adolescent obesity in the US, by asking 'In what ways are the weight (BMI) of adolescents impacted by competitive foods in schools?'

Methods
To accomplish this aim, this integrative systematic review will attempt to classify quantitative peerreviewed studies on nutrition policies in public schools which specifically address (1) competitive foods in schools and (2) the impact competitive food sources may have on adolescent weight (BMI). The methodology used in this study is an integrative systematic literature review. Data were collected through a systematic literature search.
This review will attempt to identify school-based nutrition policies of school settings which allow for competitive foods [1,32,35]. The participants in the primary studies that will be included in this integrative systematic review will be schools (public and private) in the US (Middle and High Schools) whose student population is between 13 and 18 years of age. In this review, competitive foods, often FMNV, include foods which offer less than 5% of the Reference Daily Intake (RDI) for eight selected nutrients in each serving [36], and will include SSB's, commercial foods, vending machines, al a carte venues, and school stores.
Quantitative studies that have examined schoolbased nutrition policies which include or allow for competitive foods in school settings and discuss weight (BMI) and/or ways in which weight (BMI) is impacted by competitive foods in schools, were considered.

Literature search
Cumulative Index to Nursing & Allied Health Literature (CINAHL); MEDLINE; and EMBASE databases were searched. Additional sources were examined, including bibliographies of applicable articles and relevant literature reviews, per the Methodological Expectations of Cochrane Intervention Reviews (MECIR), which advises to 'check reference lists in included studies and any relevant systematic reviews identified' [37, p. 16].
The screening process included scanning titles and brief abstracts initially, then full abstracts for relevance per inclusion criteria. The criteria for inclusion were: (1) peer-reviewed journals; (2) published 2006-2017; (3) English; (4) 13-18 years of age in middle or high schools; and (5) in the US. Criteria for inclusion also included at least one of the following measures: SSB's, competitive foods, commercial foods, vending machines, al a carte venues, school stores, and their associations with weight (BMI) for adolescents in school settings. This included either actual BMI calculation, and/or caloric intake analysis. Only peer-reviewed journal articles were indexed in MEDLINE, CINHAL, and EMBASE databases. Keywords used in the review are listed in Tables 1 and 2. A search was performed in EMBASE of six aspects, resulting in 66 articles. A quick-screening through the brief abstract of articles was completed, which resulted in 39 possible articles (66 -27 = 39). Similarly, a search was performed in EBSCO which allowed the combination of both MEDLINE and CINAHL databases to be searched simultaneously, resulting in 125 articles. Combining EMBASE results (39) with those from MEDLINE and CINAHL (39 + 125 = 164), gave a total of 164, prior to checking for duplicates (164-75 duplicates = 89).

Screening of abstracts
After the initial abstract screening of the 89 potentially qualified articles from all three databases (EMBASE, MEDLINE, and CINAHL) was performed, 24 articles were accepted, prior to quality check. Bibliographies were scrutinized to determine if anything more needed to be included, based on MECIR, and (10) additional articles were included (24 + 10 = 34).Additionally, according to the MECIR, 'a PRISMA flow chart and a table of "Characteristics of excluded studies" will need to be completed in the final review.' [37, p. 13] (see Figure 1).

Data extraction
Further in-depth examination using a data extraction method, inspired by Urzi [38] (see Appendix 1), was utilized in narrowing down the search further,  probing if the participants, settings, and interventions, are pertinent to the Key Question [39]. A pilot test of the data-extraction instrument was conducted by two reviewers, and deemed appropriate. Due to conflict in time and feasibility, the remaining 33 articles were tested by one reviewer. After utilizing the data-extraction system, and cross checking that probative information was addressed, (8) more articles were excluded with reasons (see Table 3). This left 26 articles which met inclusion criteria. Intra-rater reliability was insured through re-analysis of all data.
Once the screening was completed using the data analysis extraction form (see Appendix 1), the remaining 26 articles were examined more thoroughly via a more comprehensive reading of the 26 articles searching for pertinent aspects within each article which addressed the key question and criteria. The information discovery resulted in varying outcomes based on the criteria.

Analysis
This paper confines itself to summarizing the results in an integrative analysis as the heterogeneity of the studies investigated in this review are not suitable for a metaanalysis. Haidich [40] states that systematic reviews may not need to contain a meta-analysis as there may be occasions where it is not possible or even appropriate. Moher et al. [72]. Table 3. Excluded Articles with Reasons.
References: Reasons for Exclusion from Review: Cisse-Egbuonye et al. [32] Study identified the types of foods sold in vending machines and school stores in schools, discussed the availability and consumption associated with student purchase, but does not address weight/BMI/or caloric intake. Gordon et al. [66] Discusses availability of competitive foods in schools, school meal programs, in the prevalence of overweight and obesity. Does not address calories, BMI, or overall weight differences due to competitive foods in schools directly. Consumption, and availability, not weight. Hawkes et al. [67] Evaluates district policies addressing student wellness, developed an approach to determine best changes in policy to access to health food/drink in school cafeteria lunches. Does not address weight/BMI/caloric intake. Kristensen et al. [68] Discusses SSB taxes and impact on BMI in adolescents, but does not directly address this in school settings specifically. Kubik et al. [20] Discusses junk food in school vending machines and school stores and school nutrition policies, however does not address weight/BMI/caloric intake specifically. Consumption, not weight. Schwartz et al. [69] Discusses perceptions of weight and diet with regards to consumption of snack foods sold in schools, but actual weight, BMI, or calories were not measured or evaluated. Consumption, not weight. Thompson et al. [70] Primary purpose was to examine purchasing behaviors with school policy. Calculated purchasing behavior and dietary consumption, but did not address the research question on weight/BMI/or caloric intake. Wordell et al. [71] Study discusses school food environments, healthier school policies, and consumption and choice making of competitive foods. Does not address weight/BMI/caloric intake specifically. Consumption, not weight.

Results
The following summary of data aims to answer the objective of what impacts there are upon child weight (BMI), and/or caloric intake related to obesity when competitive foods are sold in schools, and it is the purpose of this thesis to investigate and identify any recommendations for further research as well as promote the strengthening of public health policy.

Assessment of study quality
Based on the model presented by Ackley et al. [41], seven levels of evidence were described. Each study was evaluated based on this model, and breaking down the levels of evidence reveals that seven studies (27%) were of Evidence Level III, which include non-randomized, well-designed controlled trials; and 19 studies (73%) were Level IV, well-designed case-control or cohort studies, including cross-sectional designs. According to Detsky et al. [42], who evaluated quality strength of criteria for assigning grades of evidence, seven studies (27%) were Moderate quality, with 19 studies (73%) being of Low Quality (see Appendix 2).

Eligible articles
The final analysis included 26 articles. The most common study designs, cross-sectional designs, were composed of 19 (73%) studies, followed by six (23%) quasiexperimental designs, and finally, one (4%) experimental study. Four interventions were accepted into the review [43][44][45][46]. The range of dates included were 2006-2017, to try to capture the most recent literature available at the time this review was completed. All studies included adolescent age group 13-18, including Middle and High School students. Some studies included ages 6-12, and data were then extrapolated which focused on the age of inclusion for this review. The size of sample varied anywhere between 186 students involved in a high school intervention [45], to US student population age 10-17 across 50 states in the US [47].

Synthesized presentation of research
The research synthesized is presented in Table 4, which summarizes the aspects examined most frequently within the studies. Based on Whittemore and Knafl [48], regarding methodology within integrative literature reviews, the first step in data analysis should be data reduction.
The themes most frequently appearing in the studies included were those examining elements of BMI/ Weight/Obesity, Calorie (Kcal) Consumption, Availability of Competitive Foods in Schools, Strength of Policy, and Other Effects: SES/Race/Ethnicity/Sex. Appendix 2 breaks down the literature components which were significant in the summary of each article accepted for this review. The breakdown consists of aim and objective, design of study, methods and data used in each study, evaluation measures, and level of evidence.

Summary of aspects
The following is the summary of the five aspects most frequently used in the 26 articles for this review.
BMI/obesity/weight Nineteen out of the 26 articles included in the review addressed competitive foods in schools based on BMI/Obesity/Weight as indicators. While the studies were diverse, several examined the associations between competitive food sources in the schools and adolescents weight [16,49,50]. Some studies explored the prevalence of childhood overweight trends before and after policies were implemented, as well as implementation and adherence to nutrition laws in schools [3,26,[51][52][53][54], while others examined the interaction on junk food sold in schools along with genetic components to weight [14]. Wiecha et al. [55], debated whether there was an association between BMI and grade level with intake of SSB's.
Several studies discussed the prevalence of competitive food and beverages in schools and causality towards increase in BMI and weight for adolescents [23,43,56]. A study analysis by Mendoza et al. [57], and Smith & Holloman [45], explored reductions in Kcal consumption by decreasing the energy density in foods and SSB's sold in schools and possible associations with long-term weight loss and maintenance. Fox et al. [58] discussed reductions in daily Kcalories from consumption of energy-dense foods and beverages in school with possible associations to weight of adolescents. Nanney et al. [47] details the varying levels of obesity across the country in various states and regions and any relevance it has on nutrition, physical activity, and education policies within schools. Fletcher et al. [59] examined the effects of soft drink taxes on weight among children, change in BMI, obesity prevalence, or soft drink calorie consumption.
Kcalories consumed or measured Articles which discussed weight through consumption of Kilocalories or Kilocalorie density were in 58% or 15 of the articles reviewed. Interventions assessed by Hartstein et al. [44], evaluated mean daily nutrient sales per student, per item in competitive food venues such as a la carte and snack bars in schools. Cradock et al. [43], measured data on consumption of SSB's by high school students before and after policy change implementation, noting the mean per capita daily calorie consumption for students who consume SSB's. Snelling et al. [46], compared the NSLP and competitive food offerings and purchases in students in 3 high schools, and Smith & Holloman [45], examined the impact of SSB consumption among adolescents in high school settings and the Kcalorie implications it held.
Similarly, other studies which scrutinized adolescent Kcalorie consumption of SSB's in school settings were also performed [55,60,61], including a pilot simulation by Levy & Friend [62], who assessed SSB Kcalories consumed inside and outside of school settings, and the possible effects on an SSB tax. Kakarala et al. [56], explored the mean energy intake from competitive food sources other than a la carte in schoolchildren. Briefel et al. [63], set out to test the hypothesis that children would compensate for eating fewer SSB's and low-nutrient, energy dense foods at school by consuming more of these items outside of school. Other cross-sectional analyzes examined data on Kcalorie consumption of competitive foods in schools before and after policy implementation or beverage tax imposed [51,57,59]. Fox et al. [49,58], examined associations between adolescent's weight status and school food environments, and amount of Kcalories consumed by students from low-nutrient, energy-dense foods attained in school.

Availability of competitive foods in schools
There were 20 articles (77%) which addressed the aspect of availability of competitive foods in schools and how that related to Kcalorie consumption or weight in adolescents. Studies which correlated associations between availability, consumption patterns, and student's BMI-related outcomes were performed by [14,26,49,50,59,60]. Whereas other studies considered the associations between a variety of competitive food sources (vending machines, a la carte, school store, snack bars) in schools with consumption of low nutrient, energy-dense foods [23,55,56,58,63]. Briefel et al. [61], and several others [3,43,52,54,62], explored change in consumption and weight status when policy interventions restricted the sale of SSB's and and competitive food offerings on school grounds. Snelling et al. [46], sought to compare the NSLP with competitive food offerings and the impact availability and type of competitive foods had on adolescent consumption and purchasing behaviors. Nanney et al. [47], examined cross-sectional associations to discover if youth obesity prevalence and comprehensive schoolbased obesity prevention policies were linked. Smith & Holloman [45], analyzed the purchasing patterns of SSB consumption when restrictions on availability were implemented in school settings.

Analysis of policy
Fourteen studies (54%) discussed elements involving competitive food and beverage, nutrition, and physical education policies. Addressing over-weight prevalence before and after initiation of competitive food and beverage policies was observed in a number of studies [51,53,57]. Studies examined comparisons between states with competitive food and beverage policies in place versus those states with less stringent or no policies in place [3,14,26,43,63,64].
Evaluating different competitive food and beverage policies within schools around the country and their association with adolescent obesity were also examined [47,62] and, Nanney et al. [47], and Riis et al. [52], add analysis on physical education policies in addition to school nutrition policies. Taber et al [54], analyzed state school meal laws and their associations with weight status in adolescents. Fletcher et al. [59], and Levy & Friend [62], also considered effects on adding SSB taxes to predict daily Kcalorie reductions in adolescents.  [14,16,23,26,43,45,51,57,63]

Discussion
After examining all 26 articles which met all criteria for this review, it was clear that adolescent obesity is a serious public health concern. The studies analyzed and performed within this review attempted to quantify where the problems lay, with a main assumption being that since adolescents spend an average of six or more hours per day and 180 days out of each year in school settings, the school environment might hold an inimitable influence upon the diets and eating behaviors of US schoolchildren [49].
One consistent concern was the pervasiveness of competitive foods, considered FMNV in US schools [14]. Thus far, there has been scant evidence to support the associations between school food environments with eating behaviors and weight/BMI of adolescents [49]. It is with a unified consensus therefore, that all studies in this review attempted to dissect this hypothesis, with varying results. Therefore, as 'a final step of the data analysis in an integrative review. . .the synthesis of important elements or conclusions of each subgroup' . . .is gathered. 'into an integrated summation of the topic or phenomenon,' [48, p.551], which is represented in the following text.

Positive association between competitive food and beverage policies and BMI/weight
Cradock et al. [43], found that the decline in consumption of SSB's after the policy change in Boston Public Schools relates to approximately 45 Kcals per day, where a 45 Kcal/day reduction could potentially cut the 25% to 40% of total excess Kcalories. According to Briefel et al. [60], among students who consumed SSB's at school, their energy intake over their entire day was approximately 229 Kcal higher than students who did not consume SSB's at school. According to Wiecha et al. [55], the number of items purchased at school vending machines is directly associated with SSB intake and purchase.
Fox et al. [58], discovered that in middle schools and high schools, students got 171 and 219 Kcalories from low-nutrient, energy-dense competitive foods. Fox et al. [58], discovered at the middle school level, the availability of low-nutrient, energy-dense foods in vending machines in or near the foodservice area was positively associated with a higher BMI z score. And Hartstein et al. [44] noted that two Texas schools in their study showed a reduction in Kcal density from 277 to 216, however, other reductions were modest (1 to 12 Kcal per item sold). California students exposed to more stringent school nutrition policies consumed a lower proportion of their Kcalories at school, indeed, consumed less for every measure examined, compared with students in other states [64].
Results revealed that energy density significantly declined to 2.1 + 0.78 Kcal/g in the middle school study by Mendoza et al. [57]. Nanney et al. [47] found that Food Service and Nutrition (FSN) policy groupings with the strongest associations to youth obesity are policies which pertained to competitive foods and food service standards. After adjusting for factors, nutrition policies addressing competitive foods in other venues, food service director qualifications, and BMI screening remained positively and significantly related to the odds of obesity in children, as revealed by Riis et al. [52].
Levy and Friend [62], postulated that a $0.01 tax per ounce is predicted to reduce SSB consumption by 40.3-54.2 Kcal/day. Taber et al. [54], found that the unadjusted frequency of obesity was 11% higher in students who received free/reduced price lunches at school compared with students who did not purchase lunch at school. Adolescent students gained 0.44 fewer BMI units when they were exposed to specific, consistent, required competitive food laws from 2003 to 2006 than students who were not exposed to the same requirements, according to Taber et al [3].
According to Smith & Holloman [45], students who participated in their intervention reduced both their daily consumption of SSB's as well as the number of days per week they consumed these beverages (1 serving = 150 Kcal/day). Similarly, an intervention by Snelling et al. [46], discovered that competitive food menus tend to offer foods which lack in nutrients and have higher energy densities. Kakarala et al. [56] found that the total sugar intake was higher among students who consumed competitive food/ beverages when excluding a la carte items.
Other effects are significant Briefel et al. [61], discovered when accounting for other effects such as age and race/ethnicity, non-Hispanic whites would save more total Kcalories than Hispanic middle school students (234 Kcal/day vs. 184 Kcal/day). Briefel et al. [63], also found that being non-Hispanic African American was associated with greater caloric intake of sweetened beverages in high school, by 42 Kcal, and that being Hispanic or non-Hispanic African American was associated with a 47 and 70 Kcal greater intake from low-nutrient, energy-dense foods, than for non-Hispanic whites, and females consumed 46 Kcal less from low-nutrient, energy-dense foods.
Bauhoff et al. [51], also discovered that there was a consistent decrease of soda consumption for females, when SSB reductions were put into place. Wiecha et al. [55], discovered that boys drank significantly more SSB's than girls did, and that SSB intakes were higher among Hispanics vs non-Hispanics, and African Americans vs non-African Americans. Similarly, Hennessy et al. [26] noted in their study that adolescents who were overweight/obese were more often younger, have younger parents, non-Hispanic black or Hispanic, male, less vigorously active, have a TV in their bedroom, not live in a 2parent family, and reside in a poor household. Park et al. [23] determined that in the proportions of students buying lunch from vending machines, non-Hispanic black race/ethnicity, Hispanic ethnicity, older age, and smoking were significantly higher.
Relative to other groups, fifth-grade girls in Los Angeles experienced the largest change in overweight trends, however, in the rest of California, the lower rate of increase in overweight was significant among fifthgrade boys and seventh graders [53]. In the study analyzed by Taber et al. [64], they also found that the California sample had a greater proportion of Hispanic students (76.6%) than other states that were in the sample (14.7%). However, the results were very similar when restricting the analysis to Hispanic students on consumption of caloric content of competitive foods.
Mendoza et al. [57], determined that the reduction in energy density was significantly moderated by SES, and that low-SES schools had a higher proportion of males (66%), and Hispanic students. According to Taber et al.'s [3] breakdown, it showed how states that had no 2003 laws had a relatively low proportion of students who were non-Hispanic black (9.0%) and low proportion of students in the lowest SES quintile (15.7%). However, they found that states with weak 2003 laws had a relatively high proportion of students who were Hispanic (28.0%), and who were in the lowest SES quintile (23.8%). Yet, Kakarala et al. [56] determined that competitive food use did not differ significantly between adolescents who qualified for subsidized school meals and those who did not, signifying that the cost of vended items was most likely priced so that children of any income level could buy them.
No strong effect between school food environments on BMI/weight found overall Anderson and Butcher [14], found a strong literature base which revealed robust correlations regarding a genetic component to weight, suggesting in their study that while school food policies have no affect on most students' weight, policies that increase access to junk foods and SSB's in school could be a contributing factor for those with a genetic susceptibility to weight gain. Similarly, Bauhoff et al. [51], determined there were positive effect results for obese students considering they might be more apt to purchase and consume in-school, therefore being more influenced by any nutrition policies. However, no significant statistical results regarding most student's weights and any competitive food policies were found.
Van Hook and Altman [16], also suggest that weight during early adolescence is strongly shaped by how heavy these adolescents were in their younger years.
Fletcher et al. [59, p. 1064], discovered that 'neither vending machine restrictions nor soft drink taxes will lead to noticeable weight reduction in children', noting that according to their analysis, adolescents consumed just as many soft drinks in schools with vending machine restrictions as in schools without restrictions. And Hartstein et al. [44], determined that when all schools in their intervention were included, there were no significant change in number of Kilocalories sold per student between week 1 and week 6.
Briefel et al. [61], found positive results for SSB consumption in out-of-school settings versus inschool settings, suggesting that while reducing SSB consumption in schools can be beneficial (8 Kcal/ day), a greater benefit would come from reducing SSB consumption outside of school settings (145 Kcal/day). And Park et al. [23], discovered no significant association in daily mean SSB consumption from vending machine on BMI status among middle school students in the US.
Nanney et al. [47], found that there was no significant correlation between Weight Assessment (WA) policies and youth obesity prevalence. Results from the analysis by Terry-McElrath et al. [50], showed that, after inclusion of the percentage of high school students eligible for free or reducedprice lunch (FRPL), all associations between student overweight and obesity and school food environment were not significant.
Van Hook & Altman [16], similarly discovered that their results demonstrated how changes in competitive foods sales in schools were not associated with changes in children's percentile BMI. The negative relationship between obesity prevalence and school meal environment polices in high schools was constant across cross-sectional and policy change analyzes, as was determined by Riis et al. [52]. Based on the findings from Fox et al. [49], no associations between school food environments and practices, BMI z scores, or the likelihood of obesity were significant among high-school students.

Measuring variables
A variety of different measuring variables were examined and used in each article (see Appendix 2), with some of the more common variables including 13 articles (50%) examining BMI/weight association with exposures to competitive food and beverage policies in schools, six studies (23%) compared students in school districts to those in a control group or other district. Seventeen (65%) of the 26 studies utilized questionnaires, checklists, or menu data for comparative analysis to BMI/weight/Kcal consumption, and 11 articles (42%) looked at outcome measures of per capita total Kcal consumed per day as a measuring variable. Similarly, 11 (42%) of the 26 articles analyzed in this review used comparisons within student's eating patterns as a particularly interesting measuring variable.

Method of analysis
Each article included in this review used one of four categories of analysis; Factor Analysis, Associations, Comparisons, and Descriptive Analysis (see Table 5). Seven (27%) of the 26 studies used Factor Analysis, whereas 11 (42%) of the articles applied an Association Analysis. Eleven studies (42%) employed Comparative Analysis, and lastly, most (16, or 62%) studies relied on Descriptive Analysis.

Limitations in the literature
According to this review of literature, with 26 articles included in the final analysis, each presented their limitations with regards to robustness in research. Nineteen articles (73%) were based on cross-sectional study design, which can be important in determining correlational findings, however, are not able to imply causality. It is also important to note that fifteen of the articles (58%) were based on self-reported data, while common in these types of research studies on diet and BMI/weight, could potentially alter prevalence within the data due to under or over-reporting of food and beverage consumption and purchase recall or weight/height measurements.
Five of the articles (19%) in this review mentioned that the data they were utilizing was as current as was possible to obtain, however, was still outdated and newer more vigorous research was needed. Two articles (8%) discussed the limitations in data during the pre-intervention and post-intervention phases, which would have benefited the results. Limited data collection across a variety of elements within studies was brought up in 17 of the 26 articles (65%), with most acknowledging the lack of ability to test for every factor possible to attain the strongest most unbiased results. And finally, five of the articles (19%) stated that their results were not necessarily applicable to other children, either older or in a different setting.

Methodological limitations
There were several limitations with this integrative systematic review, one being that most of the literature available are cross-sectional studies, which are unable to define causality. Another limitation might be that only three databases were searched, therefore data was only obtained from CINAHL, MEDLINE, and EMBASE, and other literature is potentially available which was not found in this review. Generalizability is scattered, with some studies being relevant over the entire adolescent age range, and others merely pertinent to middle school, or high school aged students only. Finally, due to feasibility and time, double measures of cross-checking the data and inclusion criteria was unavailable. A pilot was done with two reviewers, yet only one researcher could complete the rest of the data extraction and analysis for this review. However, intra-rater reliability was ensured through research re-analyzation of all selected data.

Conclusions
The results were inconclusive, with many studies unable to definitively prove that competitive food sources in the schools are causal towards BMI/weight of adolescents, and much of the literature revealing that there was indeed, no direct association between competitive foods in schools and BMI/weight. There were studies which did discover associations between weight and competitive foods sold in schools, but they were limited and did not have robust results.
Stronger policies overseeing competitive foods in schools proved to help in some significant ways, revealing that monitoring the sales and availability of FMNV in schools can only benefit students, considering how only a few Kcalories per day can make a difference long term. The most common theme throughout the literature was how even small fluctuations can create large changes in society, with easy access and availability towards unhealthy foods in schools impacting adolescents in a profound way, therefore, there needs to be an awareness and forethought regarding the future of our youth. It is apparent from the literature that policy makers need to pay extra attention to the more vulnerable groups such as those with a genetic susceptibility, lower socio-economic status, as well as examining more closely minority groups, and gender differences.
It was the definitive goal of this study to explore the literature currently available to answer the question of what ways competitive foods in school settings impacted adolescents BMI/weight. The hope was to bring to light the importance of how school environments are a strong influence and possibly a factor in the epidemic of adolescent obesity in America today. This literature review attempted to analyze the existing data systematically, thoroughly, and without bias. It is the opinion of this review that this was achieved. The literature which was examined, as well as the results of the studies analyzed, could very well be generalized to other countries other than just the US. Replicability of this study is ensured, and detailed throughout this review. Adolescent obesity is a growing epidemic worldwide, and a call for stronger research is imperative for this global public health problem.

Acknowledgments
I would like to thank my thesis advisor Dr. Lina Magnussen of the School of Public Health at Lund University for her oversight and assistance.

Author contributions
Kirsten Sildén was the principal researcher, writer, and editor of the study for all drafts of the paper.

Disclosure statement
No potential conflict of interest was reported by the author.

Ethics and consent
This review is comprised of articles which included established ethical clearance procured for the research, or if that was not apparent, the informed consent process was described clearly within the article thus implying that the statement of ethical consent need not be published directly, merely accompanying the submission of the article to prove the ethical quality. Therefore, this review examined and included articles if it was clear that the studies were performed in an ethical manner.

Funding information
None.

Paper Context
Obesity affects more than one-third of adolescents (age [13][14][15][16][17][18] in the Unites States. The economic burden and health implications are staggering. This review compiled the most recent research and examined competitive foods sold in schools to see if they are a contributing factor on adolescent obesity. This review can inform policy makers, educators, parents, and community about the obesity crisis facing our youth and the need for more rigorous research and policies to help fight it.
ORCID Kirsten E. Sildén http://orcid.org/0000-0003-4116-977X 4. Descriptive analyzes of calories and added sugars from SSB's and flavored milks by location and replacement simulation were completed using SAS, incorporating appropriate sampling weights for schoolchildren and design effects caused by the SNDA-III complex sample design. 5. Descriptive analyzes used.
'A t test of mean differences between population subgroups was conducted to determine which groups would benefit most from beverage changes.' Fletcher et al. [59] "In this paper, we outline important concepts related to these policies.
We present new empirical evidence of the likely effectiveness of school vending machine restrictions and increasing taxes on soft drinks on reducing obesity rates in children.
We conclude by discussing how the effectiveness of these policies might be increased." Non-experimental design: Observational; Cross 2. None provided.

3.
'Using data on soft drink consumption and vending machine restrictions from the nationally representative ECLS-K. We focused on the 5th grade (2004) and 8th grade (2007) survey waves, to measure soft drink consumption and accessing data from (NHANES) III (1988)(1989)(1990)(1991)(1992)(1993)(1994) and IV (1999IV ( -2006. Individual foods and beverages from the dietary recall were coded and classified using the USDA's Survey Nutrient Databases. We examined the effects of soft drink taxes on soft drink consumption and weight among children and adolescents using data from (NHANES) III (1988)(1989)(1990)(1991)(1992)(1993)(1994) and IV (1999IV ( -2006.' 4. Soft drink consumption and accessing data; Indiv. foods and beverages were coded and classified; soft drink consumption and weight were examined.