The eating habits of college students are generally poor(Reference Burke, Reilly and Morrell1, Reference Racette, Deusinger and Strube2) and healthful eating is often not a high priority for them(Reference Strong, Parks and Anderson3). For example, only one in twenty students eats the recommended five or more daily servings of fruits and vegetables(4). Poor diet is negatively correlated with weight status in college students(Reference Racette, Deusinger and Strube2, Reference Adams and Rini5), with more than one-third being overweight (BMI = 25·0–29·9 kg/m2) or obese (BMI ≥ 30·0 kg/m2)(6). Obesity in late adolescence predicts obesity later in life(Reference Kindblom, Lorentzon and Hellqvist7), thereby increasing the risk for CVD, hypertension and diabetes(Reference Mokdad, Marks and Stroup8). Research evidence indicates that a complex interplay of personal and environmental factors influence students’ dietary habits and weight. College students select foods based on taste, time, schedule, convenience, cost, physical/social environments, and some for health or weight control reasons(Reference Greaney, Less and White9, Reference Horacek and Betts10).
Ecological models for health behaviour change such as PRECEDE–PROCEED(Reference Green and Kreuter11) posit the importance of environmental influences upon dietary behaviours and weight status. University campus dining environments can either promote or inhibit healthy eating and weight control choices by young adults. For instance, female students residing in dormitories with on-site dining gained more weight during their freshman year than counterparts without on-site dining(Reference Kapinos and Yakusheva12), and freshmen living on campus gained more weight than students commuting from home(Reference Pliner and Saunders13). However, students living on campus ate more fruit, vegetables and dairy products than students in off-campus apartments(Reference Brunt and Rhee14). Beyond the food available in the campus dining environment, other factors can facilitate students’ selection of healthful foods. Students who utilize point-of-selection nutrition labels in dining halls often report improved food choices(Reference Driskell, Schake and Detter15) and reduced energy content of meals(Reference Chu, Frongillo and Jones16, Reference Freedman17) compared with those not using nutrition labels. College students’ eating habits improved when exposed to point-of-selection messages promoting taste, satiety, body leanness, energy value and health(Reference Peterson, Duncan and Null18). Price reduction strategies also encouraged selection of healthier food items(Reference French and Wechsler19, Reference Michels, Bloom and Riccardi20). As an ‘all-you-can-eat restaurant’, dining halls are poised to influence, either positively or negatively, students’ food choices by the array of foods offered and the techniques used to promote them.
Although there is evidence of a relationship between the campus food environment and students’ dietary habits and weight(Reference Kapinos and Yakusheva12–Reference Michels, Bloom and Riccardi20), there is insufficient evidence regarding the causal linkages(Reference Giskes, van Lenthe and Avendano-Pabon21) and a lack of clarity on how to define and describe the campus dining environment. Students no longer eat their meals only in the dining halls; they have a variety of options on and off campus from which to choose (union, food courts, snack bars, restaurants, vending and stores). Although grocery stores near college campuses provided a wide array of healthful alternatives, convenience stores on and near campus(Reference Horacek, Erdman and Reznar22) and campus vending machines(Reference Byrd-Bredbenner, Johnson and Quick23) provide few healthful foods and are more obesogenic. Numerous studies describe and evaluate food store and restaurant environments(Reference McKinnon, Reedy and Morrissette24–26), yet little attention has been given to college campus dining environments. A tool is necessary to describe and evaluate varying campus dining environments. This tool and the evaluation it enables are the essential preliminary steps for understanding baseline campus environments and for guiding intervention programme developers and policy makers in their efforts to create campus dining environments fully supportive of healthy eating. Thus, the purpose of the present study was to develop and pilot-test a tool to evaluate the restaurant and dining venues on and near post-secondary campuses at geographically diverse institutions varying in student enrolment size.
Methods
A team of researchers collaborating on the US Department of Agriculture Multistate Research Project entitled ‘Promoting Healthful Eating to Prevent Excessive Weight Gain in Young Adults’ assessed the dining environment on thirteen university campuses, one residential post-secondary training programme and one technical college. Steering committees composed of diverse teams of stakeholders (i.e. students, faculty, staff, administrators and community members) at each institution guided the research process using the principles of community-based participatory research (CBPR); more specifically, the PRECEDE–PROCEED model(Reference Green and Kreuter11). The present study comprised one portion of the full CBPR project. Each campus team established a geographic boundary extending at least 1·5 miles from its campus edge. This boundary definition was selected because many students live on or near campus, and the 1·5-mile radius was used to evaluate other aspects of this PRECEDE–PROCEED environmental assessment(Reference Horacek, White and Greene27). Given this was CBPR and not a standardized public health evaluation, each team identified the restaurants and on-campus dining venues within the boundaries that were most frequented by students and hence relevant for evaluation. Approximately 20 % of the restaurants frequently used by students were located beyond the perimeters and were also included to ensure adequate representation of dining venues frequented by students. Institutional Review Board approval was not necessary because no human subjects were involved.
Instrument
Although nearly 150 different measures are available to evaluate accessibility, availability, affordability and quality of the food or restaurant environment(Reference McKinnon, Reedy and Morrissette24–26), at the time of the present study none of these fully captured the unique mix of school, worksite, home and neighbourhood environments defining typical post-secondary institutions. Thus, the reliable Nutrition Environment Measures Survey for Restaurants (NEMS-R)(Reference Saelens, Glanz and Sallis28) was adapted for the present study to make it more applicable to the college-age population and post-secondary institution campuses. This tool was also adopted because the Nutrition Environment Measures Survey for Stores (NEMS)(Reference Glanz, Sallis and Saelens29) was used for the evaluation of stores on and near campuses(Reference Horacek, Erdman and Reznar22). All modifications were guided by a literature review(Reference Brunt and Rhee14, Reference Oldenburg, Sallis and Harris30, Reference Beydoun, Powell and Wang31), an analysis of the campus environment(Reference Erdman32), stakeholders and study researchers. The modified surveys were reviewed by five experts for face validity, comprehensiveness and clarity, and subsequently were pilot-tested at two north-eastern US college campuses. The surveys were refined based on experts’ comments and pilot testing.
The NEMS-R is designed to evaluate restaurants, including sit down, fast food and fast casual(Reference Sperber33) restaurant types. This instrument assesses facilitators of (e.g. nutrition information) and barriers to (e.g. super-size portions) healthful eating, and the availability of healthy entrées, entrée salads, fruits, vegetables and beverages. To be rated as ‘healthy’, entrées and entrée salads must meet standards(34) for energy, total fat and saturated fat.
Off-campus restaurants
For off-campus restaurants, the NEMS-R was updated to reflect current eating trends by deleting the assessment of ‘low-carb’ items and adding availability of vegetarian menu items(Reference Craig35). In addition, assessment of children's menus was omitted because this was not the target population of interest.
On-campus dining venues
On-campus dining facilities frequently have features unlike those of typical restaurants. Thus, the NEMS-R was modified to capture these unique features, resulting in the creation of the Nutrition Environment Measures Survey–Campus Dining (NEMS-CD). As shown in Table 1, a detailed review of salad bars (i.e. numbers of fresh, low-fat and no-sugar added fruit and vegetables, low-fat protein sources, and low-fat/fat-free dressings) was added. Because on-campus dining facilities supply most or all of the food consumed by college students residing on campus, the NEMS-CD also was expanded to incorporate a broader array of healthy eating constructs, such as those in the Dietary Guidelines for Americans(34), including availability and variety of 100 % fruit juices, whole grains(Reference Newby, Maras and Bakun36, Reference Rose, Hosig and Davy37), dairy alternatives (e.g. soya milk), cereals (all and those with low sugar/high fibre)(38) and vegetarian options(Reference Craig35). Evaluation of pricing was adjusted to account for dining contracts.
Comparison of dining environment constructs for assessing restaurant (NEMS-R) and campus dining (NEMS-CD) venues
*Adapted from Nutrition Environment Measures Survey for Restaurants (NEMS-R)(Reference Saelens, Glanz and Sallis28).
†Low-carb not assessed, given it is a less popular dietary management strategy; children's menus were not evaluated given the target population was the college environment.
‡Added to the original NEMS-R instrument.
§Nutrition Environment Measures Survey–Campus Dining (NEMS-CD).
NEMS-CD was used to evaluate on-campus eating venues, including dining hall cafeterias, student union/food courts and snack bars/cafés. A dining hall serves students with meal plans and their guests. A student union/food court is open to students, faculty/staff and the public, typically on a pay-per-item basis, with some venues also accommodating meal plans. Snack bars/cafés and other small dining venues on campus generally charge on a pay-per-item basis and are open to the public.
Scoring
The scoring procedures for NEMS-R and NEMS-CD followed the original NEMS-R protocols(Reference Saelens, Glanz and Sallis28). Sub-scores are described in Table 1. In brief, a total dining environment quality score was calculated for each dining establishment by summing the scores for each construct assessed. Scores for restaurants could range from −24 to +66 points. On-campus dining scores could range from −29 to +100 points; the additional points resulted from the survey modifications. Positive points were awarded for availability of nutrition information and healthful options and for a greater selection of foods rated as healthy. Negative points were given for barriers to healthful eating, such as signs encouraging overeating (e.g. ‘Get five sandwiches for only…’) or an extra charge for a shared entrée.
Data collection
Sample
Using stakeholder input, each post-secondary institution selected a sample that represented 20 % of all off-campus restaurants within its geographic boundary, 20 % of all (or at least one) campus dining halls, 20 % (or at least one) of all campus student union/food courts and 20 % of all campus snack bars/cafés. Some establishments slated for evaluation were eliminated if the data collector found that they did not meet criteria for assessment (e.g. served coffee only and no assessable food items sold). All data collectors carried a letter from the lead researcher at their institution that described the study to present upon request by dining establishment employees. Data were collected from December 2008 to March 2009 and May to June 2011.
Data collector training
Standard NEMS-R instructions were modified to include the adaptations described above and expanded to include instructions for the NEMS-CD. One investigator attended a NEMS training(Reference Honeycutt, Davis and Clawson39) in October of 2008 (www.cdc.gov/prc/training/practitoners/nutrition-environment-measures-survey.htm) and modified the materials (manual, data collection/entry tools and PowerPoint presentation) used to train the data collectors on each participating campus. Each data collector reviewed all materials and then participated in an online training seminar lasting ~2 h. After the training, data collectors practised with the modified tools at restaurants and campus dining venues not included in the study sample. A trained campus coordinator at each institution reviewed practice data collection results and worked with data collectors to clarify any deviations from the instrument scoring standards.
Prior to data collection, inter-rater reliability (IRR) was established on each campus by having each data collector independently evaluate the same establishment on the same day. The IRR scores were derived by comparing the number of constructs that were scored similarly by the data collector and the campus coordinator. Additional dining venues were assessed until all data collectors at an institution achieved an IRR greater than 80 %. Once this agreement level was reached, data collectors began assessing the sample of dining venues.
Data analysis
The SPSS statistical software package version 18·0 was used for data analysis, which was completed in winter 2011. Institution size affects the number and variety of dining venues and potentially the extensiveness of services provided. Thus, to permit meaningful comparisons, institutions were categorized into tertiles using student enrolment statistics as a proxy for institution size. Institutions with ≤15 000, 15 001–29 999 and ≥30 000 students were categorized as small-sized, medium-sized and large-sized institutions, respectively. Ranges, means and standard deviations for construct scores were calculated. ANOVA with post hoc Tukey's B and T tests were used to determine whether significant differences occurred between mean scores of on-campus and off-campus dining venues, and between institutions, geographic regions, institution sizes and data collection periods (i.e. 2009 v. 2011). To permit comparisons of all restaurant-type facilities both on and off campus including on-campus snack bar/cafés and all off-campus restaurants, only the NEMS-R constructs were considered. The percentages of dining establishments with healthy foods, facilitators and barriers were calculated and compared using χ 2 tests. Significance was set at P < 0·05.
Results
About half of the fifteen post-secondary institutions participating in the present study were from the Northeast region of the USA (n 8), one-third were from the Midwest and the remaining two were in the Southeast. There were no significant differences in results between the two data collection periods, so all data were combined. Similarly, no differences were detected by institution or region. Table 2 describes the characteristics of participating institutions and the number of dining establishments assessed. Small-sized institutions were located in significantly smaller communities than medium-sized and large-sized institutions as indicated by the differences in community population density; however, there were no differences in community median income or percentage of students living on campus.
Institution demographics and distribution of dining venues evaluated; fifteen US post-secondary institutions, 2009–2011
a,b,cMean values within a row with unlike superscript letters were significantly different as determined by pairwise comparisons using Tukey's B post hoc test (P < 0·05).
*Student enrolment data collected with NEMS (Nutrition Environment Measures Survey) data.
†Community population density is the population of the city within which the institution is located/city area in square miles according to 2010 census data.
‡ANOVA main effects indicate significant differences by institution size (P = 0·019).
§Pearson χ 2 determined a significant difference in the distribution of restaurants assessed by institution size: P = 0·002; χ 2 = 16·75, df = 4.
A total of sixty-eight campus dining venues were assessed. There were no differences by institution size in the number of campus dining venues evaluated; however, small-sized institutions evaluated significantly more fast-food restaurants (50 %), whereas large-sized institutions had significantly more sit-down restaurants (47 %). In general, dining halls had significantly higher (more healthful) scores than campus snack bars/cafés on all campus dining environment constructs except Facilitators of Healthy Eating (Table 3). There were few differences between dining hall scores by institution size, although medium-sized institutions had significantly higher (more) Healthy Side Dishes/Salad Bar scores than small-sized institutions. Student unions at large-sized institutions were less healthy, with significantly lower scores than smaller institutions in the categories of Total NEMS-CD, Healthy Side Dishes and Healthy Beverages. There were no differences in snack bars/cafés by institution size.
Comparison of campus dining environment construct scores by on-campus dining venues and institution size; fifteen US post-secondary institutions, 2009–2011
a,bMean values within a column (for all institutions combined) or within a row (for by institution size) with unlike superscript letters were significantly different as determined by pairwise comparisons using Tukey's B post hoc test (P < 0·05).
*ANOVA main effects significance level comparing campus dining environment total score by dining venues.
†ANOVA main effects significance level comparing campus dining environment sub-scores for each venue by institution size.
‡Healthy entrées (main dishes, main dish salads and low-fat/fat-free dressings; possible range = 0–12).
§Healthy sides dishes/salad bar (non-fried vegetables, no-sugar added fruit, vegetarian, wholegrain items, baked chips, salad bar; possible range = 0–35).
∥Healthy beverages (low-fat and non-fat milk and alternatives, 100 % fruit juice, diet and low-calorie soda; possible range = 0–18).
¶Facilitators of healthy eating (nutrition information, healthy signs, reduced portions; possible range = 0–30).
**Barriers to healthy eating (unhealthy signs, large portions, all-you-can-eat; possible range = −15 to 0).
††Price (dining hall contracts, salad bar by weight, combo meals, portion and healthy meal pricing; possible range = −14 to 5).
A total of 175 restaurants with a fairly equal representation across types were evaluated. There were no significant differences in NEMS-R total score or sub-scores by institution size for the restaurants; therefore Table 4 compares differences between the types of restaurants, including on-campus snack bars/cafés. For this analysis, only constructs common to both the NEMS-R and NEMS-CD were used to allow comparability. Overall, the various dining venues were similar. There were no significant differences in healthfulness among restaurant types as indicated by the lack of differences in scores for Total NEMS-R, Healthy Entrées and Barriers to Healthy Eating. However, NEMS-R sub-scores indicated that on-campus ‘restaurants’ or snack bars/cafés had more Healthy Side Dishes, and fast-food restaurants had more Facilitators of Healthy Eating. Both fast-food and on-campus restaurants offered more monetary incentives to buy meal deals (increased quantity for limited price) as indicted by the significantly lower Price scores.
Comparison of restaurant dining environment construct scores by restaurant type; fifteen US post-secondary institutions, 2009–2011
a,bMean values within a row with unlike superscript letters were significantly different as determined by pairwise comparisons using Tukey's B post hoc tests (P < 0·05).
*On-campus restaurants are the snack bars/cafés; from the campus dining environment survey only the restaurant dining environment constructs were used for this analysis to permit comparisons across restaurant types.
†ANOVA main effects significance level comparing scores by restaurant type.
‡Healthy entrées (main dishes, main dish salads and low-fat/fat-free dressings; possible range = 0–9).
§Healthy sides (non-fried vegetables, no-sugar added fruit, vegetarian, whole-wheat bread, baked chips; possible range = 0–15).
∥Healthy beverages (low-fat and non-fat milk, 100 % fruit juice, diet and low-calorie soda; possible range = 0–9).
¶Facilitators of healthy eating (nutrition information, healthy signs, reduced portions; possible range = 0–30).
**Barriers to healthy eating (unhealthy signs, large portions, all-you-can-eat; possible range = −15 to 0).
††Price (combo meals, portion and healthy meal pricing; possible range = −9 to 3).
Table 5 shows the detailed percentages of healthy food items offered, facilitators of and barriers to healthy eating, and pricing factors available at each type of dining venue. There were no significant differences among venues for the presence of healthy salads (available at 21–41 % of venues), 100 % juices, nutrition information or healthy entrées identified on the menu, signs showing healthy menu options, whether unhealthy eating or overeating was encouraged, or whether customers were charged to share entrées. Significantly more dining halls offered more healthy food items than other dining venues. More than 90 % of dining halls offered low-fat and non-fat milk and whole-wheat bread (v. 50 % of student unions, the second most likely to offer whole-wheat bread). In addition, more than 80 % had vegetarian options, a salad bar and no-sugar added fruit (v. 62 % of campus snack bars, the second most likely to serve no-sugar added fruit). More than 70 % of dining halls had healthy entrées, non-fried vegetables (followed by 60 % of sit-down restaurants), special requests encouraged and signs encouraging healthy eating, but also ‘all-you-can-eat’ pricing. Student unions had the highest percentage of locations offering baked chips (55 %) and point-of-purchase information that encouraged healthy eating (45 %). Fast-food restaurants had a significantly higher percentage of nutrition information available at the point of purchase, but also encouraged large portions (44 %) and combo pricing (57 %); student unions were a close second with 55 % having combo pricing.
Availability of dining environment constructs by dining venue; fifteen US post-secondary institutions, 2009–2011
POP, point of purchase.
*On-campus restaurants are the snack bars/cafés.
†Pearson χ 2 statistic used to assess significance in distribution between each dining venue for each construct.
Discussion
The on-campus dining environment is unique in that it has multiple competing goals related to feeding students. Some of these include responding to young adult's food preferences, food service cost-effectiveness and efficiency, and students’ health considerations(40, 41). Dining halls provided the greatest variety of healthy entrées, side dishes and beverages, but also had the most barriers to healthy eating (i.e. ‘all-you-can-eat’ pricing) as compared with student unions and snack bars. Facilitators of healthy eating, such as point-of-purchase education, could be improved to counter these barriers. Although facilitators of healthy eating did not differ significantly among the on-campus dining venues, they do help students make more healthful decisions(Reference Driskell, Schake and Detter15–Reference Peterson, Duncan and Null18, Reference Buscher, Martin and Crocker42, Reference Kolodinsky, Green and Michahelles43).
Based on the current findings, a variety of dining options are available for students on and near campus with no substantive differences in healthful choices among restaurant types. There were no significant differences in main dish salads, 100 % juice, nutrition information or signs for healthy menu options, and whether unhealthful eating or overeating was encouraged. Total NEMS-R scores for all restaurant types were only a fraction of the total maximum points (66), implying that unhealthful dining environments were prevalent.
A comparison of the findings reported here with those for the original implementation of NEMS-R in Atlanta, GA(Reference Saelens, Glanz and Sallis28) revealed that a higher percentage of the present sample of restaurants offers no-sugar added fruit, healthy entrée salads and entrées, but similar trends were evident in the offering of non-fried vegetables, baked chips and diet soda. In addition, more restaurants in the present sample facilitate healthy eating by identifying healthy entrées and providing nutrition information and reduced portion sizes. However, more restaurants in the present sample also provided more barriers to healthful eating by offering larger portion sizes. A higher percentage of sit-down restaurants in the current survey encouraged unhealthful eating or overeating and none encouraged healthful eating. The changes in fast-food restaurants are noteworthy; this current sample displays more healthful offerings and signage encouraging healthy eating. Compared with the original NEMS-R implementation, twice as many fast-food restaurants offered whole-wheat bread and 100 % fruit juice; three times as many encouraged healthy eating; and 50 % fewer encouraged unhealthy eating. Although 59 % of fast-food restaurants had combo pricing, this percentage is lower than previously reported. Some of these positive findings may result because nutrition information has become more widely available since the original NEMS-R research and/or due to the differences between Atlanta neighbourhoods and communities and post-secondary institutions. Despite these improvements, additional changes could help consumers choose healthier meals when eating out. Restaurant owners need to be encouraged to provide healthful foods and nutrition information. Although legislative efforts have forced large, franchised operations to provide healthy offerings and nutrition information, this environmental support has not been required of smaller and independent restaurants(Reference Vadiveloo, Dixon and Elbel44–Reference Roberto, Schwartz and Brownell46).
On-campus dining venues offer more healthy options than off-campus venues; however, there were some differences in the healthfulness of the on-campus dining environment according to institution size. Bigger was not necessarily better, possibly indicating differences in how resources are allocated to dining services. In the present study, large-sized institutions’ student union/food courts had significantly lower scores and similar trends for dining halls on a number of the dining environment constructs. A wider sample of comparison data for campus environments is necessary. The newly developed NEMS-CD modified from the NEMS-R, along with the campus food environment assessment(Reference Horacek, Erdman and Reznar22), can be added to the instruments designed to evaluate the college nutrition environment(Reference Freedman47, Reference Minaker, Raine and Cash48). It is imperative that evaluation of the campus food environment across varied post-secondary institutions continues given the environment's influence on young adults’ weight(Reference Kapinos and Yakusheva12–Reference Brunt and Rhee14, Reference Giskes, van Lenthe and Avendano-Pabon21) and dietary intake(Reference Brunt and Rhee14, Reference Pei-Lin49, Reference Chapman, Melton and Hammond50). Perception of the food environment is related to intake(Reference Inglis, Ball and Crawford51) and point-of-purchase information presents an opportunity to positively influence students’ food selections(Reference Driskell, Schake and Detter15–Reference Freedman17, Reference Buscher, Martin and Crocker42). Pricing strategies to improve student behaviours and health outcomes can be effective and should be considered. For example, reducing the price of fruit and salad by half resulted in a threefold increase in purchases(Reference French and Wechsler19), and a 20 % price reduction coupled with a promotional campaign resulted in modest increases in the purchase of health-promoting foods and a slight decrease in the selection of less-healthy foods(Reference Michels, Bloom and Riccardi20).
Acknowledging that campus food environments support healthful food behaviours that can help prevent obesity is critical to achieving sustainable improvements. Environmental and policy changes take time to implement and come with challenges(Reference Story, Nanney and Schwartz52). The likelihood for effective, sustained changes is increased when community-based participatory approaches are used(Reference Green and Kreuter11) and key stakeholders – students, health advocates, campus staff, local store and restaurant owners, local food producers – are involved throughout the process. Resistance to change may be addressed effectively by making gradual and modest changes in healthy food offerings, providing locally produced foods (www.FarmtoCollege.org) and adding facilitators, such as point-of-purchase education and pricing strategies that promote intake of healthier food items(Reference Driskell, Schake and Detter15–Reference Freedman17, Reference Buscher, Martin and Crocker42, Reference Story, Nanney and Schwartz52). Changes like these may be welcomed by stakeholders and could improve profitability(40, 41).
To our knowledge, the present study is the first evaluation of campus dining venues with high IRR comparing a number of institutions. The use of CBPR and the stakeholders at each institution facilitated the selection of the most appropriate establishments evaluated given limited time and resources. This is a new instrument that others can use to evaluate their campus dining venues. The modified tools and training materials are available at http://www.med.upenn.edu/nems/materials.shtml. These data provide baseline benchmarks that other campuses can use as a comparison. In addition, these data point to improvements that need to be made to improve the dining environment which can provide environmental supports needed to avoid unwanted weight gain.
Despite the strengths of the present study, it is important to also consider its limitations. The study is limited to a small percentage of the many post-secondary institutions in the USA. Although the NEMS-R is a validated measure of the healthfulness of restaurants, it is limited because it is not a comprehensive evaluation of restaurant offerings and is dependent on knowing the nutrient content of foods offered. As a result, many venues may be offering healthful/nutrient-dense foods, but if nutrition information is not provided the NEMS-R score will be lower than it actually is. NEMS-R and the NEMS-CD are useful for a detailed nutrient-based analysis but also generate extensive data to manage(Reference Lytle53). The NEMS-CD also needs to be validated to ensure it accurately reflects the nutrition/food environment of a post-secondary institution campus. A more parsimonious tool might enhance appeal and utility for use by those wishing evaluate their campus food environment. Researchers who recently developed tools to evaluate public schools(Reference Hearst, Lytle and Pasch54, Reference Krukowski, Philyaw Perez and Bursac55) found a checklist approach to evaluating foods for healthfulness to be less time-consuming and labour-intensive than the inventory method used in NEMS-R.
Future research
Not all students eat in dining halls and most also supplement their meal plan by purchasing food and snacks at local convenience and grocery stores. Thus, conducting an evaluation of the broader consumer retail food environment would round out the understanding of the full campus food environment(Reference Horacek, Erdman and Reznar22).
Validation of NEMS-CD for the campus dining environment is needed and could be accomplished via a comparative study with the Healthy Eating Index(Reference Reedy, Krebs-Smith and Bosire56) or one of the newly created campus/NEMS tools(Reference Freedman47, Reference Minaker, Raine and Cash48, Reference Lesser, Hunnes and Reyes57).
Conclusions
Most on-campus dining venues and nearby restaurants can improve their dining environments by increasing healthy food and drink options and facilitators of healthy eating and by decreasing barriers that promote obesity. Large-sized institutions in particular might have additional challenges to overcome to improve their campus dining environments. These findings should provide motivation for other campuses to evaluate the quality of their dining environments as the importance of environmental supports for obesity prevention becomes more evident.
Acknowledgements
Sources of funding: This work was supported by funding provided by Syracuse University, East Carolina University and the Experiment Stations for AL, FL, IN, NH, NJ, ME, MI, RI, SD, WI and WV. Conflict of interest: There are no conflicts of interests to report. Authors’ contributions: The Syracuse University team designed the research and the lead author analysed the data and wrote the paper. The NY, ME, SD, WI and NJ colleagues assisted with the pilot-test of the protocol, all states collected data and all co-authors assisted with the writing of the manuscript. Acknowledgements: The authors would like to thank the other state Primary Investigators: Susan Nitzke, Sharon Hoerr, Tandalayo Kidd, Onikia Brown-Esters, Beatrice Phillips, Karla Shelnutt; and graduate students: Melissa Reznar, Amanda Dent, Michelle Johnson, Jesse Morrell, Katherine DeIeso, Ginger Quick and Mallory Koenings; and all of the student data collectors for their hard work at every phase of the study. Thanks also to Dr Karen Glanz, MPH, for her review of the manuscript.
