Advertisement

Adolescent Purchasing Behavior at McDonald's and Subway

      Abstract

      Purpose

      To assess whether adolescents purchasing food at a restaurant marketed as “healthy” (Subway) purchase fewer calories than at a competing chain (McDonald's).

      Methods

      We studied 97 adolescents who purchased a meal at both restaurants on different days, using each participant as his or her control. We compared the difference in calories purchased by adolescents at McDonald's and Subway in a diverse area of Los Angeles, CA.

      Results

      Adolescents purchased an average of 1,038 calories (standard error of the mean [SEM]: 41) at McDonald's and 955 calories (SEM 39) at Subway. The difference of 83 calories (95% confidence interval [CI]: −20 to 186) was not statistically significant (p = .11). At McDonald's, participants purchased significantly more calories from drinks (151 vs. 61, p < .01) and from side dishes (i.e., French fries or potato chips; 201 at McDonald's vs. 35 at Subway, p < .01). In contrast, they purchased fewer cups of vegetables at McDonald's (.15 vs. .57 cups, p < .01).

      Conclusions

      We found that, despite being marketed as “healthy,” adolescents purchasing a meal at Subway order just as many calories as at McDonald's. Although Subway meals had more vegetables, meals from both restaurants are likely to contribute to overeating.

      Keywords

      See Related Editorial p. 429
      Implications and Contribution
      Despite being considered “healthy,” adolescents are likely to purchase just as many calories at Subway as at McDonald's. Although adolescents purchase more vegetables when dining at Subway, it is unclear whether this will reduce their risk for weight gain.
      When people dine away from home, the meals they consume are generally nutritionally inferior to what they consume at home. The amount of calories, sugar, and salt usually exceed what the Dietary Guidelines for Americans recommend [
      • Department of Agriculture Department of Health and Human Services
      Dietary Guidelines for Americans 2010.
      ] and the quantity of fruits and vegetables and other critical nutrients are too low [
      • French S.A.
      • Harnack L.
      • Jeffery R.W.
      Fast food restaurant use among women in the Pound of Prevention study: Dietary, behavioral and demographic correlates.
      ,
      • Bowman S.A.
      • Gortmaker S.L.
      • Ebbeling C.B.
      • et al.
      Effects of fast-food consumption on energy intake and diet quality among children in a national household survey.
      ]. Additionally, most studies [
      • French S.A.
      • Harnack L.
      • Jeffery R.W.
      Fast food restaurant use among women in the Pound of Prevention study: Dietary, behavioral and demographic correlates.
      ,
      • Pereira M.A.
      • Kartashov A.I.
      • Ebbeling C.B.
      • et al.
      Fast-food habits, weight gain, and insulin resistance (the CARDIA study): 15-year prospective analysis.
      ,
      • Duffey K.J.
      • Gordon-Larsen P.
      • Jacobs D.R.J.
      • et al.
      Differential associations of fast food and restaurant food consumption with 3-y change in body mass index: The Coronary Artery Risk Development in Young Adults Study.
      ,
      • Rosenheck R.
      Fast food consumption and increased caloric intake: A systematic review of a trajectory towards weight gain and obesity risk.
      ], but not all [
      • Jeffery R.W.
      • French S.A.
      Epidemic obesity in the United States: Are fast foods and television viewing contributing?.
      ], have shown a positive relationship between eating fast food and weight gain or obesity. Today, Americans consume more than one-third of their total calories on food away from home [].
      When people dine out, their choices are informed by a combination of dietary goals, the available options, and how options are presented. Marketing strategies, such as pricing, signage, promotions, and menu design influence people's choices and goals [
      • Cohen D.A.
      Obesity and the built environment: Changes in environmental cues cause energy imbalances.
      ].
      Whether it is possible for most individuals to choose a healthy meal in most away-from-home settings is an open question. The decision-making process in a fast food restaurant is very quick, a condition that favors making automatic, heuristically based decisions. Such decisions usually bypass thoughtful cognitive processes and lead customers to be guided by order, price, and volume, with larger portions being automatically preferred [
      • Cohen D.A.
      Obesity and the built environment: Changes in environmental cues cause energy imbalances.
      ,
      • Krider R.E.
      • Raghubir P.
      • Krishna A.
      Pizzas: π or square? Psychophysical biases in area comparisons.
      ].
      Yet today, several fast food restaurants are claiming that they are healthy places to eat because they offer some foods that are compatible with a healthy diet. In particular, the Subway chain claims it offers healthy fast food and helps its customers lose weight. The ZAGAT survey of fast food restaurants rated Subway as the top mega-chain for healthy options [

      Herklots T, Sampogna N. Zagat releases 2011 fast food survey. http://www.zagat.com/press/2011-fast-food-survey-results.

      ]. However, the proof of whether a restaurant is actually healthy should be measured by what customers actually purchase and consume.
      To determine if the availability of healthier options alongside less healthy options influences the total calories purchased, we studied what adolescents ordered when they went to two different fast food outlets: Subway and McDonald's. These are two of the largest fast food chains. McDonald's is the world's largest grossing fast food chain, with $24 billion in sales in 2010. Subway is its new top competitor, having just surpassed it with the most franchises in the world [
      • Jargon J.
      Subway runs past McDonald's chain.
      ].
      With pediatric obesity on the rise, parents are especially concerned about where to take their children to eat away from home. Our study examined whether an adolescent who purchases an afternoon meal will get more calories at McDonald's or Subway.

      Methods

      The study used a community partnered participatory research framework [
      • Jones L.
      • Wells K.
      Strategies for academic and clinician engagement in community-participatory partnered research.
      ] in which the University of California, Los Angeles, partnered with the Youth, Family, School and Community-Partnership In Action (YFSC-PIA), a group of mostly Filipino parents and children who actively participate in health-related activities. This collaboration selected the study sites (i.e., fast food restaurants); identified the methods used to enroll adolescents; and provided input on how the data would be collected, analyzed, and presented.

      Study sites

      Student and parent members of YFSC-PIA provided input into the study design and data collection methods, but they did not participate in the study. The advisors chose McDonald's as the restaurant most teens would categorize as “unhealthy” and chose Subway as the “healthy” restaurant. Both restaurants had a location in the same shopping plaza that was accessible to students in Carson, an urban area in Los Angeles, CA.
      This perceived difference in “healthiness” was confirmed with an objective measure called the Nutrition Environment Measures Survey for Restaurants (NEMS-R) [
      • Saelens B.E.
      • Glanz K.
      • Sallis J.F.
      • Frank L.D.
      Nutrition Environment Measures Study in Restaurants (NEMS-R): Development and evaluation.
      ]. This 25-item questionnaire counts the number of healthy options available in the restaurant as well as environmental factors, including signs promoting healthy/unhealthy eating, combo meal deals, pricing, and default choices for side items. This tool's inter-rater and test-retest reliability is generally high, with most kappa values greater than .80 [
      • Saelens B.E.
      • Glanz K.
      • Sallis J.F.
      • Frank L.D.
      Nutrition Environment Measures Study in Restaurants (NEMS-R): Development and evaluation.
      ]. The scores on NEMS-R range from −27 (most unhealthy) to 63 (most healthy).
      The principal investigator (L.I.L.), who had formal training in administering the NEMS-R, scored the two restaurants: McDonald's had a score of 22 and Subway had a score of 41.

      Participants

      Adolescents 12 to 21 years of age in the South Los Angeles area were eligible to participate and were recruited at community events, through individual contacts, and through approaching potential participants onsite. In the first method, an adult leader in the YFSC-PIA contacted adolescents and parents at community and church events, where she handed out recruitment flyers that contained the dates and restaurant locations of the study as well as parental consent forms. In the second method, students of YFSC-PIA posted the study website on their Facebook pages and texted their friends to encourage them to visit the study website. The study website contained the study information sheet with dates and locations of the study and parental consent forms. The flyers and the websites directed participants to show up on specific days at either one of the two restaurants. Finally, we also approached potentially eligible individuals visiting the restaurants on the study days and asked them to participate. Participants were not informed of the hypothesis of the study; they were told that we were interested in seeing what adolescents ate at different restaurants.

      Data collection

      The study was conducted during the month of May 2011. Study days and times included Monday through Thursday from 3 to 5 p.m. At the first restaurant visit, which could either be at the Subway or McDonald's, participants checked in with the research assistant (RA) by providing their first and last names and contact numbers. Students were instructed to purchase a “meal” and return their receipt to the RA. They then received a short written survey to document their year of birth, ethnicity/race, mother's educational level, height, and weight.
      The RA and YFSC-PIA student advisors then used purchase receipts and restaurant menus to fill out food intake surveys. The food intake surveys gathered details about the contents of the meal necessary to calculate calorie intake. Because purchase receipts occasionally excluded details about a meal, the RA asked participants additional questions. For McDonald's meals, participants were asked about the following: specific meal selected from category (e.g., Extra Value Meals, Dollar Menu), fountain soda drink selected, and number and type of condiment packets (ketchup, mustard, ranch sauce, barbecue sauce) and dressings. For Subway meals, toppings (vegetables, seasonings, sauces, oils, and cheeses) added to sandwiches, type of chips selected, and fountain drinks selected were identified. Before leaving the restaurant, students were given a reminder card indicating the upcoming dates for their second visit at the alternate restaurant. During the second visit, students were required to check in and provide their first and last name to the RA in order to verify it was the participant's second restaurant visit and that the participant was visiting the correct restaurant. Participants were instructed to purchase a meal and return the receipt to the RA or YFSC-PIA student advisors who recorded the food intake surveys in the same manner.
      We created a database of all the possible food orders at the two restaurants based on nutrition information available on their website. The only exception was for the fountain soda drinks. We used McDonald's nutrition information for all the fountain drinks because it was the only information posted online. For drinks that were only available at Subway (e.g., PowerAde, Nestea), we used information from Coca Cola Company's customer service telephone line.

      Consent and incentives

      The RAND Corporation's Institutional Review Board approved the study. Adolescents younger than 18 years of age were required to present a signed written consent form from a parent or guardian. After completion of both meals, all participants received a $20 Target gift card. Participants were also entered into a contest to win an active video game system.

      Statistical methods

      Power calculation

      To calculate a sample size for this study, we used estimates from other studies of fast food consumption. We erred on the side of conservative estimates. The estimated standard deviation of one fast food meal was 443 calories, based on a study of fast food consumption in children [
      • Zoumas-Morse C.
      • Rock C.L.
      • Sobo E.J.
      • Neuhouser M.L.
      Children's patterns of macronutrient intake and associations with restaurant and home eating.
      ]. To estimate the correlation between what the participants would eat at two different meals, we used an estimate of the correlation between what adolescents ate on two different days [
      • Bowman S.A.
      • Gortmaker S.L.
      • Ebbeling C.B.
      • et al.
      Effects of fast-food consumption on energy intake and diet quality among children in a national household survey.
      ]. The correlation between the calories they consumed on these 2 days was .53. Using this correlation and the standard deviation of a fast food meal, we calculated a standard deviation of the difference of 430 calories. Finally, the difference in calories we could expect to find between meals purchased at the two restaurants was estimated to be 123 from a study in which adults were surveyed eating at restaurants including McDonald's and Subway [
      • Bassett M.T.
      • Dumanovsky T.
      • Huang C.
      • et al.
      Purchasing behavior and calorie information at fast-food chains in New York City, 2007.
      ]. To have 80% power to detect a 123-calorie difference with a two-tailed alpha of .05, 95 participants needed to be enrolled.

      Data analysis

      Nutrients at the two restaurants were compared with a paired t-test. After looking at the primary outcome (calories), we performed secondary analyses looking for differences in other nutrients and food groups. We did not adjust our significance level for these multiple comparisons because most values were either highly significant or insignificant. For comparisons of proportions, we used McNemar's test.
      At the beginning of the study, we did not realize that ketchup packets at McDonald's were not listed on the itemized receipts. Fifty-two participants had incomplete ketchup data. (Each ketchup packet contains 15 calories.) To correct this problem, we imputed the number of ketchup packets for those with missing data, based on the total amount of calories each participant ordered and whether they ordered French fries.
      We also used a linear regression to test whether any individual characteristics of the participants predicted the difference in calories between the restaurants. Our analyses were all done with STATA, version 12 (College Station, TX).

      Results

      We recruited 97 adolescents who ate at both restaurants. Ten adolescents only completed a meal at one restaurant and were not included in the analysis. Although adolescents could start at either restaurant, 69% had their first meal at McDonald's. The average age was 16.9 (SD 2.3) with 61% being male and 47% Asian. Most (52%) of the participants reported that their mothers had graduated from college (Table 1).
      Table 1Characteristics of the participants
      CountPercent
      Male5960.8
      Age (mean and SD)16.9 (2.3)
      Ethnicity
       Latino/a1717.5
      Race
       White44.1
       Black22.1
       Mexican, Central/South American1212.4
       Asian4647.4
       Other or mixed race3232.1
       Refused11.0
      Highest grade mother completed
       Less than high school1313.4
       High school1616.5
       Some college133.3
       College4142.3
       Graduate school88.3
       Don't know/refused66.2
      Adolescents purchased 1,038 calories (standard error of the mean [SEM] 41, degrees of freedom = 96) at McDonald's and 955 calories (SEM 39) at Subway (Table 2). This difference of 83 calories (95%CI: −20 to 186) was not statistically significant (p = .11). When we undertook the analysis without any ketchup packets included (recorded or imputed), we found similar results: 1,022 calories versus 955 (p = .19).
      Table 2Purchases by adolescents at Subway and McDonald's
      VariableMcDonald's, mean (SEM)Subway, mean (SEM)Difference of McDonald's - Subway, mean (SEM)p Value
      Calories (kcal) (primary outcome)1,038 (41)955 (39)83 (53).11
      Cost of meal ($)4.46 (.20)6.14 (.17)−1.67 (.24)<.01
      Origin of calories
       Main dishes572 (28)784 (31)−212 (39)<.01
       Drinks151 (16)61 (11)90 (18)<.01
       Sides201 (20)35 (8.2)166 (19)<.01
       Condiments63 (9.5)46 (12)17 (13).19
       Desserts51 (16)28 (13)23 (21).28
      Nutrients
       Carbohydrates (g)128 (5.7)102 (4.6)26 (6.9)<.01
       Fiber (g)5.9 (.32)6.7 (.35)−.84 (.44).06
       Sugar (g)54 (3.9)36 (3.3)18 (4.6)<.01
       Fat (g)45 (2.1)42 (2.5)3.9 (2.9).18
       Saturated Fat (g)12.6 (.65)13.5 (.80)−.87 (.95).36
       Protein (g)32 (1.4)41 (1.8)−9.8 (2.1)<.01
       Sodium (mg)1,829 (99)2,149 (93)−320 (120).01
      Types of food
       Fruits (cups).01 (.01).01 (.01)0 (.01).95
       Vegetables (cups).15 (.02).57 (.04)−.42 (.04)<.01
       Drinks64% (4.9)28% (4.6)36% (6.7)<.01
       (% who purchased)
       Sides purchased58% (5.0)13% (3.5)44% (7.1)<.01
       (% who purchased)(Fries)(Chips)
      SEM = standard error of the mean.
      The participants obtained their calories from different types of food in the two restaurants. For main dishes (i.e., sandwiches), participants purchased 572 calories at McDonald's and 784 at Subway (p < .01). Participants purchased significantly more calories from the drinks category at McDonald's (151 vs. 61, p < .01). They also purchased more calories as side dishes (i.e., French fries or potato chips): 201 at McDonald's versus 35 at Subway (p < .01). In contrast, they purchased fewer cups of vegetables at McDonald's (.15 vs. .57, p < .01).
      At McDonald's, the teenagers purchased more grams of carbohydrates (128 vs. 102, p < .01) and sugars (54 g vs. 36 g, p < .01) than at Subway. There was no significant difference between the restaurants in fat or saturated fat. Participants purchased fewer grams of protein at McDonald's than at Subway (32 vs. 41, p < .01). Overall sodium amounts were less at McDonald's (1,829 mg) than at Subway (2,149 mg, p = .01). Adolescents spent significantly less at McDonald's ($4.46), compared to Subway ($6.14, p < .01).
      The regression revealed no associations between ethnicity/race, mother's educational status, sex, body mass index, age, or whether they ate at Subway first and the calorie differences between the two restaurants (data not shown).

      Discussion

      Fast food has been present in the United States for almost a century. The increase in the availability of fast food has coincided with the worldwide obesity epidemic, and has been implicated as a cause [
      • St-Onge M.P.
      • Keller K.L.
      • Heymsfield S.B.
      Changes in childhood food consumption patterns: A cause for concern in light of increasing body weights.
      ]. To counter this concern, many fast food restaurants have changed menus and highlighted healthy items [
      • Parker L.
      • Spear M.
      • Holovach N.F.
      • Olson S.
      Legal strategies in childhood obesity prevention.
      ]. The National Restaurant Association reported that chefs in the United States list “healthful kids' meals” as the third-most popular trend for 2012 [
      ].
      Subway tries to attract customers with promotions that claim they are the healthiest fast food chain. These promotions include “six subs with 6 grams of fat or less” and providing apples and milk with kid's meals [

      Subway menu. Doctor's Associates, Inc. Available at: http://www.subway.com/menu/default.aspx.

      ]. The most famous of all their promotions uses Jared, a customer who apparently lost weight by mostly eating at Subway [].
      The true measure of a healthy restaurant, however, is whether people actually consume a healthy meal there. If the healthy options do not look attractive, are not displayed prominently, are priced higher, or are offered with high-calorie side dishes, consumers may be less likely to choose those foods [
      • Wansink B.
      Environmental factors that increase the food intake and consumption volume of unknowing consumers.
      ].
      The prespecified primary outcome of our study was the number of calories adolescents purchased. Although there is some evidence that calories from different foods may have different effects on satiety and resting energy expenditure [
      • Ebbeling C.B.
      • Swain J.F.
      • Feldman H.A.
      • et al.
      Effects of dietary composition on energy expenditure during weight-loss maintenance.
      ], there is currently no better predictor of future weight gain than the total number of calories an individual consumes. In this study, using the metric of calories to compare two fast food chains, we found no difference between McDonald's and Subway. Adolescents at both restaurants consumed about 1,000 calories for an afternoon “meal”. Even if this was to be their lunch or dinner for the day, it still exceeds the amount of calories recommended for adolescents at one meal. The Institute of Medicine (IOM) states that school lunches for adolescents should be no more than 850 calories [
      • Institute of Medicine (U.S.). Committee on Nutrition Standards for National School Lunch and Breakfast Programs
      ].
      Despite finding no differences in the number of calories ordered, we did find differences in the nutritional value. Youth ordered fewer calories from sugary drinks and salty snacks at Subway. They also purchased more vegetables at Subway. However, these results should be interpreted with caution, because although Subway may be “healthier,” it still may not be healthy. Adolescents still only purchased about a half a cup of vegetables at Subway, less than the IOM recommends for a lunch (i.e., 1 cup) [
      • Institute of Medicine (U.S.). Committee on Nutrition Standards for National School Lunch and Breakfast Programs
      ]. They also purchased more sodium at Subway than McDonald's, and at a level (2,149 mg) that is nearly three times higher than the IOM recommends for school lunches (i.e., 740 mg) [
      • Institute of Medicine (U.S.). Committee on Nutrition Standards for National School Lunch and Breakfast Programs
      ]. The higher salt amounts at Subway may be an indicator of processed food and meat purchasing. Processed food consumption in adolescents may be associated with obesity [
      • Libuda L.
      • Kersting M.
      • Alexy U.
      Consumption of dietary salt measured by urinary sodium excretion and its association with body weight status in healthy children and adolescents.
      ]. Processed meat consumption is associated with a higher risk of cardiovascular disease and cancer [
      • Micha R.
      • Wallace S.K.
      • Mozaffarian D.
      Red and processed meat consumption and risk of incident coronary heart disease, stroke, and diabetes mellitus.
      ,
      • Marmot M.
      • Atinmo T.
      • Byers T.
      • et al.
      Food, nutrition, physical activity, and the prevention of cancer: A global perspective.
      ].
      The main limitation of this study is that we did not examine what the participants ate for the rest of the day. Many of the students knew that they would be participating in the study before ordering the meal. They could have compensated for this by skipping other meals, knowing that they would be eating a mid-afternoon meal. However, the students in our community partnership advised us that many of their peers regularly eat an early lunch at school and then often (i.e., about three times a week) go after school to a fast food restaurant to get a meal to “hold them over” until dinner. The participants could have also could have compensated for these calories later in the day, such as by skipping dinner. Research shows that normal-weight adolescents compensate for fast food calorie intake by decreasing calories throughout the rest of their day, whereas overweight adolescents tend not to compensate [
      • Ebbeling C.B.
      • Sinclair K.
      • Pereira M.
      • et al.
      Compensation for energy intake from fast food among overweight and lean adolescents.
      ]. Additionally, the adolescents in this study ordered similar amounts of calories as those in other studies [
      • Yamamoto J.A.
      • Yamamoto J.B.
      • Yamamoto B.E.
      • Yamamoto L.G.
      Adolescent fast food and restaurant ordering behavior with and without calorie and fat content menu information.
      ].
      We surveyed adolescents in a working-class area of urban Los Angeles. Many of the participants were of Asian origin, with many of mixed race and ethnicity. The patterns of purchasing may not be generalizable to certain other groups, such as those living in rural areas and very high- or low-income areas.
      Our study's strength comes from using each participant as his or her own control. Other studies of fast food have looked what different people exiting two different restaurants order [

      Boutelle KN, Fannin H, Newfield R, Hamak L. Nutritional quality of lunch meal purchased for children at a fast-food restaurant. Childhood Obes 2011:7:316–22.

      ,
      • Chandon P.
      • Wansink B.
      The biasing health halos of fast-food restaurant health claims: Lower calorie estimates and higher side-dish consumption intentions.
      ]. These cross-sectional studies suffer from an endogenicity problem: people who eat at a “healthy” restaurant (e.g., Subway) may be different (i.e., more health conscious) than those eating at another chain (e.g., McDonald's). Another weakness of other studies is presenting adolescents with menus in artificial scenarios [
      • Yamamoto J.A.
      • Yamamoto J.B.
      • Yamamoto B.E.
      • Yamamoto L.G.
      Adolescent fast food and restaurant ordering behavior with and without calorie and fat content menu information.
      ,
      • Tandon P.
      • Wright J.
      • Zhou C.
      • et al.
      Nutrition menu labeling may lead to lower-calorie restaurant meal choices for children.
      ]. We eliminated both of these problems by having the same adolescents order meals at both restaurants, with their own money, in a usual setting. Thus, we can be confident that our results reflect the true effect of changing the restaurant environment on an individual.
      We used two standards to classify Subway as the “healthy” fast food restaurant. First, we used the NEMS-R survey instrument, which scored Subway as healthier. Because we found no difference in calories in restaurants rated differently by NEMS-R, public health officials should not rely on the NEMS-R in its current form to predict which restaurants may be larger contributors to obesity. Further research may be able to adjust the survey to predict more accurately calorie purchases. However, NEMS-R could be said to predict higher vegetable purchases, a noteworthy goal to prevent chronic disease [
      • Department of Agriculture Department of Health and Human Services
      Dietary Guidelines for Americans 2010.
      ].
      The second standard we used to determine healthiness was our community's perception. Our community partner advised us that adolescents think of Subway as the healthy place to go for an afternoon meal. This sentiment may also hold for the general public. Yet, despite the perceived public opinion that Subway is the healthy fast food restaurant, adolescents in our study purchased no fewer calories as this site than at McDonald's. Subway may indeed offer more healthy items. But, Subway's marketing may counteract the healthy offerings. For instance, pricing (such as combo meals), promotions ($5 foot-longs), and signage (promoting a Philly-cheese steak sub) all can urge people to purchase more calories. In addition, the “healthy halo” of Subway may make people think they are eating healthier, even when they are not. This could result in individuals actually purchasing more calories than they intend [
      • Chandon P.
      • Wansink B.
      The biasing health halos of fast-food restaurant health claims: Lower calorie estimates and higher side-dish consumption intentions.
      ].
      Ultimately, the nutritional quality of the food people actually consume should be the outcome we examine when determining whether a restaurant is healthy. Public health authorities may want to survey restaurants in their communities and rate them based on how many calories consumers purchase. This could allow the public to choose restaurants that actually help people purchase fewer calories. Until then, parents should steer their children away from fast food restaurants. No matter which one they choose, they are likely to purchase too many calories.

      Acknowledgments

      This study was funded by the Robert Wood Johnson Foundation Clinical Scholars Program and conducted while L.I.L. was at the University of California, Los Angeles. The foundation had no role in the design or the analysis of the study. The authors would like to thank the members of YSFC-PIA, especially the youths who helped design, collect, and analyze data: Virginia-Cleo V. Velasquez, Nathan C. Suarez, and Luigi L. Alejandrino. The authors report no conflicts of interest. L.I.L. had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Preliminary results were presented at a private meeting for the Robert Wood Johnson Foundation in October 2011 and at the 2012 meeting of the American Public Health Association.

      References

        • Department of Agriculture Department of Health and Human Services
        Dietary Guidelines for Americans 2010.
        Government Printing Office, Washington, DC2010 (Available at:)
        • French S.A.
        • Harnack L.
        • Jeffery R.W.
        Fast food restaurant use among women in the Pound of Prevention study: Dietary, behavioral and demographic correlates.
        Intl J Obesity. 2000; 24: 1353-1359
        • Bowman S.A.
        • Gortmaker S.L.
        • Ebbeling C.B.
        • et al.
        Effects of fast-food consumption on energy intake and diet quality among children in a national household survey.
        Pediatrics. 2004; 113: 112-118
        • Pereira M.A.
        • Kartashov A.I.
        • Ebbeling C.B.
        • et al.
        Fast-food habits, weight gain, and insulin resistance (the CARDIA study): 15-year prospective analysis.
        Lancet. 2005; 365: 36-42
        • Duffey K.J.
        • Gordon-Larsen P.
        • Jacobs D.R.J.
        • et al.
        Differential associations of fast food and restaurant food consumption with 3-y change in body mass index: The Coronary Artery Risk Development in Young Adults Study.
        Am J Clin Nutr. 2007; 85: 201-208
        • Rosenheck R.
        Fast food consumption and increased caloric intake: A systematic review of a trajectory towards weight gain and obesity risk.
        Obesity Rev. 2008; 9: 535-547
        • Jeffery R.W.
        • French S.A.
        Epidemic obesity in the United States: Are fast foods and television viewing contributing?.
        Am J Public Health. 1998; 88: 277-280
      1. What We Eat in America, NHANES 2007-2008. Available at: http://www.ars.usda.gov%2FSP2UserFiles%2FPlace%2F12355000%2Fpdf%2F0708%2FTable_9_AWY_GEN_07.pdf.

        • Cohen D.A.
        Obesity and the built environment: Changes in environmental cues cause energy imbalances.
        Intl J Obesity. 2008; : S137-S142
        • Krider R.E.
        • Raghubir P.
        • Krishna A.
        Pizzas: π or square? Psychophysical biases in area comparisons.
        Marketing Sci. 2001; 20: 405-425
      2. Herklots T, Sampogna N. Zagat releases 2011 fast food survey. http://www.zagat.com/press/2011-fast-food-survey-results.

        • Jargon J.
        Subway runs past McDonald's chain.
        Wall Street Journal. 2011; (Available at:)
        • Jones L.
        • Wells K.
        Strategies for academic and clinician engagement in community-participatory partnered research.
        JAMA. 2007; 297: 407-410
        • Saelens B.E.
        • Glanz K.
        • Sallis J.F.
        • Frank L.D.
        Nutrition Environment Measures Study in Restaurants (NEMS-R): Development and evaluation.
        Am J Prev Med. 2007; 32: 273-281
        • Zoumas-Morse C.
        • Rock C.L.
        • Sobo E.J.
        • Neuhouser M.L.
        Children's patterns of macronutrient intake and associations with restaurant and home eating.
        J Am Dietetic Assoc. 2001; 101: 923-925
        • Bassett M.T.
        • Dumanovsky T.
        • Huang C.
        • et al.
        Purchasing behavior and calorie information at fast-food chains in New York City, 2007.
        Am J Public Health. 2008; 98: 1457-1459
        • St-Onge M.P.
        • Keller K.L.
        • Heymsfield S.B.
        Changes in childhood food consumption patterns: A cause for concern in light of increasing body weights.
        Am J Clin Nutr. 2003; 78: 1068-1073
        • Parker L.
        • Spear M.
        • Holovach N.F.
        • Olson S.
        Legal strategies in childhood obesity prevention.
        National Academies Press, Washington, DC2011 (1–105)
      3. What's hot in 2012. National Restaurant Association, Washington, DC2011
      4. Subway menu. Doctor's Associates, Inc. Available at: http://www.subway.com/menu/default.aspx.

      5. Jared's journey. Doctor's Associates, Inc. Available at: http://www.subway.com/subwayroot/freshbuzz/website/jareds_journey/default.aspx.

        • Wansink B.
        Environmental factors that increase the food intake and consumption volume of unknowing consumers.
        Annu Rev Nutr. 2004; 24: 455-479
        • Ebbeling C.B.
        • Swain J.F.
        • Feldman H.A.
        • et al.
        Effects of dietary composition on energy expenditure during weight-loss maintenance.
        JAMA. 2012; 307: 2627-2634
        • Institute of Medicine (U.S.). Committee on Nutrition Standards for National School Lunch and Breakfast Programs
        Stallings V.A. Suitor C.W. Taylor C.L. School Meals. The National Academies Press, Washington, DC2010
        • Libuda L.
        • Kersting M.
        • Alexy U.
        Consumption of dietary salt measured by urinary sodium excretion and its association with body weight status in healthy children and adolescents.
        Public Health Nutr. 2011; 15: 433-441
        • Micha R.
        • Wallace S.K.
        • Mozaffarian D.
        Red and processed meat consumption and risk of incident coronary heart disease, stroke, and diabetes mellitus.
        Circulation. 2010; 121: 2271-2283
        • Marmot M.
        • Atinmo T.
        • Byers T.
        • et al.
        Food, nutrition, physical activity, and the prevention of cancer: A global perspective.
        World Cancer Research Fund/American Institute for Cancer Research, Washington, DC2007
        • Ebbeling C.B.
        • Sinclair K.
        • Pereira M.
        • et al.
        Compensation for energy intake from fast food among overweight and lean adolescents.
        JAMA. 2004; 291: 2828
        • Yamamoto J.A.
        • Yamamoto J.B.
        • Yamamoto B.E.
        • Yamamoto L.G.
        Adolescent fast food and restaurant ordering behavior with and without calorie and fat content menu information.
        J Adolesc Health. 2005; 37: 397-402
      6. Boutelle KN, Fannin H, Newfield R, Hamak L. Nutritional quality of lunch meal purchased for children at a fast-food restaurant. Childhood Obes 2011:7:316–22.

        • Chandon P.
        • Wansink B.
        The biasing health halos of fast-food restaurant health claims: Lower calorie estimates and higher side-dish consumption intentions.
        J Cons Res. 2007; 34: 301-314
        • Tandon P.
        • Wright J.
        • Zhou C.
        • et al.
        Nutrition menu labeling may lead to lower-calorie restaurant meal choices for children.
        Pediatrics. 2010; 125: 244