Increased Portion Size Leads to Increased ... - Wiley Online Library

Increased Portion Size Leads to Increased Energy Intake in a Restaurant Meal Nicole Diliberti,* Peter L. Bordi,† Martha T. Conklin,† Liane S. Roe,* and Barbara J. Rolls*

Abstract DILIBERTI, NICOLE, PETER L. BORDI, MARTHA T. CONKLIN, LIANE S. ROE, AND BARBARA J. ROLLS. Increased portion size leads to increased energy intake in a restaurant meal. Obes Res. 2004;12:562–568. Objective: Eating frequently in restaurants is one of the behaviors associated with obesity. This study examined whether increasing the portion size of an entreé affected energy intake at a restaurant meal. Research Methods and Procedures: In a cafeteria-style restaurant on different days, the size of a pasta entreé was varied from a standard portion (248 g) to a large portion (377 g). The entreé price was not changed. Intake of the entreé was determined by covertly weighing each dish before and after the meal; intake of all other foods was determined by estimating the percent consumed. The 180 adult customers who purchased the entreé also completed a survey in which they rated characteristics of the meal, including the appropriateness of the entreé portion size and the amount that they ate compared with their usual meal. Results: Portion size had a significant effect on intake of the entreé (p ⬍ 0.0001). Compared with customers who purchased the standard portion, those who purchased the larger portion increased their energy intake of the entreé by 43% (719 kJ; 172 kcal) and of the entire meal by 25% (664 kJ; 159 kcal). There was no difference between the two groups of customers in ratings of the appropriateness of the portion size or of the amount that was eaten in relation to their usual meal. Discussion: In a restaurant setting, increasing the size of an entreé results in increased energy intake. These results sup-

Received for review August 19, 2003. Accepted in final form January 15, 2004. The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. *Department of Nutritional Sciences and †School of Hotel, Restaurant, and Recreation Management, The Pennsylvania State University, University Park, Pennsylvania. Address correspondence to Barbara J. Rolls, The Pennsylvania State University, 226 Henderson Building, University Park, PA 16802-6502. E-mail: [email protected] Copyright © 2004 NAASO

562

OBESITY RESEARCH Vol. 12 No. 3 March 2004

port the suggestion that large restaurant portions may be contributing to the obesity epidemic. Key words: serving size, body size, food intake, dining out, adults

Introduction Eating away from home has been shown to be associated with increased energy intake as well as with overweight and obesity (1– 4). One study among men and women 19 to 80 years of age has found that the frequency of consuming restaurant food is positively associated with body fatness (1). Another study has found that, in women, frequent consumption of meals at fast food establishments is associated with increased BMI (kilograms per meters squared), and in men, eating at both restaurants and fast food establishments is associated with increased BMI (2). In the year 2000, Americans spent 47% of their food expenditures on foods consumed away from home, and this proportion continues to increase (5). When consumers eat away from home, they are offered a large variety of low-cost, energydense foods in large portions (6,7); all of these factors may encourage overconsumption of energy and lead to obesity. In this study, we examined whether increasing the portion size of an entreé affected energy intake at a restaurant meal. Several researchers have noted that during the last 25 years, portion sizes of commercially available foods have increased in parallel with the increase of the incidence of overweight and obesity (8 –10). Recent research has also demonstrated that, in laboratory-based experimental studies, the portion size of food directly influences the amount that is consumed. Rolls et al. (11) showed that when participants were served a portion of macaroni and cheese that was 50% bigger than a standard portion, their energy intake increased by 19%, and when the portion was increased by 100%, intake increased by 30%. Similar increases were found in laboratory studies when researchers increased the portion size of packaged potato chips, deli-type sandwiches, or all foods consumed over 2 days (12–14). To further understand the influence of portion size on energy intake, its effect on intake in more naturalistic eating

Restaurant Portion Size Affects Intake, Diliberti et al.

environments must be examined. Apart from one study showing that the portion of popcorn given to movie-goers affected the amount eaten (15), there have been no direct demonstrations in a naturalistic setting that the portion size of food affects intake. The main objective of this study was to measure the effect on meal energy intake of increasing the portion size of an entre´ e that was sold in a commercial restaurant. A secondary objective was to evaluate customer perceptions of the portion size of the entre´ e and of the amount that they ate. The participants in the study were restaurant customers who completed a survey but were unaware that their intake was being monitored. We hypothesized that when the portion size of an entre´ e was increased by 50%, customers would consume significantly more energy than when the standard portion was offered. We were also interested in whether customers would compensate for the increased intake from the entre´ e by reducing their intake of other foods at the meal and whether we could identify any characteristics of customers that would predict the degree of their response to increased portion size.

Research Methods and Procedures Study Design The study used a between-subjects, parallel-group design and was conducted in the setting of a public cafeteria-style restaurant on a university campus. On 10 days over 5 months, we covertly recorded the food intake of customers who purchased a baked pasta entre´ e from a serving line at lunch. On 5 of the days, the portion size of the entre´ e was the standard (100%) portion, and on 5 different days, the size was increased to 150% of the standard portion. The same portion size of the entre´ e was sold on two consecutive days of a given study week (Monday to Thursday). Study weeks were separated by at least 2 weeks, and the portion size sold in a given week was randomly determined. Customers who purchased the entre´ e were also asked to complete a brief survey. All aspects of the study were approved by the Office for Research Protections of The Pennsylvania State University. Subjects We included as subjects all customers who purchased the pasta entre´ e on study days and completed a short survey. Customers were excluded from the study if they had purchased the entre´ e on a previous study day or if they shared their meal with another person. On the 10 study days, 180 customers met the inclusion criteria; 89 customers purchased the 100% portion, and 91 customers purchased the 150% portion. The two groups of customers did not differ in sex or body size but did differ in age and university status (Table 1). Compared with the customers who purchased the 100% portion, more of the customers who purchased the 150% portion were 25 to 34 years of age and were graduate students.

Table 1. Distribution of customer characteristics* 100% Portion 150% Portion (n ⴝ 89) (n ⴝ 91) Sex Male Female Estimated body size† Underweight or normal weight Overweight or obese Age category 18 to 24 years 25 to 34 years 35 to 44 years ⱖ45 years University affiliation Undergraduate student Graduate student Faculty or staff Visitor

38 (44%) 49 (56%)

41 (48%) 44 (52%)

61 (69%) 28 (31%)

57 (64%) 32 (36%)

36 (41%) 16 (18%) 13 (15%) 23 (26%)

30 (33%) 43 (47%)‡ 8 (9%) 10 (11%)‡

28 (32%) 18 (20%) 36 (40%) 7 (8%)

28 (31%) 34 (37%)‡ 28 (31%) 1 (1%)‡

* Numbers do not always add up to total sample size because of missing data on the customer survey. † Estimated by study coordinator. ‡ Frequency is significantly different from that of customers in the 100% portion condition (p ⬍ 0.02).

Foods An entre´ e of baked pasta in a cheese sauce, a popular item at the restaurant, was chosen for the experiment. The entre´ e was served in an individual casserole dish, and its portion size could be increased relatively unobtrusively (Figure 1). The ingredients of the entre´ e were ziti pasta, canned diced tomatoes, four cheeses (ricotta, mozzarella, provolone, and romano), heavy cream, fresh basil, garlic, salt, and pepper. The macronutrient composition of the entre´ e was 27% energy as carbohydrate, 19% as protein, and 54% as fat, and the energy density was 7.1 kJ/g (1.7 kcal/g). The standard portion of the pasta entre´ e that was sold in the restaurant was used as the experimental control (100% portion). After baking, the 100% portion had a mean weight of 248.4 ⫾ 0.4 g and an energy content of 1766 ⫾ 2.8 kJ (422 ⫾ 0.7 kcal), and the 150% portion had a mean weight of 376.6 ⫾ 0.6 g and an energy content of 2647 ⫾ 3.9 kJ (633 ⫾ 0.9 kcal). On study days, the entre´ e was prepared by study staff in a research kitchen separate from the restaurant kitchen so that the ingredients, preparation methods, and portion size could be controlled. The pasta entre´ e was served with two accompaniments that were standard in portion size: one-half a tomato topped with pesto and a OBESITY RESEARCH Vol. 12 No. 3 March 2004

563


Figure 1: Standard (100%) and larger (150%) portion sizes of the baked pasta entre´ e are shown on the left and right, respectively, for comparison. The two portions were served on different days; therefore, customers never saw both of them simultaneously.

white bread roll with a butter packet. On study days, the accompaniments to the pasta entre´ e and all other foods in the serving lines were prepared by restaurant employees. Food Intake Data The pasta entre´ es in the individual casserole dishes were weighed after baking and before being placed on the restaurant serving line. After the meal, the uneaten portion of each entre´ e was weighed in the research kitchen. Entre´ e weights were recorded to the nearest 0.1 g. For the pasta entre´ e, energy intake was calculated from the weight consumed and the energy density. Customers were classified as consuming the entire entre´ e if there was ⱕ10 g remaining in the casserole dish. Because of the nature of the restaurant setting, the other food items (side dishes, desserts, and beverages) and entre´ e accompaniments (tomato, roll, and butter packet) were not weighed before or after the meal. Instead, study staff identified the additional foods purchased by each customer and estimated the percent of each serving that was consumed. Energy intake was calculated from the energy content of a single serving, using information from food suppliers or restaurant recipe analysis. Customer Survey Data Customers who purchased the pasta entre´ e were approached during the meal and asked by study staff to com564


plete a short survey. Because the restaurant is operated by a university food service department, customer satisfaction surveys are commonly distributed. Using a seven-point scale, customers rated the appropriateness of the portion size, how the amount they ate compared with their usual lunch, the value for money of the meal, and the taste of the entre´ e. Customers were also asked to indicate their sex, age group, and university affiliation (faculty, staff, graduate student, undergraduate student, or visitor). The study coordinator unobtrusively recorded an estimate of the body size of each subject as 1) normal weight or underweight, 2) overweight, or 3) notably obese, using a chart of body silhouettes (16). Statistical Analysis Data were analyzed using the SAS System (Version 8.1; SAS Institute, Cary, NC). A linear mixed model was used to analyze differences in energy intake and survey responses between the two conditions of portion size; survey responses were treated as continuous variables. Energy intakes were analyzed for the pasta entre´ e alone; the separate meal components of the accompaniments, side dishes, desserts, and beverages; and the meal in its entirety. The subject characteristics (age, sex, estimated body size, and university affiliation) were added as factors to the model to determine whether they influenced the relationship between portion size and the outcome measures. Student’s t tests


Figure 2: Mean ⫾ SE energy intake at lunch from all meal components by portion size of entre´ e (n ⫽ 180). Significant differences in intakes between conditions of portion size are marked (*p ⬍ 0.015; **p ⬍ 0.0001).

were used to compare the characteristics of customers in the two portion size conditions. Results were considered significant at p ⬍ 0.05.

groups of customers. Overall meal intake was increased by 25% (664 kJ; 159 kcal) when the larger portion was offered compared with the standard portion (p ⬍ 0.0001; Table 2). The effect of portion size on intake remained significant (p ⬍ 0.001) when the group of customers who consumed the entire entre´ e was excluded from the analysis (61% of customers who purchased the standard portion and 44% of customers who purchased the larger portion). Among the customers who did not consume the entire entre´ e, those who purchased the larger portion consumed 649 kJ (155 kcal) more than whose who purchased the standard portion. The effect of portion size on entre´ e intake was also significant among customers who purchased side dishes, desserts, or beverages in addition to the pasta entre´ e and those who did not purchase additional food (data not shown). Effect of Subject Characteristics None of the subject characteristics affected the relationship between portion size and energy intake. Thus, there was a significant increase in energy intake of the larger entre´ e in both men and women, in both normal weight and overweight customers, and in customers of all age groups and university affiliations. In all of these customer subgroups, overall meal energy intake also increased when the larger portion size was purchased.

Results Energy Intake Portion size had a significant effect on energy intake of the entre´ e. When the 150% portion size was offered, customers who purchased it consumed 43% more energy from the entre´ e (719 kJ; 172 kcal) than customers who purchased the 100% portion size (p ⬍ 0.0001; Figure 2). Customers who purchased the larger portion also ate significantly more of the entre´ e accompaniments (tomato, roll, and butter) than those who purchased the standard portion, even though the portion size of the accompaniments was standard for all customers. Energy intake from side dishes, desserts, and beverages did not differ significantly between the two

Survey Responses Responses to the survey are shown in Table 3. Overall, the rating of appropriateness of the portion size of the entre´ e did not differ between customers who purchased the 150% portion and the 100% portion. There was, however, an effect of subject body size on this rating. Among underweight and normal weight customers, those who purchased the 150% portion rated it as closer to the “too large” end of the seven-point scale than those who purchased the 100% portion (mean rating, 4.9 vs. 4.3; p ⬍ 0.006). Among overweight and obese customers, however, this rating did not differ significantly with portion size (mean rating, 4.2).

Table 2. Energy intakes (mean ⫾ SE) Meal component

100% Portion, kJ(kcal)

150% Portion, kJ(kcal)

Pasta entre´ e Entre´ e accompaniments‡ Optional items§ Entire meal

1671 ⫾ 13 (399 ⫾ 3) 507 ⫾ 37 (121 ⫾ 9) 457 ⫾ 97 (109 ⫾ 23) 2636 ⫾ 98 (630 ⫾ 24)

2390 ⫾ 32 (571 ⫾ 8)* 612 ⫾ 31 (146 ⫾ 8)† 298 ⫾ 68 (71 ⫾ 16) 3300 ⫾ 77 (789 ⫾ 19)*

* Significantly different from intake in the 100% portion condition (p ⬍ 0.0001). † Significantly different from intake in the 100% portion condition (p ⬍ 0.015). ‡ Tomato topped with pesto, roll, and butter packet were included with purchase of the pasta entre´ e. § Side dishes, desserts, and beverages could be purchased in addition to the pasta entre´ e.


565


Table 3. Survey ratings (mean ⫾ SE) Survey question (answer anchors) Was the entre´ e portion served today appropriate for you? (1 ⫽ “way too small,” 7 ⫽ “way too large”) How would you compare how much you ate today with what you normally consume at lunch? (1 ⫽ “less than normal,” 7 ⫽ “more than normal”) How do you perceive today’s entre´ e as giving you value for your money? (1 ⫽ “poor value,” 7 ⫽ “great value”) How would you rate the entre´ e today? (1 ⫽ “off-putting,” 7 ⫽ “delicious”)

100% Portion (n ⴝ 89)

150% Portion (n ⴝ 91)

4.4 ⫾ 0.1

4.6 ⫾ 0.1

4.7 ⫾ 0.1

5.0 ⫾ 0.1

5.1 ⫾ 0.2 6.1 ⫾ 0.1

5.6 ⫾ 0.1* 6.1 ⫾ 0.1

* Significantly different from rating in the 100% portion condition (p ⬍ 0.05).

When customers rated the amount that they had eaten at the meal (compared with their usual lunch), the rating did not differ significantly between those who had purchased the standard portion and those who had purchased the larger portion. Thus, despite substantial differences in intake, customers in the two groups did not differ in their estimation of the amount of food they had eaten. Regardless of portion size, the rating of amount eaten compared with usual intake was significantly higher (closer to “more than normal”) among female customers than among male customers. No other subject characteristics significantly affected this rating. There was a significant difference in the customer rating of the value of the meal between the two portion size conditions (Table 3). Customers who purchased the 150% portion rated it as offering greater value compared with customers who purchased the 100% portion. The only subject characteristic that affected this rating was age. Regardless of portion size, customers ⱖ35 years old rated the value of the entre´ e higher than the younger customers. Thus, the higher rating of value by the customers who bought the larger portion cannot be attributed to the greater number of younger customers in that group. The rating of the taste of the entre´ es did not differ by condition of portion size or by any of the customer characteristics.

Discussion The results of this study support the hypothesis that large portions in restaurants are associated with increased energy intake. We found that customers who purchased a pasta entre´ e that was 50% larger consumed 43% more energy from the entre´ e and 25% more energy in the entire meal than those who purchased the standard portion. The influence of portion size on energy intake was previously demonstrated in laboratory-based studies (11–14). This study extends these findings to a more naturalistic setting of a 566


commercial restaurant, where individuals selected and paid for the food that they consumed, and to a larger number of participants. The effect of portion size was seen in customers of both sexes and in different categories of age and body size. Thus, as seen in previous laboratory experiments, the effect of portion size on intake seems to be robust and to affect individuals with a range of characteristics. Responses to the customer survey showed that underweight and normal weight subjects who purchased the 150% portion rated it as closer to “too large” compared with those who purchased the 100% portion, whereas overweight and obese subjects rated both portions as equally appropriate in size. The other noteworthy difference in ratings of the two portions was in perceived value, in that customers who bought the 150% portion rated it as having greater value. Thus, the survey responses suggest that some customers had an awareness of the size of the portion in terms of the amount they received for a given cost. Customers seemed unaware, however, of how much food they consumed; despite a 43% difference in food intake, the two groups of customers did not differ in their ratings of the amount of food they had eaten in relation to their usual intake. In a previous laboratory-based study, there was a similar lack of awareness of the amount consumed (11); hunger and fullness were rated similarly after two meals of macaroni and cheese, even though intake increased by 30% when the portion was doubled. Thus, the results of studies in the laboratory and in a restaurant indicate that many individuals are unaware of differences in portion size and that even when they are aware, they still eat more of the larger portion. The portion size of the entre´ e that was sold on a particular day may have affected its purchase by customers with particular characteristics. For example, a customer’s level of hunger may have influenced the decision to purchase the portion that was sold on a given day, but because of the covert nature of the experiment, we could not make this


assessment. In laboratory-based studies, however, the effect of portion size on intake was not dependent on differential hunger at the start of the meal (11). Other customer characteristics related to energy needs or cognitive factors could have affected the decision to purchase a given portion size. The results show, however, that despite the fact that men usually eat significantly more than women, similar numbers of men and women purchased the two portion sizes. Customers of different body size were also equally distributed in the two conditions. We did find that more of the customers who were 25 to 34 years of age and who were graduate students purchased the bigger portion, whereas more of those older than 45 years of age purchased the smaller portion. It should be noted, however, that, despite this difference in purchase pattern, the increase in energy intake from the larger portion occurred in customers of all categories of age, sex, body size, and university affiliation. We found that energy consumption from side dishes, desserts, and beverages did not differ significantly between those who purchased the standard and larger portions of the entre´ e. Thus, it seems that participants did not compensate for the increased intake of the larger entre´ e by reducing their consumption of other foods. Because intake of the other foods was estimated rather than weighed, there is a greater likelihood of error in these data. Consumption of the pasta entre´ e, however, was accurately measured, and this accounted for the majority of the energy intake at the meal. An unexpected finding was that customers who purchased the larger portion of pasta had a greater intake of the entre´ e accompaniments (tomato, roll, and butter), even though the portion size of these items was standard. It is possible that the larger serving of pasta affected the perceived size of the other items on the plate, or it may be that consumption of the accompaniments increased together with the pasta because these foods provided a sensory contrast to the entre´ e. Another possibility is that, on the days that the larger portion was offered in the serving line, it attracted individuals with larger appetites, who, therefore, ate more of the accompaniments. The finding that the intake of side dishes and desserts did not increase in customers who purchased the larger portion and that similar numbers of men and women purchased the larger portion weakens the support for this possibility. A question raised by this study is whether the findings can be generalized to other situations where consumers eat away from home. One previous study has found that the amount of food offered in a naturalistic eating environment affects intake. In that study, participants who attended a movie were given different portions of popcorn (15). The results showed that participants who were served a double portion of popcorn consumed 53% more than the participants who were given a standard portion. Additionally, a recent survey of over 1000 Americans (17) found that 69% of respondents reported that when they dined at a restaurant,

they finished their entre´ es most or all of the time. Of these, 60% described the portion sizes served in restaurants as “just right.” A total of 30% of all consumers reported that they base the amount they eat on the amount they are served. These results suggest that portion size affects intake in a variety of settings encountered by consumers when dining away from home. This study adds to the accumulating body of evidence that large portions of energy-dense foods are associated with increased energy intake and that this intake is contributing to the rising incidence of obesity (6). Laboratorybased studies have shown that the portion size of different types of foods [such as those that are amorphous in shape (11), those that come in discrete units (13), and those that are prepackaged (12)] affects the amount that is consumed by individuals with a variety of characteristics. Furthermore, the effect persists from meal to meal over a 2-day period, with no evidence of compensation for the increased intake (14). As in the laboratory-based studies, the customers in this study ate significantly more when the portion was increased, and their responses to the survey indicated that many were unaware that the portion was larger than normal or that they had eaten more food. Future studies should focus on strategies to moderate the effects of portion size on intake. These may include alterations in the energy density of the available foods, clearer information about the energy content of the foods, and strategies to educate consumers about appropriate portion sizes.

Acknowledgments This work was supported by National Institutes of Health Grant DK59853. We thank Jennifer S. Meengs and other staff and students of the Laboratory for the Study of Human Ingestive Behavior for assistance in conducting the experiment. We also thank The Pennsylvania State University Housing and Dining Services for sharing their facility with us and the School of Hotel Restaurant and Recreation Management for allowing this collaborative study to take place.

1.

2.

3.

4.

References McCrory MA, Fuss PJ, Hays NP, Vinken AG, Greenberg AS, Roberts SB. Overeating in America: association between restaurant food consumption and body fatness in healthy adult men and women ages 19 – 80. Obes Res. 1999;7:564 –71. Binkley JK, Eales J, Jekanowski M. The relation between dietary change and rising US obesity. Int J Obes Relat Metab Disord. 2000;24:1032–9. French SA, Harnack L, Jeffery RW. Fast food restaurant use among women in the Pound of Prevention Study: dietary, behavioral and demographic correlates. Int J Obes Relat Metab Disord. 2000;24:1353–9. Clemens LHE, Slawson DL, Klesges RC. The effect of eating out on quality of diet in premenopausal women. J Am Diet Assoc. 1999;99:442– 4. OBESITY RESEARCH Vol. 12 No. 3 March 2004

567


5. Frazao E, ed. America’s Eating Habits: Changes and Consequences. Washington DC: Economic Research Service, U.S. Department of Agriculture; 1999. 6. Rolls BJ. The supersizing of America: portion size and the obesity epidemic. Nutr Today. 2003;38:42–53. 7. Guthrie JF, Lin B-H, Frazao E. Role of food prepared away from home in the American diet, 1977–78 versus 1994 –96: changes and consequences. J Nutr Educ Behav. 2002;34:140 –50. 8. Young LR, Nestle M. The contribution of expanding portion sizes to the US obesity epidemic. Am J Public Health. 2002; 92:246 –9. 9. Smiciklas-Wright H, Mitchell DC, Mickle SJ, Goldman JD, Cook A. Food commonly eaten in the United States, 1989 –1991 and 1994 –1996: are the portion sizes changing? J Am Diet Assoc. 2003;103:41–7. 10. Nielsen SJ, Popkin BM. Patterns and trends in food portion sizes, 1977–1998. JAMA. 2003;289:450 –3. 11. Rolls BJ, Morris EL, Roe LS. Portion size of food affects energy intake in normal-weight and overweight men and women. Am J Clin Nutr. 2002;76:1207–13.

568


12. Rolls BJ, Roe LS, Kral TVE, Meengs JS, Wall DE. Increasing the portion size of a packaged snack increases energy intake in men and women. Appetite. 2004;42:63–9. 13. Rolls BJ, Roe LS, Meengs JS, Wall DE. Increasing the portion size of a sandwich increases energy intake. J Am Diet Assoc. 2004 (in press). 14. Kral TVE, Meengs JS, Wall DE, Roe LS, Rolls BJ. Effect on food intake of increasing the portion size of all foods over two consecutive days. FASEB J. 2003;17:A809. 15. Wansink B, Park SB. At the movies: how external cues and perceived taste impact consumption volume. Food Qual Prefer. 2001;12:69 –74. 16. Sørensen TIA, Stunkard AJ. Does obesity run in families because of genes? An adoption study using silhouettes as a measure of obesity. Acta Psychiatr Scand Suppl. 1993;370: 67–72. 17. American Institute for Cancer Research. Awareness and Action: AICR Surveys on Portion Size, Nutrition and Cancer Risk. Washington, DC: American Institute for Cancer Research; 2003.