点击显示 收起
1 From the Laboratory for the Study of Human Ingestive Behavior, The Pennsylvania State University, University Park.
2 Supported by grant DK 59853 from the National Institutes of Health. 3 Address reprint requests to BJ Rolls, The Pennsylvania State University, 226 Henderson Building, University Park, PA 16802-6501. E-mail: bjr4{at}psu.edu.
ABSTRACT
Background: Increases in both the portion size and energy density of food have both been shown to increase energy intake, but the combined effects of such increases have not been investigated.
Objective: The objective was to determine the combined effects of energy density and portion size on energy intake in women.
Design: This study used a within-subjects design. Once a week for 6 wk, 39 women were served breakfast, lunch, and dinner ad libitum. The main entrée at lunch was formulated in 2 versions that varied in energy density (5.23 or 7.32 kJ/g), each of which was served in 3 different portion sizes (500, 700, or 900 g). The 2 versions were matched for macronutrient composition and palatability. Breakfast and dinner were standard meals.
Results: Increases in portion size and energy density led to independent and additive increases in energy intake (P <0.0001). Subjects consumed 56% more energy (925 kJ) when served the largest portion of the higher energy-dense entrée than when served the smallest portion of the lower energy-dense entrée. Subjects did not compensate for the additional intake by eating less at the subsequent meal. Despite substantial differences in energy intake, no systematic differences in ratings of hunger and fullness across conditions were observed.
Conclusions: The energy density and the portion size of a food act independently to affect energy intake. The findings indicate that large portions of foods with a high energy density may facilitate the overconsumption of energy.
Key Words: Portion size • energy density • energy intake • food intake • obesity • women
INTRODUCTION
Over the past few decades, the increase in body weight among Americans has coincided with significant changes in the eating environment. Not only has the variety of foods and beverages available to consumers increased significantly, but so have the portion sizes of food at take-out establishments and family-type restaurants (1) and in the home (2). Thus, the easy access to highly palatable, energy-dense foods that are available in large portions at a relatively low cost may be contributing to increased energy intakes and to the growing incidence of obesity. The purpose of the following experiment was to determine how energy density and portion size interact to influence food and energy intakes.
Previous studies have shown that the energy density (kJ/g) of foods significantly affects energy intake in both men and women (3–7). In these experiments the energy density of test meals was varied, whereas the macronutrient composition was held constant. The results showed that the subjects tended to eat a fairly consistent amount of food over the course of several days and that, therefore, their energy intakes were directly related to the energy density of the foods consumed. The subjects had significantly higher energy intakes when they ate meals of a higher energy density than when they consumed meals of a lower energy density.
Recently, many studies have shown that increasing the portion size of foods increases energy intake (8, 9). This effect was seen with the consumption of a single meal consisting of a wide range of foods including amorphous foods (foods with an undefined shape) (10, 11), discrete foods (foods with a clearly defined shape; 12), and prepackaged snack items (13). In another experiment (14), the portion sizes of all foods served over a 2-d period were increased. The results showed that energy intakes increased with the increasing portion sizes, and there was no evidence of meal-to-meal compensation for higher intakes.
It seems likely that it is not portion size alone that is the main contributor to the overconsumption of energy but rather the consumption of large portions of foods high in energy density that facilitates increased energy intakes. The hypothesis of this experiment was that when both the energy density and portion size of a meal are increased, the effects of energy density and portion size on energy intake are independent and will add together to increase energy intake.
SUBJECTS AND METHODS
Subjects
Women were recruited for the study through advertisements in the local university community. Interested persons completed a standard telephone interview to ensure that they met the initial criteria for inclusion in the study: 20–45 y of age, in good health, nonsmoking, not dieting, not in athletic training, not pregnant or lactating, not using medications known to affect food intake or appetite, consuming meals (including breakfast) at regular intervals, no change in body weight (±4.5 kg) in the previous 6 mo, and no food allergies or food restrictions. In this first experiment of the combined effects of energy density and portion size, we included only female subjects. We did this to avoid the substantial increase in variability in food intake that is caused by including subjects of both sexes. This allowed us to keep the number of subjects needed in the experiment within a reasonable limit.
Potential subjects came to the laboratory to complete several screening questionnaires and to have their heights and weights measured. The subjects completed a Demographic and Health Questionnaire and the Eating Inventory (15), which assesses dietary restraint, disinhibition, and hunger. Dietary restraint is defined as an individual's cognitive attempt to consciously limit food intake for the purpose of regulating body weight, whereas dietary disinhibition refers to the loss of the self-imposed cognitive control of eating. Subjects also completed the Eating Attitudes Test (16), which assesses aberrant attitudes toward food and eating; the Zung Self-Rating Depression Scale (17), which measures cognitive indicators of depression; and the Dutch Eating Behavior Questionnaire (18), which measures the degree of external and emotional eating as well as dietary restraint. We accepted in the study normal-weight and overweight women [body mass index (BMI; in kg/m2): 19–29.9] who scored <20 on the Eating Attitudes Test-26, 40 on the Zung Self-Rating Depression Scale, and who were unaware of the purpose of the research conducted in the laboratory. We included subjects with a range of characteristics such as age, BMI, body fat, dietary restraint, and disinhibition to assess whether potential differences in food intake can be related to differences in these individual characteristics.
Before the start of the study, subjects signed a consent form acknowledging receipt of a description of the experimental procedures and participation in the study. To prevent experimental bias, the consent form indicated that the aim of the study was to investigate the effects of food on taste. Each subject received financial compensation after completion of all sessions. The study was approved by the Office for Research Protections of The Pennsylvania State University.
Forty-five women were recruited for participation in the study. Three subjects withdrew from the study before it started, for personal reasons; one subjects did so after her second session. Two subjects were excluded from the analysis because they did not meet the minimum requirements for intake (=" BORDER="0">100 g) and ratings of pleasantness of taste (=" BORDER="0">35 mm) of the manipulated entrée. Thus, a total of 39 women completed the study (Table 1).
View this table:
TABLE 1. Subject characteristics1
Study design
The study used a within-subjects design with repeated measures. The subjects reported to the laboratory for breakfast, lunch, and dinner on the same day of the week for 6 wk. At lunch, subjects received a main entrée that was formulated in 2 versions that varied in energy density (5.23 and 7.32 kJ/g), each of which was served in 3 different portion sizes (500, 700, and 900 g). The 2 versions of the entrée were matched for macronutrient composition and palatability. The order of presentation of conditions was counterbalanced across subjects.
The portion sizes of the entrées were determined on the basis of the intake data from previous experiments. We ensured that the amount of food and calories offered in the smallest portion size of the entrée was sufficient for the subjects to feel full and satisfied after the completion of the meal.
Procedures
The women were instructed to refrain from eating and drinking (except for water) after 2200 the night before each test day, not to consume alcoholic beverages during the 24 h preceding and throughout their test day, and to maintain similar exercise levels throughout the day. On the morning of their first test day, percentage body fat was determined by bioelectrical impedance analysis (Biodynamics model 310e; Biodynamics Corporation, Seattle).
On arrival at the laboratory before each meal, the subjects were seated in individual cubicles. They completed a questionnaire about their physical well-being, intake of medications and alcohol in the previous 24 h, and any food intake since their last meal. The questionnaire was reviewed for compliance with the study protocol; the women who failed to comply had their test day rescheduled. Rescheduling was infrequent and the maximum period between sequential test days was 2 wk.
Breakfast and dinner were standard across conditions. All foods and beverages were consumed ad libitum. A list of foods offered at each meal, including the weight and energy contents of the foods, is available on request from the first author. Lunch consisted of one main entrée that was formulated to vary in energy density and portion size (see Manipulated main entrées). A small amount of compulsory foods (standard in type and amount) accompanied each lunch to add some variety to the meal. The subjects consumed the main entrée at lunch ad libitum and the compulsory foods in full. They were not allowed to request any additional food. All foods were weighed before and after each meal to determine the amount consumed to the nearest 0.1 g. Energy and macronutrient intakes were determined by using manufacturers' nutrition information and food-composition tables (19).
For a given individual, meals were scheduled at the same time on all test days, with a minimum of 4 h between breakfast and lunch and 4.5 h between lunch and dinner. The subjects were instructed to consume only foods and beverages provided by the laboratory on test days. They were allowed to drink water up to 1 h before their meal. At the end of their last test day, the women completed a discharge questionnaire. This questionnaire asked the subjects what they thought was the purpose of the study and whether they had noticed any differences between the test days.
Manipulated main entrées
The main entrée for lunch, an Italian pasta bake, was formulated in 2 versions that varied in energy density (5.23 and 7.32 kJ/g), each of which was served in 3 portion sizes (500, 700, and 900 g). The energy density of the entrées was varied by changing the proportion of vegetables, pasta, and full-fat and reduced-fat cheeses. Entrées were formulated using food composition software (Nutritionist V, v.2.0., First Databank, San Bruno, CA) based on nutrient composition data from manufacturers and food-composition tables (19). Recipes for the main entrées can be obtained by contacting the first author. The ingredients that were used for the main entrées consisted of commercially available national brands. The entrées were assembled in the laboratory by trained research personnel. Each ingredient was weighed on a calibrated scale to the nearest 0.1 g, and cooking times and procedures were carefully standardized.
The lunch entrées were formulated to contain 25% of energy from fat, 60% of energy from carbohydrate, and 15% of energy from protein. To ensure that the entrées were equally well-liked and perceived as being similar in sensory properties such as fat content and saltiness, 2 preliminary sensory tests were conducted in a previous experiment. Results from these sensory tests showed that no significant differences were detected in any of the palatability ratings between the 2 versions of the pasta bake of different energy densities (P >0.05). The subjects who participated in the sensory tests were excluded from participation in this study.
Visual analogue scale ratings
Before each meal was served and again after consumption of each meal, the subjects completed a series of 100-mm visual analogue scales (VAS), rating their degree of hunger, thirst, perception of how much they could eat (prospective consumption), nausea, and fullness. For example, hunger was rated on a 100-mm line preceded by the question "How hungry are you right now?" and anchored on the left by "not at all hungry" and on the right by "extremely hungry." Other questions combined "not at all" and "extremely" with the adjectives "thirsty," "full," and "nauseated." The anchors for the question on prospective consumption were "nothing at all" and "a large amount."
At lunch only, the subjects completed an additional series of 100-mm VAS to assess the palatability of the entrée. Before beginning the meal, the subjects rated their answers to the questions "How pleasant is the appearance of this food right now?" and "How pleasant is the odor of this food right now?". The lines for these 2 questions were anchored with "not at all pleasant" and "extremely pleasant." The subjects were also asked "How does the size of this serving compare with your usual portion?". For this question the line was anchored with "a lot smaller" and "a lot larger." Referring to the first bite and then to the last bite of the main entrée, the subjects were instructed to answer the following questions: "How pleasant is the taste of this food right now?", "How pleasant is the texture of this food right now?", "How much salt do you think this food has?", "How much fat do you think this food has?", and "How much of this food do you think you could consume right now?". For these questions, the lines were anchored with "extremely pleasant" and "not at all pleasant," "no salt (fat) at all" and "extremely high in salt (fat)," or "nothing at all" and "a large amount."
Data analysis
Data were analyzed by using the SAS System for WINDOWS (version 8.1; SAS Institute, Cary, NC). Food intake (g), energy intake (kJ), VAS ratings (mm), and the change in VAS ratings from before to after the meal were analyzed with the use of a mixed linear model (PROC MIXED) with repeated measures. The fixed-factor effects in the model were energy density (5.23 and 7.32 kJ/g) and portion size (500, 700, and 900 g); the subjects were treated as a random effect. The interaction between energy density and portion size was tested for significance before analyzing the main effects of these factors.
The residuals from the model for food and energy intakes at lunch were examined for normality and equal variance. We considered subjects who left <30 g of the main entrée at lunch uneaten to have consumed the entire portion and defined them as "plate cleaners." Mean intakes were analyzed both with and without subjects who cleaned their plate in any condition. The effect of subject characteristics on the main outcome was analyzed by analysis of covariance. Regression analysis was used to identify significant predictors of food intake. The results were considered significant at P <0.05. Tukey's test was used for post hoc pairwise comparisons of means for significant effects. The results are reported as means ± SEMs.
RESULTS
Food and energy intakes
Meal and total energy, food, and beverage intakes across conditions are presented in Table 2.
View this table:
TABLE 2. Energy, food, and beverage intakes from each of 3 meals and over the entire day by energy density (5.23 or 9.32 kJ/g) and portion size (500, 700, or 900 g) of the meal1
Food intake
There was a significant effect of portion size (P <0.0001) and a significant effect of energy density (P <0.0001) on the weight of food consumed at lunch. The interaction between the factors was not significant (P = 0.09), indicating that the portion size and energy density of foods act independently to affect the amount of food consumed. The subjects consumed 20% more food (61.5 g) when served the largest portion of the entrée compared with when served the smallest portion; they consumed 10% less food (36.1 g) when served the entrée of higher energy density compared with that of lower energy density (Figure 1). Food intakes at breakfast and dinner did not differ significantly between conditions (P >0.05). When all 3 meals were considered together, there was a significant effect of portion size (P = 0.0004) and a significant effect of energy density (P = 0.02) on the total daily amount of food consumed (Table 2). There were no significant effects of study week on food intake at lunch (P = 0.42) or on total daily food intake (P = 0.07).
FIGURE 1.. Mean (±SEM) food intakes from the main entrée at lunch for the 3 portion sizes (500, 700, and 900 g) with the 2 energy densities (5.23 and 7.32 kJ/g) combined (A) and for the 2 energy densities with all 3 portion sizes combined (B). The data were analyzed by using a linear mixed model and Tukey's test for post hoc pairwise comparisons of the means. The main effects of energy density and portion size on food intake were significant (P <0.0001), but their interaction was not significant (P = 0.09). Bars with different lowercase letters are significantly different: P <0.03 (for portion size) and P <0.0001 (for energy density).
Energy intake
There was a significant effect of portion size (P <0.0001) and a significant effect of energy density (P <0.0001) on energy intake at lunch. Again, the interaction between the factors was not significant (P = 0.27), indicating that the portion size and energy density of foods acted independently to affect energy intake. Energy intakes at lunch were directly related to the energy density of the main entrée. Despite the lower weight of food consumed in the condition of higher energy density, energy intakes at lunch were 26% (494 kJ) greater when subjects consumed the entrée of higher energy density compared with the entrée of lower energy density. Subjects consumed 20% more energy (377 kJ) when served the largest portion of the entrée compared with when served the smallest portion of the entrée. Thus, the combined effect of energy density and portion size led to a 56% increase in energy intake (925 kJ) when subjects were served the largest portion of the higher energy-density entrée compared with when they were served the smallest portion of the lower energy-density entrée (Figure 2).
FIGURE 2.. Mean (±SEM) energy intake by experimental condition (n = 39). The main effects of energy density and portion size on energy intake were significant (P <0.0001), but their interaction was not significant (P = 0.27). The figure illustrates the combined effects of the 2 experimental variables.
The subjects' energy intakes did not differ significantly at breakfast and dinner across conditions (P >0.05). Furthermore, energy intake at breakfast did not affect the experimental variables and did not significantly affect the main outcome. When all meals were considered together, there was a significant effect of portion size (P = 0.003) and a significant effect of energy density (P <0.0001) on total daily energy intake. This difference in total daily energy intakes across conditions resulted from changes in energy intakes at lunch and was not due to differences in intakes at either breakfast or dinner.
Plate cleaners
Five subjects ate all of the smallest portion size (500 g) of the lunch entrée (4 subjects finished the higher energy-dense entrée, and 1 subject completed the lower energy-dense entrée). When the data were analyzed without these observations, the results for food intake remained unchanged.
Visual analogue scale ratings
Ratings of palatability
Before lunch, there were no significant differences in subjects' ratings of the pleasantness of appearance, taste, or texture, or the perceived salt and fat contents between the lower- and higher energy-density versions of the entrée. The mean taste ratings for the entrées of lower and higher energy density were 73.5 ± 1.5 and 71.4 ± 1.7 mm, respectively, which indicated that both versions were equally well liked. There was, however, a significant effect of energy density on odor ratings before lunch (P = 0.02). The subjects rated the pleasantness of odor significantly higher for the entrée of lower energy density than for the entrée of higher energy density (81.9 ± 1.3 compared with 78.7 ± 1.4 mm, respectively).
After lunch, there was a significant effect of portion size on the change in ratings (rating measured after the meal minus rating measured before the meal) for the pleasantness of taste and texture of the main entrée (P < 0.05). The decrease in the ratings for both taste and texture was greater after consumption of the entrées of larger portion size than after consumption of the entrées of smaller portion size. We also tested whether the change in taste rating was related to the amount of energy consumed. The results indicated that the change in ratings of taste per joule consumed did not differ by energy density or portion size (P >0.05).
Ratings of portion size
There was a main effect of portion size on the subjects' perception of how the size of the serving of food compared with their usual portion. Ratings of the perceived portion size significantly increased as the portion size of the entrées increased (Figure 3). In addition, there was a main effect of portion size on ratings of prospective consumption. The subjects indicated that they could consume a significantly greater amount of the entrée when served the 500-g portions than when served the 900-g portions.
FIGURE 3.. Mean (±SEM) ratings of the perceived portion size of the main entrée at lunch relative to the subjects' usual portion (n = 39). The data were analyzed by using a mixed linear model and Tukey's test for post hoc pairwise comparisons of the means. Bars with different lowercase letters are significantly different, P <0.05. The dotted line indicates a rating equivalent to the subjects' usual portion size.
Ratings of hunger and satiety
There were no significant differences in subjects' ratings of hunger, thirst, prospective consumption, nausea, and fullness across conditions either before or after consumption of breakfast, lunch, and dinner (P >0.05). In particular, despite the differences in energy intake at lunch, the subjects did not rate their hunger or fullness any differently after lunch across conditions.
Subject characteristics
None of the subject characteristics affected the relation between the experimental variables and food intake. Thus, subjects of various ages and body sizes and with different scores for dietary restraint, disinhibition, depression, and eating attitudes responded to variations in the portion size and the energy density of their meals by increasing their food intake.
Regression analysis was performed to identify predictors of food intake at lunch by using the subject characteristics in conjunction with the experimental variables. Significant predictors of food intake at lunch, listed in the order of their predictive value, were as follows: external disinhibition score, dietary restraint score, emotional disinhibition score, and height. External disinhibition, dietary restraint from the Dutch Eating Behavior Questionnaire, and height were positively associated with food intake, whereas emotional disinhibition and dietary restraint from the Three-factor Eating Questionnaire were negatively associated with food intake. Together these variables explained an additional 16% of the variance in the weight of food consumed.
Discharge questionnaire
Only one subject correctly identified that a purpose of the study was to investigate whether the portion size of the lunch entrée affected food intake. Nine subjects (23%) related the purpose of the study either to ratings of hunger and fullness or to ratings of taste or food intake in general. Twenty-nine subjects (74%) had no knowledge or incorrect knowledge about the purpose of the study.
When asked whether they were aware of differences between any of the sessions, 21 subjects (54%) mentioned that they noticed changes in portion size of the lunch entrée; 2 subjects thought incorrectly that the portion sizes at dinner had also changed. Eight subjects reported noticing changes in the composition of the pasta bake, and 3 subjects reported noticing differences in the taste and flavoring of the pasta bake. Ten subjects (26%) did not report noticing any differences between their test days. The effect of portion size and energy density on energy intake was the same regardless of whether the subjects noticed portion-size differences in the lunch entrée.
DISCUSSION
The findings of this study showed that when both the energy density and the portion size of a food are increased, both factors act independently to affect energy intake. Thus, when increased simultaneously, the effects of energy density and portion size add together to affect energy intake. The subjects consumed 925 kJ more energy when served the largest portion of the lunch entrée of higher energy density compared with when served the smallest portion of the lunch entrée of lower energy density.
It was of interest to see whether the additional energy that was consumed at lunch was compensated for by reducing intake at dinner. The results showed that the subjects consumed the same amount of energy at dinner across conditions despite substantial differences in lunch energy intakes. Therefore, when lunch energy intake increased, total daily energy intake increased as well. This finding is similar to the results from a recent study, which investigated the effect on energy intake of increasing the portion size of all foods over 2 consecutive days (14). The results from that experiment showed that when the portion size of all foods consumed over 2 d was doubled, the cumulative difference in energy intake amounted to 6694 kJ for men and 4184 kJ for women by the end of the second day. Thus, when the portions served were 100% larger, daily energy intakes increased by an average of 26% (2218 kJ for women and 3360 kJ for men). Most importantly, there was no evidence of any meal-to-meal compensation for the increased energy intakes when larger portions were served. Hence, it seems plausible that various environmental factors, one of which is portion size, may override an individual's ability to self-regulate energy intake (8).
Previous studies (3, 20, 21) of the effects of energy density on intake have shown that persons tend to eat a fairly consistent weight or volume of food over the course of several days, regardless of the energy density of their meals. Therefore, the finding in this study that subjects had lower intakes of entrées of higher energy density differs from the findings of some previous studies. A similar disparity in intake between entrées of different energy densities, however, was found in one previous experiment (6) in which the same type of entrée (pasta bake) was served as one of the test meals. This suggests that despite the fact that entrées of both energy densities had similar ratings for taste and texture, differences in other sensory properties of the 2 versions could have led to changes in intake. For instance, subjects rated the odor of the lower energy-dense entrées to be significantly more pleasant than that of the higher energy-dense entrées. It is possible that the different amounts and kinds of cheeses used in the 2 versions of the pasta bake may have accounted for the differences in the odor ratings. Other differences in the sensory properties of the 2 versions of the entrées may have remained undetected by the measures used in this experiment. It is known that the sensory characteristics of foods directly influence food intake (22).
It is also possible that the finding of lower intakes of the entrées of higher energy density may be attributed to energy compensation. At higher energy densities, individuals may sense the greater energy content in a given portion of food. In an attempt to reduce their caloric intake, or because of the initiation of satiety signals, persons may reduce the amount of food consumed. It, however, needs to be pointed out that the reduction in food intake was small, amounting to 36 g (262 kJ), and therefore did not negate the effect of the higher energy density on energy intake.
The significant effect of portion size, but not energy density, on change ratings for the pleasantness of taste and texture from before to after lunch may be related to sensory-specific satiety. Sensory-specific satiety refers to the decrease in the pleasantness of a food as it is consumed, whereas that of other foods remains relatively unchanged. In a recently published study conducted by Bell et al (23), the results showed that doubling the volume, but not the energy content, of a liquid food significantly decreased the ratings of pleasantness of taste and, thus, increased sensory-specific satiety. This suggests that the volume of food consumed has a greater effect on perceptions of a food's pleasantness of taste than does its energy content. The same may be true for the amount (weight) of a solid food consumed, as seen in this experiment.
The subjects' ratings of portion size in relation to their usual portion indicated that they did notice differences in the size of the entrées. Thus, although subjects rated all 3 portion sizes as larger than their usual portion and noticed the differences in the size of the entrées, they still responded to the increasing portion size by increasing their intake.
The fact that none of the subject characteristics affected the relation between the experimental variables and food intake suggests that females with a range of individual characteristics are affected by changes in the portion size and the energy density of foods. We found that subjects' scores for external and emotional disinhibition, and their scores for dietary restraint were significant predictors of food intake. Thus, the degree of dietary disinhibition and restraint predicted how much food an individual consumed at lunch. It is possible that the presentation of larger portion sizes to subjects who scored high for dietary restraint or disinhibition represented an environmental trigger that counteracted subjects' self-imposed cognitive control over food intake. Clearly, more research is needed to analyze the possible influence of these individual characteristics on food intake in response to dietary manipulations of portion size and energy density.
The results from the present experiment are likely to have important implications for weight management. Increasing the portion size and the energy density of a main entrée resulted in significant increases in energy intake during a single meal. There was no evidence of significant energy compensation occurring at the subsequent meal for the increased energy intake at lunch. Therefore, foods that are of a higher energy density and served in larger portions are likely to facilitate the overconsumption of energy. So far, the long-term effects of habitual consumption of such foods on energy intake and body weight regulation have not been established. The energy density of foods, however, has been identified as a key determinant of energy intake. Cognitive, behavioral, and sensory cues related to the weight or volume of food consumed are known to affect food intake regulation (24, 25).
Few data are available that relate portion size and energy density of foods to weight status. In an analysis of food diaries, Westerterp-Plantenga et al (26) found that obese women took larger portions of food with a higher energy density than did nonobese women. Obese women also consumed 24% of their total energy intake from foods of the highest energy density category and 24% of their total energy intake from foods of the lowest energy density category as opposed to 13% and 38% consumed by nonobese women, respectively. Thus, it is possible that the regular consumption of large portions of high energy-dense foods may be one of the factors contributing to the increasing prevalence of obesity in Western nations.
There are several ways in which this research needs to be extended. This study included normal-weight and overweight women only. It will be important to broaden the subject population and also to conduct the experiment in males and obese subjects of both sexes. In addition, the dietary manipulation was carried out with a single food only. It will be of interest to investigate the effects on food intake when the energy density and portion size of more than one meal is manipulated. Last, to increase the ecologic validity of the study, we suggest that this type of experiment be conducted in real-life settings.
In conclusion, the results from this experiment provide evidence that the energy density and the portion size of foods add together to independently affect energy intake. This association is of particular importance in that it identifies the availability of large portion sizes of energy-dense foods as a significant contributor to increased energy intakes. The understanding of this relation can aid in the development of dietary strategies to prevent and treat obesity.
ACKNOWLEDGMENTS
TVEK contributed to the study design, food formulation, data collection, data analysis, and writing of the manuscript. LSR contributed to the study design, data analysis, and writing of the manuscript. BJR contributed to the study design and the writing of the manuscript. None of the authors had any financial or personal interest in the organization sponsoring the research. BJR is a past member of the Advisory Council of the National Institute of Diabetes and Digestive and Kidney Diseases.
REFERENCES