Super Bowls: Serving Bowl Size and Food Consumption

To the Editor: Obesity has been linked in part to the expanding portion sizes of prepackaged or preserved foods.^1-3 However, adults frequently serve themselves the food they will eat for a meal or snack. We investigated how the size of serving bowls influences how much food a person decides to serve and consume in a natural environment.

Graduate students were recruited to attend a Super Bowl party at 5:30 PM. On arrival, 40 individuals orally consented to participate in an institutional review board–approved study in which they “may be asked questions about food and commercials in party environments, such as at a Super Bowl party.” No reference was made to the hypotheses being examined. Each participant was led in an alternating order to 1 of 2 identical buffet tables on opposite sides of an adjoining room and asked, “Would you care for some snacks before the game?” – had identical amounts of high energy density snacks (assorted roasted nuts and a pretzel/chip variety mix, both approximately 5.26 kcal/g). On one table these snacks were offered in 2 large (4-L capacity) serving bowls. On the second table, an equal quantity of the same 2 snacks was offered in 4 medium (2-L capacity) serving bowls that were otherwise identical to the larger bowls.

Participants served the snack mix to themselves on 10-inch plates using a 1-cup serving scoop. Immediately following their self-serving, a research assistant blinded to the purpose of the study weighed the plate. The participant was then led to a table to watch the game. Beverages were not available until after the food was served. After the last participant had taken a serving, all food was removed. One hour later, each participant completed a questionnaire. Each plate was collected and plate waste was recorded and subtracted from how much each had initially taken.

Participant characteristics between the study groups were compared using t tests. The volume of food served and eaten was analyzed with a 2-way analysis of covariance, which used bowl size and sex as factors between participants. Body mass index (calculated as weight in kilograms divided by the square of height in meters), body weight, hours since prior meal, age, and education were used as covariates. A power analysis indicated a power of 0.91 for detecting an effect size of 0.50 at the 5% confidence level with a sample of 20 in each cell.

Five participants (1 woman in the large-bowl group; 3 women and 1 man in the small-bowl group) did not take a snack and were not included in the primary analyses. A sensitivity analysis was conducted that assumed the nonparticipant in the large-bowl group had consumed the least amount recorded among the women in that group, and that the nonparticipants in the small-bowl group had consumed the greatest amount recorded among same-sex participants in that group.

All analyses were performed using SPSS statistical software (version 11.0, SPSS Inc, Chicago, Ill). A P value <.05 was considered statistically significant.

Comparing the study groups, the participants who took snacks were similar with respect to their body mass index, weight, hours since their prior meal, age, education, and sex (Table 1). There were no significant differences in these characteristics between these participants and the 5 persons who did not take a snack.

Participants serving from large bowls took 53% (146 calories) more and consumed 56% (142 calories) more than those who served from small bowls (P = .02 and P = .01, respectively) (Table 2). The effect of serving bowl size on intake was not significantly influenced by body weight (P = .22), hours since prior meal (P = .25), age (P = .20), or education (P = .71). Men ate more than women (P = .04), and the effect of serving bowl size on consumption was statistically significant for men (P = .02) but not women (P = .17).

In the sensitivity analysis to estimate the potential impact of the 5 nonparticipants, the effect of bowl size remained significant (P = .02).

Small environmental factors can have a large influence on food consumption.⁴ At this party, large serving bowls led to a 56% greater intake (a mean of 142 more calories/person). The size of a serving bowl (or of a portion) may provide a consumption cue that implicitly suggests an appropriate amount to eat.⁵ Larger bowls, like larger packages or portions, may suggest that a proportionately larger amount is appropriate to consume. Although this study was not conducted in a medical setting, it is possible that if a physician giving diet-related advice recommends using smaller serving bowls, patients may serve themselves smaller portions.

Portion distortion has generally focused on how consumption cues lead people to overeat less healthy, energy-dense foods. An appropriate area for further research is whether these same cues, ie, larger serving bowls, can be used to encourage people to eat greater amounts of healthier foods such as fruits and vegetables.

Letters Section Editor: Robert M. Golub, MD, Senior Editor.

Retraction: A notice of Retraction was published on September 19, 2018; and a notice of Expression of Concern was published on April 13, 2018.