Effects of Longer Seated Lunch Time on Food Consumption and Waste in Elementary and Middle School–age Children

Key Points Question What is the effect of longer seated lunch time on children’s consumption and waste of lunches that were prepared according to the National School Lunch Program nutrition standards? Findings In this crossover trial of 241 lunch observations from 38 children, 20 minutes of seated lunch time were associated with significantly more consumption and significantly less waste of fruits and vegetables compared with 10 minutes of seated lunch time. Entree and beverage consumption was not significantly different between the 10-minute and 20-minute seated lunch conditions. Meaning Results from this study support policies that require 20 minutes of seated lunch time for school-age children, which could lead to better diet quality and reduced food waste.

June 28 D 10 minutes *There are 5 different menus (labeled A-E). Each menu was served 4 times over the course of the study, consisting of two menu-lunch time dyads occurring during the study period. The lunch time was randomly determined for the first and third occurring menu date, and the alternate lunch time was assigned to the second and fourth occurring date accordingly.
Treatment Arms: Due to logistics, there will be only one sequence of menu and length of lunch throughout the study.

STUDY RECRUITMENT AND STUDY POPULATION
Recruitment: Study participants will be recruited from a university summer camp. All children who sign up for a university summer camp will be invited to participate in the study. Lunch will be provided for free to all children as a benefit, regardless of study participation. There were no additional incentives to participate in the study.
Parents will be sent information about the study, a consent form, and demographic form electronically when signing up for the summer camp. Parents can complete the consent form electronically or bring a hard copy the first day of camp. Parents will also fill out a demographic information form for their child along with the consent form. During the first day of camp, children will provide written assent to participate on the first day of camp each week.
Children can sign up for 1-4 weeks of camp and therefore were not expected to participate in every study day.

Menu Plans:
The five menus to serve throughout the study (serving size and NSLP meal component): Grilled chicken sandwich on a whole grain bun (2 oz meat, 1-2 oz grain) OR Whole grain pasta w/ marinara sauce (2 oz grain) and meatballs (2 oz meat) Menus will conform to the National School Lunch Program nutrition standards. Food will be prepared by a University chef with assistance from graduate and undergraduate student researchers in a commercial kitchen. All persons involved in preparing meals had a certificate of food handler training. Temperature monitoring and other food safety procedures were strictly followed during food preparation and service. Menus are designed using MOSIAC software and featured the same foods served to the local school district using the same distributor when possible.

Training of Research Staff:
Research assistants participate in several trainings prior to the start of the study. The first training taught researchers how to complete their respective roles for data collection (e.g., observational data, food waste measurement) and allowed them to practice with each other. The second training consisted of a complete rehearsal of an entire study, including meal production and reference weight collection, participant observation of seated lunch time and social behaviors, and measuring wasted food. During the second training, table layout and placement was refined so that observers could have an unobstructed view of all participants. All research staff also complete CITI ethics trainings and, for research assistants involved in food preparation, food handler training.

Lunchroom Setup and Layout:
Tables are spaced around the lunchroom so that observers are at their own tables with clear sight lines to participants. (Lunchroom layout shown below). Additional observers may stand along the walls (wherever allows clear sightlines) of the room to provide extra observations of participants so that observers reliability can be assessed. Seated observers are provided smaller portions of the same food served to participants to make their observations less noticeable to the participants.
Once participants exit the lunch line, they are permitted to sit at whichever open seat they choose but are not allowed to move chairs.

Study Protocol for:
Time for Lunch: The Impact of Lunch Time Constraints on Child Eating Behaviors 5 Food Preparation: For weighing pre-consumption standardized reference weight, five extra servings of each food item will be made, portioned out via standard sized serving utensils, and weighed to the nearest 0.5 grams. The weights of these five samples will be averaged. This average weight will be used as the reference weight.
Foods that are individually packaged (e.g., chicken sandwiches) and have larger weight variation range will be weighed and labeled individually in lieu of reference weights. These items will be marked with a sticker (labeled A-Z) that will be used to track the item's specific weight. A sticker weight log will be used to keep track of stickers and their corresponding weights so that this information will not be visible to participants.
For all foods, inedible portions (cores, peels, etc.) are extracted before the reference weight is measured.
Participants were also offered a healthy snack prepared by research staff at the same time each morning and did not participate in any physical activity before lunch each day. Lunch took place inside the Spice Box, which is a dining space within the department of Food Science and Human Nutrition at the XXXX. At 12:30 pm each day of camp, participants would enter the Spice Box for lunch and immediately be informed if there would 10 or 20 minutes of seated lunch time. Participants were then directed to enter the lunch line, which was located in the adjacent commercial kitchen, very close to the Spice Box entrance.

Study Protocol for:
Time for Lunch: The Impact of Lunch Time Constraints on Child Eating Behaviors 6 Lunch Period: At the start of lunch, participants will form a line to enter the food serving area. Once participants are lined up, a research staff member will make an announcement about the length of seated lunch time for that day. Participants will retrieve their participant ID card and then move through the lunch line.
All food items will be pre-portioned, meaning there will be no option to select portion sizes of items. Children will be instructed to select menu items that qualify as a reimbursable lunch as defined by the National School Lunch Program. Thus, participants are required to select three different components and one must be a fruit or vegetable. (All entree items constituted two meal components: grain and protein). A researcher will check each participant's tray to ensure they have a reimbursable meal after exiting the serving line.
While selecting food items in the lunch line, participants will verbally ask for the food items they want, and research staff will serve the pre-portioned meal items (offer vs. serve format of the National School Lunch Program). Foods will be pre-portioned on small plates, food boats, or bowls. Food components that are hot will be served first due to the arrangement of the buffet equipment.
As participants exit the lunch line, a pre-consumption photo is taken of their selected food items. An ID card with the participant ID number will be included in each tray photograph.
Seated lunch time will officially begin when the last participant makes it through the line and sits down. At this time, a researcher will announce the exact time that lunch ends and will write the time on a poster. The researcher will make an additional announcement when five minutes remain for lunch.
Participants will not be allowed to get seconds on food items, however participants are allowed refills of water. Participants can also get up for condiments and utensils during lunch.
At the end of lunch, a final announcement is made for participants to walk their trays to the corner of the room to collect tray items for weighing. 6. Cover sheet: student researchers will fill out a cover sheet for each study day that includes information about weather and what time lunch started and ended each day.

Data Collection During Lunch
Data Collected Immediately Following the Lunch Period: 1. Food preference surveys: after handing in their trays, participants will fill out a post-meal survey indicating their preferences on how the food looked and tasted.
2. Food Waste Measurement: A post-consumption photo of each tray will be taken at the end of lunch prior to weighing food items. Each food component (entree, fruit, vegetable) remaining on the tray as food waste will be weighed separately to the nearest 0.5 gram after the lunch period and recorded on weight sheets. Inedible portions of food items will not be weighed (e.g. food bowls, napkins). Beverage waste remaining on the trays will be recorded to the nearest 0.5 mL using a graduated cylinder.

Measures:
Sample Size: The sample size for this study was fixed based on the summer camp enrollment. A sample size or power calculation was not conducted due to this restraint.

Seated Lunch Time:
Seated lunch time will be measured in two ways in this study. The first measure will be a categorical variable that indicates the randomly assigned seated lunch time: 10 minutes or 20 minutes. There will be a total of 10 10-minute days and 10 20-minute days throughout the study.
The second measure will use the actual time a participant is seated eating lunch. This continuous variable will be calculated using the actual observed seated lunch time. The total seated lunch time is calculated as the time a participant stood up at the end of lunch minus the time a participant sat down at the beginning of lunch. Additional non-seated time recorded on the observer sheet will be subtracted from this lunch time to account for time a participant may have gotten up after initially sitting down.
Inter-rater reliability was assessed for 27.4% of trays during the study that had multiple observers for the same participant. The inter-rater reliability for both the length of lunch and non-seated time had excellent agreement 1 as measured by the intra-class correlation coefficient The nutritional profile of all foods served will be determined using the Nutrition Data System for Research. For each food item, nutrients consumed and nutrients wasted will be calculated as follows: Nutrients consumed: Percent consumed*reference nutrient (where the reference nutrient is the pre-consumption value) Nutrients wasted: Percent wasted*reference nutrient (where the reference nutrient is the preconsumption value) Nutrients will be calculated for individual meal components as well as the total meal. The total meal nutrients will be a sum of the nutrients consumed/wasted from each meal component.

Observed Social Behaviors:
Talking and phone use ratings from the observer sheets will range from a scale of zero to five during the first part of lunch and the last five minutes. An overall rating during lunch will be calculated as the average of the rating for the first part of lunch and the last five minutes.
Inter-rater reliability was assessed for 32.8% of trays during the study that had multiple observers for the same student. The inter-rater reliability for both talking and phone use had excellent agreement 1 as measured by the intra-class correlation coefficient: Talking: 0.954 (95% CI: 0.932, 0.967) Phone use: 0.967 (95% CI: 0.954, 0.976)

Analysis Plan:
Descriptive statistics: Descriptive statistics will include a count of participants and total trays, frequencies of categorical demographic information, and mean and standard deviation of continuous demographic information.
Selection: Percent of participants who select each food item overall and broken down by 10-minute and 20-minute seated lunch time. Chi square tests can compare the differences in selection distribution between 10-minute and 20-minute seated lunch time .
Model parameters: Mixed effects linear regression models will assess all continuous outcome measures and will include the following: Random effects: participants (to control for repeated measures) Fixed effects: menu (5 levels), prior menu exposure (4 levels), average talking rating, average phone use rating. For amount wasted and all nutrient models: an additional fixed effect of the reference weight/volume or reference nutrient (to control for variability of available food/nutrients preconsumption) For models that include the categorical time (10-minute and 20-minute seated lunch time), the estimated marginal means (emmeans) will be used to compare the mean outcome for 10-minute and 20-minute seated lunch time. The contrast between the means (10 minute-mean minus 20-minute mean) will also be reported along with a p-value. Models that used continuous time variable will assess the beta coefficient, t-value, and p-value. Beta coefficients will represent the change in outcome for a one minute increase in seated lunch time.