Reshaping Food Choice at University of Michigan

NCT05131529 | Completed

• 1 other identifier: HUM00198619
• Study Type: interventional
• Target: 3,003
• Locations: 1 country
• Sites: 1

Brief Summary

Nearly a quarter of global greenhouse gas emissions originate from agriculture. Consumer demand for food, especially carbon-intensive red meats, largely underlies these emissions. Therefore, reducing consumption of red meat has the potential for transformative impacts on climate change. Our interdisciplinary team of researchers and dining staff aims to determine the carbon impact of evidence-based behavior change interventions to reduce red meat consumption at University of Michigan dining halls. Using an experimental design, we will evaluate the impacts of three "nudge" interventions (i.e., modifying environmental cues and incentives) to reduce red meat consumption in three randomly assigned treatment dining halls vis-à-vis three paired control dining halls. Nudges will include changing default food items and altering food labels. We will: 1) collect data on meals served during each intervention periods to assess the separate and combined effects of the interventions on red meat consumption; 2) conduct e-mail interviews with students to understand treatment effect heterogeneity and sociodemographic determinants of food choice; and, 3) conduct focus group discussions and interviews, respectively, with students and dining staff to assess implementation feasibility. We expect that this research will generate scalable, replicable solutions for reducing greenhouse gas emissions through dietary change at universities and similar institutions, will significantly advance the food choice literature, and will influence sustainability strategies at university dining operations nationally given the proposed team's close partnership with regional and national dining programs.

Conditions

Climate Change; Food Preferences

Outcome Measures

Primary Outcomes (3)

• The difference in the change in the average amount of red meat served in treatment vs. control dining halls from the pre-intervention to intervention period based on service data from dining hall operations staff

  We will assess the amount of red meat served across project dining halls from the the pre-intervention to the intervention period using "post-cost" data collected by operations staff (i.e., the amount of each dish prepared prior to the start of a meal period as well as the amount of each dish remaining at the end of the meal period with units of measure standardized to kg. We will assess the average weight of red meat served in each dining hall during the pre-intervention baseline period compared to the average weight of red meat served in each dining hall during the intervention period. We will do a difference-in-difference analysis to identify whether the change in the amount of red meat served at treatment vs. control dining halls differed during this time period. Difference-in-difference analyses examine the difference in the change in an outcome between a treatment and control group.

  11 week intervention period (during each of 3 academic semesters)

• The difference in the change in the average amount of red meat purchased by dining halls in treatment vs. control dining halls from the pre-intervention to intervention period

  We will assess the average amount of red meat purchased by each of the 6 project dining halls during the pre-intervention period and during the intervention period. We will do a difference-in-difference analysis to identify whether the change in the amount of red meat purchased at treatment vs. control dining halls differed during these time periods.

  11 week intervention period (during each of 3 academic semesters)

• The difference in the change in the average amount of red meat consumed in treatment vs. control dining halls from the pre-intervention to intervention period based on 30-day food frequency questionnaire intake

  We will assess the amount of red meat consumed by a subsample of dining hall patrons (n = 1,000) at each of the 6 project dining halls from the start to the end of each intervention period. This will be done through a self-administered food frequency questionnaire that asks respondents to report their frequency of consumption of red meat items in the previous 30 days. We will do a difference-in-difference analysis to identify whether the change in the amount of red meat consumed by respondents who frequently patron treatment vs. control dining halls differed during this time period.

  11 week intervention period (during each of 3 academic semesters)

Secondary Outcomes (2)

• The change in the average amount of red meat served at meal stations in treatment vs. control dining halls from the pre-intervention to intervention period based on service data from dining hall operations staff

  15 weeks (4 week baseline period + 11 week intervention period (during each of 3 academic semesters)

• The change in the average amount of red meat purchased by dining halls in treatment vs. control dining halls from the pre-intervention to intervention period

  15 weeks (4 week baseline period + 11 week intervention period (during each of 3 academic semesters)

Study Arms (2)

Control

NO INTERVENTION

The 3 dining halls allocated to this arm will receive none of the three interventions during the 3 semesters of the project.

Treatment

EXPERIMENTAL

The 3 dining halls allocated to this arm will receive one intervention during each of the 3 semesters of the project (only one intervention per semester).

Other: Taste-focused labels; Other: Carbon impact labels; Other: Changing default menu options

Interventions

Taste-focused labels OTHER

We will implement taste-focused labels on plant-based dishes in each of the 3 treatment dining halls to re-brand dishes with attributes focused on the taste of the dish (e.g., renaming a "Squash with Ginger" side to "Aromatic Thai Curry Kabocha Squash With Zesty Ginger").

Treatment

Carbon impact labels OTHER

We will add labels next to all dishes served in treatment dining halls that indicate whether production of the ingredients in the dish contribute a high, moderate or low amount of carbon emissions. These labels will be absent from the same dishes served in control dining halls. Labels will be added to point-of-service signs, digital menu boards, and menu information on a mobile app.

Treatment

Changing default menu options OTHER

We will make red meat items an opt-in (as opposed to opt-out) item across all meal stations in treatment dining halls. All menu stations will exclude red meat by default, and dining hall patrons will have to request the addition of red meat to meals. At control dining halls, red meat items will be provided by default at meal stations as normal.

Treatment

Eligibility Criteria

• Age: 18 Years+
• Sex: all
• Healthy Volunteers: Yes
• Age Groups: Adult (18-64), Older Adult (65+)

Student Survey: * An individual must have been a frequent diner at one of the six project dining halls during the first three weeks of the academic term * An individual must have attended only treatment or only control dining halls for 90% or more of their dining hall visits during the first three weeks of the academic term Student Focus Group: * Must have participated in the student survey * We will stratify the sample by selecting respondents whose red meat consumption either increased, decreased or remained the same (based on baseline to endline comparisons of data in the student survey) Dining Staff Interviews: \- Must have been a non-student member of the Michigan Dining staff at one of the six project dining halls during the relevant intervention period.

Contact the study team to discuss eligibility requirements. They can help determine if this study is right for you.

Sponsors & Collaborators

• University of Michigan: lead
• Graham Sustainability Institute of the University of Michigan: collaborator

Study Sites (1)

University of Michigan

Ann Arbor, Michigan, 48109, United States

Location

MeSH Terms

Conditions

Food Preferences

Condition Hierarchy (Ancestors)

Feeding Behavior; Behavior

Study Officials

• Andrew D Jones, PhD

  University of Michigan

  PRINCIPAL INVESTIGATOR

Study Design

• Study Type: interventional
• Phase: not applicable
• Allocation: RANDOMIZED
• Masking: NONE
• Masking Details: Participants are not randomly allocated to treatment groups, but rather, dining halls are randomly allocated to treatments. This is what "Randomized" refers to under "Allocation" below. Under "Enrollment" below, the number of participants refers strictly to the number of participants in an e-mail survey that will go out to students during each of the three semesters of the project. We anticipate 1,000 respondents during each semester. These respondents do not include the entire population of dining hall patrons that will interact with the dining halls (and therefore the interventions) during the three semesters of the study.
• Purpose: PREVENTION
• Intervention Model: SEQUENTIAL
• Sponsor Type: OTHER
• Responsible Party: PRINCIPAL INVESTIGATOR
• PI Title: Associate Professor

Model Details: We will carry out the proposed project within six of the University of Michigan's MDining's nine residential dining halls on the Ann Arbor campus. Using a cluster-randomized controlled trial, we will create three matched pairs among the six dining halls based on the size of each facility. One dining hall from each of the pairs will be randomly assigned to a treatment and control group, respectively. Participants will receive interventions based on which dining halls they visit regularly and most often.

Study Record Dates

• First Submitted: November 1, 2021
• First Posted: November 23, 2021
• Study Start: August 30, 2021
• Primary Completion: December 18, 2022
• Study Completion: December 18, 2022
• Last Updated: January 12, 2023

Record last verified: 2023-01

Data Sharing

• IPD Sharing: Will not share

Locations

US (1)