UCLA

Current dietary patterns threaten human and planetary health. In the United States, individuals must shift to dietary patterns higher in plant-based foods and lower in animal-based foods to reduce chronic disease risk and maintain stability of the Earth system. Despite high scientific agreement that we can simultaneously improve health and environmental sustainability through dietary shifts, interventions targeting these dual outcomes remain understudied. This dissertation employed a community-based participatory research approach to investigate how academics and non-academic foodservice leaders can collaborate to address gaps in the development, implementation, and evaluation of interventions to promote healthier, more environmentally sustainable diets. Research was focused on the university setting and took place at the University of California, Los Angeles.

Guided by the diffusion of innovation framework, Study One qualitatively described and examined the process of developing and implementing the Impossible™ Foodprint Project—an intervention to reduce animal-based protein consumption in university dining. Intervention components included: 1) the addition of new menu items with Impossible™ plant-based meat, and 2) a complementary social marketing campaign framed around climate change. Findings from Study One highlight the value of the university’s involvement in existing health and sustainability initiatives for intervention agenda-setting and collaboration among academics and non-academic partners. In addition, results suggest university foodservice leaders may be particularly open to strategies such as piloting new menu items and providing education—rather than taking existing menu options away. Furthermore, co-creation of intervention materials and feedback from multiple data sources enhanced capacity for foodservice leaders to expand efforts to promote low-carbon-footprint foods. Lack of coordination with restaurant operators emerged as a barrier to initial implementation of the social marketing campaign, while cost prevented scale-up of Impossible™ menu items beyond the pilot intervention restaurant.

Study Two utilized routinely collected sales and nutritional data from FoodPro, a widely used foodservice data management platform. A natural experiment with a pre-post non-equivalent comparison group design was used to evaluate 1) whether the Impossible™ Foodprint Project intervention met foodservice leaders’ goal of reducing animal-based entr�e sales, and 2) the impact of the intervention on the healthfulness and environmental sustainability of entr�es sold. The analytic sample included 645,822 entr�es sold at the three study sites during the Fall 2018 (pre) and Fall 2019 (post) academic quarters. During the post period, new menu items with Impossible™ plant-based meat comprised over 11% of entr�e sales at the intervention site. At the same time, the proportion of animal-based entr�e sales decreased by 9% (raw change 7%, 83% to 76%), a significantly greater decrease than the two comparisons sites.

Healthfulness was operationalized as a decrease in the proportion of red meat entr�es sold and improvement in the nutritional quality of entr�es sold. While the proportion of red meat entr�es sold significantly decreased by about 8% at the intervention site (raw change 4%, 45% to 41%), a similar decrease was observed at one of the comparison sites, resulting in an unclear intervention effect. Small but statistically significant nutritional changes were observed at the intervention site: On average, each entr�e sold contained 21.3 fewer calories (kcal) and lower quantities of nutrients of concern: 0.2 fewer g saturated fat and 26.9 fewer mg sodium. Quantities of other nutrients also decreased: 0.7 fewer g protein, 0.1 fewer g fiber, and 1.5 fewer g unsaturated fat. However, nutritional outcomes varied when stratifying by entr�e type (i.e., build-your-own vs. special), resulting in a conditional assessment of the intervention’s nutritional impacts, described within.

Environmental sustainability was operationalized as reduction in climate impact level (low, medium, high) and carbon footprint of entr�es sold. There were clear positive intervention effects on these outcomes. For example, the proportion of low-impact entr�e sales increased by over 50% at the intervention site (raw change 7%, 14% to 21%), a significantly greater increase than the two comparison sites. This corresponded with an 8% decrease in the mean carbon footprint of each entr�e sold at the intervention site, from 1,522 to 1,405 g CO2-equivalent (117 g decrease). With 141,321 entr�es sold at the intervention site in Fall 2019, this equates to approximately 16.4 metric tons of CO2 saved—the equivalent of driving 42,000 miles.

In line with foodservice leaders’ priorities, we also conducted a brief customer survey (n=215). Results suggest a diverse range of students was open to trying the new Impossible™ menu items, and customer satisfaction was high. In comparing one-time versus repeat consumers, we found significant differences across most behavioral and cognitive factors measured. In general, repeat consumers reported consuming less animal-based protein and were more likely to believe Impossible™ is delicious and a satisfying alternative to animal meat. We also found evidence that values and race/ethnicity may affect beliefs about the sensory experience of eating Impossible™, which in turn affects repeat consumption.

Finally, Study Three utilized a true experiment through Qualtrics to test whether environmental sustainability framing is more effective than health framing in “nudging” university consumers to choose a plant-based menu option. Participants were randomly assigned to one of three menu framing conditions—control (no framing), health framing, and environmental sustainability framing—and given the choice between chicken enchiladas and plant-based tacos. Of the 450 participants recruited for the study, 437 were maintained in the analytic sample, including 352 (79%) undergraduate students and 85 (21%) university staff. There were no statistically significant differences in choice across menu framing conditions. Approximately 39% of participants chose the plant-based tacos in the control condition, 36% in the health framing condition, and 40% in the environmental sustainability framing condition. In short, we found no main or conditional effects of environmental sustainability framing, compared to control. In contrast, we found some evidence that, compared to control, health framing may have negative effects among some subgroups, including university staff. Despite observed null effects of environmental sustainability framing, this approach may still be preferable to health framing given potentially counteractive health framing effects. In ancillary analyses described within, we found that, compared to health framing, environmental sustainability framing may improve anticipated enjoyment of a plant-based dish—even if it does not affect choice.

In sum, Study One sheds light on how and why interventions take shape, with an emphasis on collaboration between academic and non-academic foodservice partners. Study Two provides novel insight into the benefits and tradeoffs of promoting low-carbon-footprint foods and introducing new plant-based meat alternatives into institutional food environments. Experimental findings from Study Three suggest some nudges may be insufficient to affect choice of a plant-based menu item, while others may be counteractive. Taken together, results of this dissertation build capacity for academics and foodservice leaders to advance intervention action and research to improve human and planetary health through food.