Title: Development of personalized healthy food incentives to improve diet and cardiovascular risk
ABSTRACT
It is imperative to address the complex factors that drive food decisions and promote unhealthy dietary
patterns. This mentored career award will support a rigorous training and research plan that will apply mixed
methods to promote healthier grocery purchases and improve dietary intake and cardiovascular (CV) health.
Through strong mentorship and training opportunities, I will adapt and test the novel automated machine-
learning based Smart Cart 2.0 platform to deliver personalized recommendations and incentives for healthier
grocery purchases among Rhode Island adults with CV risk factors (i.e., Body Mass Index (BMI) > 30 kg/m2
and/or hypertension). To date, interdisciplinary research from my Big Data and Eating Decisions lab has
focused on investigating determinants of food choice and diet quality in large population cohorts and using
those insights to develop dietary interventions. I previously developed and pilot tested a semi-automated
personalized healthy food incentive platform using decision tree logic and found that it significantly improved
grocery purchase quality among healthy adults in the 9-month randomized controlled Smart Cart Study. The
proposed project extends this research by using novel applications of grocery sales data to adapt and evaluate
the automated `Smart Cart 2.0' platform to encourage healthier food purchases and dietary quality. To adapt
and scale this platform to adults at high CV risk, it is essential to understand its's feasibility, acceptability, and
preliminary effectiveness to facilitate adoption of a healthier diet among adults with CV risk factors. This project
also advances methodology to leverage technology and machine learning to automatically evaluate sales data
and deliver personalized dietary recommendations in real time. The proposed research will use the Obesity-
Related Behavioral Intervention Trials (ORBIT) framework to determine whether the Smart Cart 2.0 platform
adapted and tested with input from adults with CV risk factors promotes a clinically significant change in diet
quality and CV risk factors. In Aim 1, I will evaluate how well the Smart Cart 2.0 content and function meet the
needs of adults at CV risk using focus groups and surveys. In Aim 2, I will conduct a 1-week 3-arm mock
shopping trial in a simulated online grocery store to see how much the platform affects willingness to purchase
(WTP) recommended foods relative to control using a discrete choice task comparing a) personalized
recommendations b) personalized recommendations plus incentives to c) generic education and incentives
(control). In Aim 3, I will conduct a pilot 6-month randomized controlled trial to test the acceptability, feasibility,
and preliminary effectiveness of the Smart Cart 2.0 platform for improving real-world dietary behaviors, diet
quality, BMI, and blood pressure. Results of this project will provide critical pilot data to support a larger
randomized controlled trial. The training plan and collaborations established through this award will position me
to pioneer a policy-relevant translational research program focused on promoting healthier dietary patterns and
CV health.