GroceryDB: Prevalence of Processed Food in Grocery Stores

Abstract

The offering of grocery stores is a strong driver of consumer decisions, shaping their diet and long-term health. While processed food has been increasingly associated with unhealthy diet, information on the degree of processing characterising an item in a store is virtually impossible to obtain, limiting the ability of individuals to make informed choices. Here we introduce GroceryDB, a database with over 50,000 food items sold by Walmart, Target, and Wholefoods, unveiling the degree of processing characterizing each food. GroceryDB indicates that 73% of the US food supply is ultra-processed, and on average ultra-processed foods are 52% cheaper than minimally-processed alternatives. We find that the nutritional choices of the consumers, translated as the degree of food processing, strongly depend on the food categories and grocery stores. We show that there is no single nutrient or ingredient “bio-marker” for ultra-processed food, allowing us to quantify the individual contribution of over 1,000 ingredients to ultra-processing. GroceryDB and the associated http://TrueFood.Tech/ website make this information available, aiming to simultaneously empower the consumers to make healthy food choices, and aid policy makers to reform the food supply.

Ultra-processed food staples dominate mainstream U.S. supermarkets. Americans more than Europeans forced to choose between health and cost

“…Moreover, the prevalence of UPF markers also varies according to the type of supermarket in the United States (35), as confirmed by our results showing 69% more UPF markers in products found in budget-friendly supermarkets (Walmart and Target) compared to products found in premium supermarkets (Whole Foods). The number of UPF markers might be linked to the industry's desire to save production costs (18) and produce ultra-palatable products, replacing unprocessed products in the shopping trolley.…”

You are where you eat: Effect of mobile food environments on fast food visits

“…In addition to identifying locations for intervention, our methodology could also be used to inform individual-level interventions promoting or encourage visits to food outlets located in food environments that have more diverse, healthy food options, for example using a mobile app. Design of such individual-level interventions would require combination with other data about food intake (e.g., delivery, nutrients 50 , degree of food processing 51 ), food preference and sentiment 52,53 , and price sensitivity 46,54,55 .…”

Machine Learning Prediction of Food Processing

“…Given that our algorithm only needs the Nutrition Facts, information already accessible to consumers on packaging and via smartphone apps, web portals, and grocery store and restaurant websites, F P ro can help monitor the reliance of an individual's diet on less or more processed food. Building on the portability of our model, we were able to extend the F P ro assessment to over 50,000 products collected from major US grocery store websites, a first step towards a systematic characterization of the food environments [82].…”

“…Requirements for Branded Products . While in this study we focus on survey data, our model based on nutritional values can easily work on different food and cohort databases, as proven by our analysis of over 50,000 products collected from major grocery store websites [62]. When facing real-world food data, we have to account for the regulations introduced by government agencies, like the FDA.…”