The long-standing concern over “food deserts”—areas where residents struggle to access nutritious food options—has gained new analytical depth from an MIT-led study. Examining three major global cities with a novel, highly granular methodology, researchers concluded that a scarcity of nearby healthy eating choices significantly correlates with elevated rates of obesity and other adverse health outcomes.
Researchers moved beyond traditional geographic mapping, instead undertaking a comprehensive analysis of the nutritional value of millions of food items found on approximately 30,000 restaurant menus. This detailed examination provided a significantly more precise assessment of the link between neighborhoods and nutritional outcomes.
A recent study, co-authored by MIT researcher Fabio Duarte, establishes a direct correlation between the types of food sold in restaurants and public health. Duarte underscores the critical influence of the “food landscape” on people’s well-being.
Researchers unveiled an open-access study this week in Nature: Scientific Reports, titled “Data-driven nutritional assessment of urban food landscapes: insights from Boston, London, Dubai.”
The collaborative research featured contributions from a team spanning Wageningen University in the Netherlands and MIT’s Senseable City Lab. From Wageningen University, the co-authors included PhD student Michael Tufano and senior researcher Guido Camps. The Senseable City Lab, which employs data-driven methods to analyze cities as dynamic systems, was represented by its associate director Duarte; postdoc Martina Mazzarello; research fellow Javad Eshtiyagh; and Carlo Ratti, who serves as both professor of the practice and director of the lab.
Guests perused the array of food and beverage options.
During the summer of 2023, researchers conducted a detailed analysis of menus from popular food-delivery platforms in Boston, Dubai, and London, compiling an extensive database containing millions of food items. These items were then evaluated against the USDA’s FoodData Central database, a comprehensive resource cataloging 375,000 distinct food products. The study’s analytical framework relied on two primary metrics: the Meal Balance Index and the Nutrient-Rich Foods Index.
A recent study analyzed a vast collection of menu items across three major global cities, revealing disparities in USDA database inclusion. Researchers scrutinizing approximately 222,000 menu entries from over 2,000 Boston-area restaurants found 71 percent were represented in the USDA database. In Dubai, an examination of roughly 1.6 million items from nearly 9,000 eateries showed a significantly lower figure, with only 42 percent matching the database. London’s larger dataset, encompassing around 3.1 million menu items from some 18,000 establishments, registered 56 percent in the USDA database.
Researchers evaluated the nutritional content of menu items and correlated this data with health outcomes in Boston and London. The analysis revealed a clear link in London between neighborhood food offerings and obesity levels. A similar, though slightly less definitive, correlation was found in Boston. Neighborhoods that offered food options rich in dietary fibers, often including fruits and vegetables, consistently demonstrated improved health data.
In Dubai, researchers, while lacking specific health data, identified a strong correlation between a neighborhood’s rental prices and the nutritional value of its available food. This finding suggests that the city’s wealthier residents likely benefit from superior dietary options.
Less nutritious food items are directly linked to a rise in obesity cases, according to Tufano. He points out that this problem is exacerbated in lower-income neighborhoods, which not only have a higher concentration of fast food establishments but also offer options with significantly reduced nutritional value.
Redefining the Food Map
This new study significantly deepens the analysis of “food deserts,” building upon prior research. While previous groundbreaking work successfully identified neighborhoods and regions with insufficient access to quality food, the current investigation offers a more comprehensive assessment of what residents actually consume. The research moves towards evaluating the complex array of food options available in any given area, a complexity that can persist even in locales with seemingly limited choices.
Duarte challenges the conventional definition of a “food desert,” expressing dissatisfaction with the notion that an area is categorized as such solely if it offers only fast food, while the presence of a store like Whole Foods exempts it from that classification. Duarte contends that this distinction is “not necessarily like that,” implying a more complex reality.
This latest research provides Senseable City Lab researchers with an innovative technique for understanding urban dynamics and the impact of city environments on public health. The lab’s previous work has frequently focused on areas such as urban mobility, and extended to critical issues like the connection between mobility and air pollution.
The ability to conduct granular studies of food and health within individual neighborhoods underscores a key advantage of data-rich environments: the capacity for exceptionally detailed analysis across various aspects of life.
Initially, efforts to analyze urban data were significantly limited by low resolution, Ratti explained. However, the current explosion in data volume now presents an unprecedented opportunity to meticulously examine the profound influence of city environments on public health. This emerging field represents a crucial new frontier for their lab, enabling remarkably precise insights into urban health dynamics.
