Millions of grocery transactions reveal how London’s access to healthy food is influenced by income, culture, and transportation.
Study: Food purchase data reveals the locations of London’s ‘food deserts’. Image credit: bobby20/Shutterstock.com
Food deserts are areas where nutritious food is unaffordable or inaccessible in necessary quantities. A recent study by researchers from the University of Adelaide and the University of Nottingham, published in the open-access journal PLOS Complex Systems, traces the occurrence of these areas across London using grocery shopping data.
Mapping London’s nutrition crisis
An unhealthy diet plays a major role in about 13% of deaths in London, while over 60% of its residents are overweight or obese. People living in urban food deserts are more likely to eat unhealthy food, which has been associated with higher rates of obesity, diabetes, and cardiovascular disease. The present study aimed to identify food deserts across London.
The conventional approach to identifying food deserts consists of mapping store locations, on the premise that where stores offer healthy foods, it ensures nutritional access to the neighborhood. The current study used another strategy based on tracking food purchases by a set of consumers.
This helped overcome the uncertainty surrounding the actual purchase of food, in addition to revealing other factors that potentially influenced food choices. These include lack of transportation and financial incapacity, as well as the presence or absence of outlets offering healthy food in the neighborhood.
Tracking grocery data
The dataset comprised 420 million food item purchases in 2015, involving over 1.6 million buyers who used Tesco Clubcards in 411 Tesco supermarkets in London. Using card data, researchers traced residential patterns while preserving anonymity.
Food purchases were classified into 12 categories, including grains, fruits and vegetables, fish, poultry, red meat, soft drinks, and so on. The overall nutritional quality of the purchases was compared with social and demographic factors, including car ownership, income, and ethnic context.
Food purchases were analyzed at a fine geographical scale, known as Lower Layer Super Output Areas (LSOAs), which represent small neighborhoods of approximately 1,000 to 3,000 residents.
Drivers of poor diet
Food deserts were found in large groups across London. Mostly, they were located in East London boroughs, including Newham, Redbridge, and Barking, with several also in parts of West London, such as Ealing.
Food deserts were characterized by purchases that favored ultra-processed, high-sugar foods, with high total carbohydrate content. Purchases of fresh foods, including fiber-rich plant foods and protein-rich foods, were relatively low.
In contrast, inner northwest London boroughs had nutrient-rich purchases. This underlines the disparity in nutritional quality across London.
The next step was to analyze the food-buying patterns by sociodemographic factors. This revealed that people expected to have better food access, as indicated by greater purchasing power, such as higher income and car ownership, were confined to specific regions, for the most part. Yet their food choices sometimes ran counter to expectations. One example came from East and West London, where higher incomes were linked to nutritionally poor purchases in contrast to inner West London.
Owning a car might be expected not only to indicate a higher net worth and purchasing power, but also to enable better access to nutritious food outside one’s own area. However, car owners tended to buy healthier food only in a part of one northwestern district of London. Conversely, car owners in east, west, and some parts of northwestern London were more likely to buy nutrient-poor food.
A third set of factors driving nutritional patterns was linked to ethnic background. Black, Asian, and minority ethnic (BAME) communities in west, northeast, and inner east London were more likely to purchase low-quality food. This warrants further investigation to identify factors like poverty (twice as prevalent among BAME vs White communities), local cultures, and lack of access as potential barriers to buying healthy food, creating a continuing cycle
Obviously, communities and neighborhoods that rely on unhealthy food pose a significant public health threat. This emphasizes the need to frame locally relevant food interventions and policies to help people living in food deserts gain access to healthy foods.
Rethinking access to nutrition
This study aimed to unravel the complexity of meeting urban food access challenges by combining big commercial data with sociodemographic and geographic markers. Rather than creating a theoretical model, the researchers used actual purchase data to pinpoint nutritionally deficient dietary patterns by locality and community. Thus, they shifted from dependence on the presence of physical infrastructure, in this case, fresh food outlets, as the source of their data. This is important in framing effective and evidence-based public health guidelines.
The study breaks new ground in utilizing large-scale consumer data to effectively explore public health and social issues, while upholding privacy and maintaining high ethical standards. Food retail chains provide a valuable source of data related to health for urban health planners to develop more universally accessible healthy food systems.
It is clear that to transform these food deserts into hubs of healthy eating, interventions must start with the infrastructure. Additionally, they must embrace economic, cultural, logistical, and educational insights into the groups at risk.
Our findings demonstrate the importance of purchase data analysis to accurately identify food deserts and their multifaceted drivers. This allows for local public health strategies that are context-sensitive and better attuned to the lived realities of urban residents.
Lead author, Tayla Broadbridge
Next steps for food equity
Rather than a simple one-on-one association of food outlets with food deserts, this study combined geographical and spatial factors with sociodemographic markers to provide rich insights into food-buying patterns across London. The study highlights the significant need for more accurate and locally relevant interventions that address the multifaceted obstacles to equitable access to healthy food in this city, thereby improving public health.
However, the authors noted that the dataset reflects only Tesco supermarket purchases from 2015, excluding other retailers and food sources, and therefore represents associations at the neighborhood level rather than individual behavior.
Download your PDF copy now!
Social media food shows like Mukbang are changing how people eat