The research map of the food environment

In 2017, the UN High-Level Panel of Experts on Food Security and Nutrition identified three key research gaps in the study of the food environment (Fig. 1)⁽⁴⁾. The first gap concerns the effect of the food environment on human health (the 1^st mechanism). The second gap explores how external factors, such as transportation, land use and sociocultural environments, influence the food environment (the 2^nd mechanism). The third gap focuses on developing a comprehensive theoretical framework to measure and understand the dynamics of the food environment for future intervention.

Figure 1. Food environment research map according to the gaps summarised by the High-Level Panel of Experts report (2017).

Previous decades of research have focused on the health effects (the 1^st mechanism), but inconsistent findings have hindered the development of obesity mitigation strategies^{(Reference Cobb, Appel and Franco10–Reference Westbury, Ghosh and Jones15)}. However, understanding the 1^st mechanism does not represent the entirety of research on food environments. The food environment is linked to various urban subsystems and other systems and is constrained by them. For example, transportation issues can impede food distribution and urban planning can indirectly shape food environments. To generate meaningful findings for human health policies, it is essential to address these gaps comprehensively in a broader context. This involves clarifying its external links with multiple urban subsystems (the 2^nd mechanism), internal differences and connections with other food system components, and defining the food environment more precisely (the intervention goal). The most critical research gap lies in the development of a theoretical framework that accurately defines and measures the food environment, which is crucial for uncovering the two mechanisms. Therefore, our priority lies in thoroughly examining and dissecting the challenges in defining and measuring the food environment to develop a conceptual framework tailored to the Chinese context.

Deconstructing the traditional definition of food environment: identifying and highlighting key elements

Defining the food environment and its key components poses a great challenge^{(Reference Turner, Aggarwal and Walls3)}. Existing definitions, which integrate physical, economic, political and sociocultural contexts and various influencing factors, have not effectively established a link between the food environment and population health outcomes. This is because these contexts not only influence the food environment but also directly or indirectly affect population health through other pathways. The complexity involved makes it exceptionally challenging to explain the relationship between food environment and population health. Therefore, the definition of food environment should identify its distinctive characteristics and specify the specific pathways through which it affects population health.

Tracing back to the inception of food environment research may shed light on this question. The primary focus of these studies was initially to address the rising obesity epidemic, as the overabundance of food provided by the food environment is the main driver of this global health concern^{(Reference Swinburn, Egger and Raza16)}. Additionally, the dietary risk factors identified in studies on the global burden of disease are all based on evidence of the nutritional composition of food^{(Reference Murray, Aravkin and Zheng17)}. Thus, food and nutrition, but not other attributes such as food quality and safety, are the core components of the food environment^{(Reference Cong, Zhao and Kwan18)}. In contemporary society, the accessibility of food is not ubiquitous but rather concentrated in specific locations. Spatial proximity between food and consumers plays a pivotal role in determining the dynamics of the food environment and its interaction with individuals. Furthermore, the spatial distribution of food plays a fundamental role in establishing evidence regarding factors that influence the food environment. Therefore, the spatial location of food is the second key component of the food environment.

When considering the quantity or other characteristics of food as attributes of the food environment, it is essential to recognise that the physical, economic, political and sociocultural contexts outlined in the High-Level Panel of Experts definition should be regarded as impact factors that contribute to changes in the food environment, rather than constituting the food environment itself. To this extent, we identified the fundamental components that constitute the food environment as well as the external systemic factors that influence it.

A noticeable issue arising from the vague definition of traditional food environments is the proliferation of various terms derived from it. Different studies refer to the food environment using terms such as market food environment^{(Reference Toure, Herforth and Pelto19)}, retail food environment^{(Reference Wilkins, Radley and Morris11,Reference Wilkins, Morris and Radley20)} , community/neighbourhood food environment^{(Reference Zenk, Tarlov and Wing21)}, public food environment^{(Reference Cong, Zhao and Kwan18,Reference Cong, Zhao and Gong22)} , local food environment^{(Reference Cobb, Appel and Franco10,Reference Cong, Zhao and Kwan18)} and workplace food environment^{(Reference Burgoine and Monsivais23)}. However, the distinctions and connections between these terms are overlooked.

Measurement issues

Measuring the food environment in multiple dimensions may address several challenges.

First, primary measurements in community-scale studies often target specific outlets, such as supermarkets^{(Reference Jalbert-Arsenault, Robitaille and Paquette24,Reference Harmer, Jebb and Ntani25)} and restaurants^{(Reference Joseph, Cassandra and Wesley26–Reference Chen, Zhao and Yang28)}, or specific settings, such as residential areas or schools^{(Reference O’Halloran, Eksteen and Gebremariam29)}. In some cases, the availability of certain types of food has been assessed^{(Reference Williams, Wang and Taniguchi30,Reference Wray, Martin and Doherty31)} . On a broader scale, food environments have often been measured by counting the number of different food outlets in many studies^{(Reference Burgoine and Monsivais23,Reference Shearer, Rainham and Blanchard32–Reference Hall, Huang and Yi34)} . However, the reliability of these methods has been questioned^{(Reference Cong, Zhao and Kwan18,Reference Wilkins, Radley and Morris35–Reference Thomson, Goodman and Landry37)} . Given that dietary risk factors identified by the global burden of disease research are universally applicable^{(Reference Afshin, Sur and Fay38)}, it may be more effective to develop scale-independent measurement methods based on these factors as demonstrated by a recent study^{(Reference Cong, Zhao and Kwan18)}.

Second, the incorporation of the digital dimension into the food environment framework has been under-discussed. It remains unclear whether the digital food environment (DFE) requires a separate measurement method or whether it can be integrated into the overall framework. Although a recent study touched on digital settings and culture, the theoretical framework for comprehending the DFE and its integration into the overall framework is in its infancy^{(Reference Granheim, Løvhaug and Terragni39)}. Simply adding the digital dimension to the existing measurement dimensions of the food environment will only complicate measurements, without fundamentally solving the problem.

Third, the lack of transparency regarding the development of measurements and extraction of information on the different dimensions of food environments from their data should be addressed^{(Reference Wilkins, Radley and Morris35)}. One review found that only half of the studies followed rigorous systematic steps, hindering the comparability and replicability of research findings^{(Reference Nahid, Mohammad and Leila40)}. A protocol to standardise food environment measurements within the overall framework is needed to ensure an effective comparison of results across contexts and foster a more comprehensive understanding of the impacts of food environments on health and sustainability.

A framework for food environment research

In light of the above discussions, we characterise the food environment as a specialised environment, encompassing food items and their nutritional value, as well as the physical locations where they are available^{(Reference Cong, Zhao and Kwan18)}. As the core of the food system, it bridges the extensive Earth system and persistently interacts with the relevant population via technological, economic, social and cultural factors, ultimately influencing human health. Accordingly, we propose a conceptual framework as depicted in Fig. 2.

Figure 2. A conceptual framework for diet, food, food environment, food system and broader systems in the digital age.

The traditional food supply chain involves several steps, from harvesting food to making it available to consumers at retail and service stages, including production, storage, distribution, transportation, processing, packaging and retailing⁽⁴⁾. However, the digital age has brought significant changes, providing consumers with unprecedented access to food products. Digital platforms for food sales and services, such as online delivery, e-commerce, live streaming and modernised delivery systems, have broadened consumer access to diverse food choices. These changes have broken down the traditional barriers and allowed people to connect directly with producers, processors, retailers and distributors. The resulting public food environment encompasses various foods and their suppliers throughout the food supply chain^{(Reference Cong, Zhao and Gong22)}, including all market and retail food environments.

Individual food items are aggregated into dishes or meals during food services or cooking at home, ultimately forming part of the consumer’s diet. The goal of both the retail and food service sectors is to deliver food to the consumer’s location for consumption. This food, delivered to communities, represents the aggregated diet of people in those places and constructs the community food environment. The community food environment, as defined here, includes the neighbourhood food environment, local food environment, workplace food environment and other community areas where people receive food. The community food environment plays a crucial role in exploring the association between individual dietary choices and the public food environment. A thorough understanding of the public food environment is essential for formulating effective macro-intervention policies, whereas a comprehensive understanding of the community food environment is valuable for implementing practical intervention practices.

The food environment is influenced by various factors, including natural disasters, urbanisation, political and economic shocks, sociocultural norms and technological advancements, all of which also impact human diet and health. The complexity of these factors requires a multidisciplinary approach involving fields such as geography, food and nutrition science, economics, public health, urban planning and computer and data science. By combining these diverse disciplines, we can gain a comprehensive understanding of the factors influencing the food environment and develop effective intervention strategies and urban planning approaches to address obesity and promote healthier food environments.

Digital technologies have infiltrated nearly every facet of human existence, including the food environment and its contributing factors. These technologies serve as crucial tools for monitoring and analysing the food environment and its impact factors. However, it is important to note that the foundational definition of the food environment currently in use does not differentiate between the physical food environment^{(Reference Downs, Ahmed and Fanzo41)} and the DFE^{(Reference Granheim, Løvhaug and Terragni39)}. This is because food purchases frequently originate from physical food outlets, even though digital technologies have significantly transformed the way individuals engage in food environments. Essentially, integrating digital aspects into the food environment for DFE does not change the core relationship between the food environment and human diet/health because the digital dimension of the food environment cannot exist independently of the physical food environment. In our framework, the digital aspects are identified as belonging to the technological factor domain. The current adoption of terms such as DFE not only directs researchers’ attention to the digital aspects of the food environment but also emphasises the crucial role of digital technology in facilitating comprehensive monitoring of the food environment. This has the potential to catalyse a ‘data revolution’ in food environment research, ultimately leading to a paradigm shift in the field⁽⁴⁾.

In addition, although the transportation system serves as the key means of delivering food from providers to consumers, it does not influence the core relationship between the food environment and human diet/health, which is similar to digital technology. Thus, the transportation system was not highlighted in our framework.

Our framework provides several benefits. At the micro level, integrating nutrition into food environment definitions enhances measurement approaches informed by global disease burden insights. This can lead to effective and interpretable measurements that may be incorporated into the Food Systems Countdown Initiative^{(Reference Schneider, Fanzo and Haddad42)}. At the macro level, the framework enables food environment research to extend beyond merely addressing the global obesity epidemic from a nutritional or epidemiological perspective and serves as a critical tool in addressing the global syndemic (including obesity, undernutrition and climate change)^{(Reference Swinburn, Kraak and Allender43)} and other environmental issues. Furthermore, it broadens the scope of Earth system modelling to include human social systems^{(Reference Steffen, Richardson and Rockström44)}, given that the food environment is a central link among various societal systems (as shown in Fig. 1), impacting both human and planetary health.