Background

Obesogenic environments have been defined as ‘the sum of influences that the surroundings, opportunities or conditions of life have on promoting obesity in individuals or populations’⁽Reference Swinburn and Egger¹⁾. Ten years ago the Foresight Tackling obesities: future choices ⁽²⁾ report was published. This report described obesity in terms of complex systems⁽Reference Savona, Rutter and Cummins³⁾. While, 10 years ago, this concept was relatively new, it brought together evidence that linked the built and food environments in novel ways⁽Reference Lake and Townshend^4, Reference Lake, Townshend and Alvanides⁵⁾. Over 10 years the publications and evidence relating to obesogenic environments have grown exponentially. Most focus on how aspects of the built environment may contribute to current obesity levels, by influencing physical activity and dietary behaviours at the individual and community level⁽Reference Townshend and Lake⁶⁾. The 2014 McKinsey report⁽⁷⁾ described that overcoming obesity will require multiple solutions, involving many sectors from policy and practice through to industry and consumers. The report describes how we need to ‘reset the default’ in order to normalise and make healthy behaviours easier, relying less on the individual⁽⁷⁾. Swinburn et al. have described how ‘dramatic actions’ are needed, globally, to address food environments and thereby impact the on the rise in obesity and diabetes⁽Reference Swinburn, Kraak and Rutter⁸⁾. The worldwide rise in obesity has been ‘driven’ by significant changes in the global food system⁽Reference Swinburn, Sacks and Hall⁹⁾. This food system produces readily available, processed food which is marketed at populations⁽Reference Swinburn, Sacks and Hall⁹⁾. Looking at trends in high-, middle- and low-income countries, research has concluded that increases in the food energy supply, alongside increasing sedentary behaviour, explains the increases in population body weight, particularly in high-income countries⁽Reference Vandevijvere, Chow and Hall¹⁰⁾. The food environment has a role to play in helping to develop obesity, and therefore a role to play in preventing obesity. The present paper focuses on how the environment influences dietary behaviours mediated through the food environment.

Food environments

Food choice and eating behaviours are influenced by a multitude of factors from culture to personal preferences. The relationship between our food choices and the environment in which we make these choices is now widely acknowledged; from nudging and choice architecture⁽Reference Bucher, Collins and Rollo¹¹⁾ to conceptualising how the local food environment influences eating behaviour⁽Reference Caspi, Sorensen and Subramanian¹²⁾. We can define the food environment as any opportunity to obtain food; it includes physical, socio-cultural, economic and policy influences at both micro- and macro-levels⁽Reference Townshend and Lake¹³⁾. The broader food environment includes the home food environment, food policies and school food policies in addition to the neighbourhood food environment⁽Reference Lake, Townshend, Burgoine, Buttriss, Welch, Kearney and Lanham-New¹⁴⁾. Story et al. ⁽Reference Story, Kaphingst and Robinson-O'Brien¹⁵⁾ have developed an ecological framework to illustrate the influences on dietary behaviours. The present paper focuses on the neighbourhood food environment.

The neighbourhood food environment is defined as a mixture of retail outlets (e.g. small convenience stores to supermarkets) as well as restaurants and take-away (fast food) outlets and is not limited to the residential neighbourhood⁽Reference Lake, Townshend, Burgoine, Buttriss, Welch, Kearney and Lanham-New¹⁴⁾. The neighbourhood food environment influences individual food choice and food intake through the concept of food access. The relatively simple concept of access, in terms of the food environment actually includes five dimensions which are: availability, accessibility, affordability, acceptability and accommodation⁽Reference Penchansky and Thomas¹⁶⁾ (See for more details: Lake et al.Reference Lake, Townshend, Burgoine, Buttriss, Welch, Kearney and Lanham-New¹⁴).Research has focused on the availability and accessibility of neighbourhood food outlets. Two recent systematic reviews, one exploring the local food environment in relation to obesity⁽Reference Cobb, Appel and Franco¹⁷⁾ and one exploring the food environment in relation to diet⁽Reference Caspi, Sorensen and Subramanian¹²⁾ have been inconclusive in their findings. This is, in part, due to the complexity of the measures and the quality of the studies. However some important patterns emerged; for example, in adults, Cobb et al. ⁽Reference Cobb, Appel and Franco¹⁷⁾ found evidence that supermarket availability was negatively associated with obesity and fast food availability was positively associated. The Janssen et al. ⁽Reference Janssen, Davies and Richardson¹⁸⁾ review suggested that the strongest determinants of out-of-home food availability are the density of food outlets and deprivation within the built environment.

Our surrounding environment is going to impact on our food choice and ultimately on our eating behaviour and consequently our energy balance, weight gain and obesity⁽Reference Rosenheck¹⁹⁾. Neighbourhood food environments are important and much attention has been paid to fast food and takeaway outlets. We know the food served within these outlets tends to be nutrient poor and energy dense⁽Reference Jaworowska, Blackham and Davies^20, Reference Lachat, Nago and Verstraeten²¹⁾. Public Health England estimates that in 2014 there were over 50 000 fast food and takeaway outlets, fast food delivery services, and fish and chip shops in England⁽²²⁾ and a greater proportion of these are in deprived areas. Data from the UK National Diet and Nutrition Survey indicates that between a fifth and a quarter of people in the UK eat meals out once per week or more, with one-fifth eating takeaway meals at home once per week or more⁽Reference Adams, Goffe and Brown²³⁾.

Using data from Norfolk (England), researchers report that takeaway food outlet density increased from 1990 to 2008⁽Reference Maguire, Burgoine and Monsivais²⁴⁾. Takeaway food outlet density was significantly higher in more deprived areas at all time points. Worryingly, over the time period, there were increases in socioeconomic disparities in takeaway food outlet density⁽Reference Maguire, Burgoine and Monsivais²⁴⁾. This concentration effect has been observed in other studies where takeaway and fast food outlets tend to cluster in more deprived areas⁽Reference Macdonald, Cummins and Macintyre²⁵⁾.

A recent Danish cross-sectional study reported that fast food access is associated with fast food intake in the capital region of Denmark⁽Reference Bernsdorf, Lau and Andreasen²⁶⁾. Cross-sectional research in Norfolk, England reported that greater exposure to fast food and a lower educational level is associated with greater fast food consumption, BMI and odds of obesity⁽Reference Burgoine, Forouhi and Griffin²⁷⁾. However, this finding is not consistent with other studies. An Australian study of disadvantaged women explored the longitudinal associations between the fast food environment around their home and BMI⁽Reference Lamb, Thornton and Olstad²⁸⁾. Over the 5-year study, they found no association between increases in a major chain fast food outlet availability and increases in BMI over time. There are a number of reasons why fast food outlets around the home are not associated with a change in BMI. People lead complicated lives and do not simply source food from their residential neighbourhood. Focusing on residential address alone excludes other food environment exposure opportunities⁽Reference Burgoine and Monsivais²⁹⁾. In their study, Burgoine and Monsivais⁽Reference Burgoine and Monsivais²⁹⁾ explored the difference in food environments between homes, workplaces and along commuting routes between home and work for residents in the East of England. Perhaps unsurprisingly there was a significantly greater density of, and proximity to all food outlet types at work compared with home, particularly restaurants and takeaways. Most recently, Mason et al. ⁽Reference Mason, Pearce and Cummins³⁰⁾ used cross-sectional baseline data from the UK Biobank (project 17380) to explore associations between fast food and physical activity environments and adiposity for adults in mid-life. This large and unique UK study, spanning a geographically diverse area, found a weak association between access to fast food and adiposity. However, they reported high densities of physical activity facilities were associated with lower adiposity for these adults in mid-life. The authors attribute this to limitations in the metric and measurements used in the food environment analysis, an opinion strongly supported by an editorial in the same publication⁽Reference Monsivais and Burgoine³¹⁾.

These examples illustrate how reliable measures of the food environment are the foundation of research that will help to inform obesity-related policy⁽Reference McKinnon, Reedy and Morrissette³²⁾. A broad evidence base is required, ranging from spatial analyses to within store audits, alongside individual and neighbourhood level data⁽Reference Caspi, Sorensen and Subramanian¹²⁾. In a recent editorial, Cummins et al. ⁽Reference Cummins, Clary and Shareck³³⁾ explored the methodological challenges in estimating the effect of the food environment on obesity. While the use of census tracts or postcodes to define exposure to food environments has been used in research, they do not represent the environments an individual is actually exposed to. Rather, an activity-space approach is advised which considers the individual's behaviour in space and time⁽Reference Perchoux, Chaix and Cummins³⁴⁾.

As well as focusing on exposure to different types of outlets, interventions have also focused on the food served in outlets, promoting healthier ready-to-eat meals (to eat in, to take away or to be delivered). Hillier-Brown et al.⁽Reference Hillier-Brown, Summerbell and Moore³⁵⁾ conducted a systematic review to assess the impact of such interventions. The majority of the included studies were conducted in the USA (twenty-seven out of thirty), two in Australia and one in the UK. Most studies focused on adults and eighteen were within chain food outlets. The quality of evidence was generally considered to be poor; however, the present study provides useful insight into these types of interventions. Intrusive interventions that restricted or guided choice appeared to have an impact on food-outlet and customer level outcomes, while interventions that enabled choice or provided intervention had little impact.

However, these types of interventions are being delivered by local authorities and rarely get published within the academic literature⁽Reference Adams, Hillier-Brown and Moore³⁶⁾. In addition to the systematic review, this larger piece of work also explored what interventions were being delivered by local authorities (in England) around providing healthier ready-to-eat meals (to eat in, to take away, or to be delivered)⁽Reference Hillier-Brown, Summerbell and Moore³⁷⁾. This systematic mapping and evidence synthesis of interventions to promote healthier ready-to-eat-food⁽Reference Hillier-Brown, Summerbell and Moore³⁷⁾ found seventy-five interventions, of which most were ‘award’ type interventions (defined as those that involved an assessment of food outlet practice(s) targeted by the intervention using pre-defined criteria, together with some sort of accreditation if the food outlet met the criteria). The interventions were delivered by local authority staff, in most cases Environmental Health Officers. Interventions tended to be time-limited and evaluation was limited⁽Reference Hillier-Brown, Summerbell and Moore³⁷⁾. This research suggested that there was much activity across local authorities in England (as well as elsewhere) and that sharing good practice as well as robust evaluation would be beneficial. The analyses suggested that business owners were generally positive about such interventions. Those that were cost neutral and that were not obvious to their customers were perceived more positively. This programme of research has also explored the perspectives of intervention deliverers⁽Reference Goffe, Penn and Adams³⁸⁾. Interviews found barriers and facilitators to interventions including lack of funding for interventions and the difficulties of dealing with this sector of the food industry⁽Reference Goffe, Penn and Adams³⁸⁾.

In the developed world, a sophisticated food supply chain and food system, as well as the domination of a small number of companies, have been noted in the literature⁽Reference Lang³⁹⁾. While there has been much focus on fast food, there is a lack of clarity about the role of supermarkets in terms of obesity⁽Reference Burgoine, Mackenbach and Lakerveld⁴⁰⁾. Using a large UK sample, Burgoine et al. ⁽Reference Burgoine, Mackenbach and Lakerveld⁴⁰⁾ explored the independent and combined associations of supermarket distance and education with BMI, overweight and obesity. Their results indicated that greater supermarket distance was independently associated with higher BMI and odds of both overweight and obesity⁽Reference Burgoine, Mackenbach and Lakerveld⁴⁰⁾.

Food outlets selling low-cost energy-dense food is one aspect of the modern UK high street. Townshend has coined the term ‘toxic high street’ to describe the co-location of money lenders, betting shops and fast food restaurants in more deprived neighbourhoods⁽Reference Townshend⁴¹⁾. Since the economic recession of 2007/2008, Townshend has described how this scenario has become embedded as a feature of the British high street, in contrast to more affluent areas where bistros, delis and boutiques flourish. While some areas have empty boarded up shops, other areas have an unhealthy or ‘toxic’ mix of uses including; takeaways and ‘all you can eat’ buffet style restaurants; sub-prime money lenders (offering instant cash and pay-day loans at high-interest rates); and betting shops; also tanning salons, body piercing parlours, shops selling cut-price (sometimes counterfeit) alcohol and tobacco⁽Reference Townshend⁴¹⁾.

Food environments and eating behaviours

Few studies have explored individual's diets in relation to their food environment, mainly due to the fact it is time-consuming and methodologically challenging. Focusing on young adults (n 86, mean age 17 years), a study by Tyrrell et al. ⁽Reference Tyrrell, Greenhalgh and Hodgson⁴²⁾ provided detailed information regarding where young people obtain food and the nutritional consequences of choosing those food environments. Their respondents completed 4-day self-complete food diaries, recorded what food they consumed and where food was sourced. Sources of food items were coded as home (including friends or relatives homes), and out-of-home. Food items sourced out-of-home were further classified using an updated and a modified version of the Lake et al. food environment classification tool⁽Reference Lake, Burgoine and Greenhalgh⁴³⁾. The tool contained fifteen out-of-home food outlet categories with eighty-eight detailed sub-categories. With the addition of ‘home’, use of sixteen possible food environments was recorded. Food was obtained by these young adults from a wide range of environments. Over 4 days of observation, all respondents sourced food from home and from an average of 4·3 different out-of-home food environments. Excluding home and school, takeaway and fast food environments were the most commonly used with 53% of respondents sourcing food from these environments; 41% obtained food from convenience stores, at least once over the 4-day period. Food sourced from specialist outlets, convenience outlets, and retail bakers (i.e. national commercial bakers) were the most energy dense. The highest percent energy from fat was from foods sourced from retail bakers, takeaway and fast food and specialist outlets (47%, 43% and 42%, respectively)⁽Reference Tyrrell, Greenhalgh and Hodgson⁴²⁾. This work highlights the eating behaviours of this transitional age-group and points to the importance of interventions around schools, colleges (the school fringe) and the wider food environment. The importance of the school fringe has been highlighted in a number of studies⁽Reference Gallo, Barrett and Lake^44–Reference Walton, Pearce and Day⁴⁶⁾.

In a larger cross-sectional study of 839 mothers with young children, Vogel et al. ⁽Reference Vogel, Lewis and Ntani⁴⁷⁾ used a survey including an FFQ, demographic characteristics and frequently visited locations. The authors developed scores for the mother's diet and their food environment. Using Geographical Information Systems alongside demographic information their findings suggested that there is a relationship between health and place as well as educational attainment for these mothers. For example, less healthy food environment scores were associated with better dietary quality in mothers with degrees, but poorer dietary quality in mothers with lower educational attainment. Their findings also suggested that the majority of respondents were exposed to less healthy food environments. These studies⁽Reference Tyrrell, Greenhalgh and Hodgson^42, Reference Vogel, Lewis and Ntani⁴⁷⁾, both cross-sectional and both with their limitations, link the environment to food behaviours of two population groups and highlights the importance of the environment in relation to eating behaviour.

Urban planning and the food environment

The environment has been acknowledged as a determinant of health⁽Reference Barton and Grant⁴⁸⁾. Historically, the urban planning and public health professions originated from the same need to deal with the health inequalities due to both the rapid industrialisation and urbanisation of the 19th century⁽Reference Northridge and Freeman⁴⁹⁾. In the 21st century, it is internationally acknowledged that professions need to work together to address our ‘dynamic, complex and interconnected health concerns’⁽Reference Northridge and Freeman⁴⁹⁾. Examples of this collaboration include The Healthy Cities movement which, amongst other things links planning and health has raised awareness around healthy urban planning⁽Reference Barton and Grant⁴⁸⁾.

An acknowledgment of the potential role of the built environment and planning in creating healthier communities was reflected in the in the 2012 National Planning Policy Framework for England⁽⁵⁰⁾. In England, in 2013, many public health responsibilities were moved from the National Health Service to local authorities⁽Reference Atkins, Kelly and Littleford⁵¹⁾. This included responsibility for obesity, community nutrition and increasing physical activity⁽Reference Lake, Henderson and Townshend⁵²⁾. A recent umbrella literature review⁽⁵³⁾ assessed the impact of the built and natural environment on health. The review concentrated on five key built environment topics: neighbourhood design, housing, healthier food, natural and sustainable environment, and transport. These are environmental issues that can be shaped by planners and have the potential to influence health.

Policy documents have highlighted the role that local authorities have in tackling obesity^(54–56). There has been a recent interest in the role of local authorities in shaping the food environment⁽²²⁾, particularly in engaging with small businesses⁽⁵⁷⁾ and with planning departments⁽⁵⁸⁾.

Planning policy in relation to the food environment has tended to focus on the restriction of hot-food takeaways. Using The Town and Country Planning (Use Class) Order 1987, outlets are classified according to the use class order of the premises they occupy, dependent upon their primary operating model and premise size (The Town and Country Planning (Use Classes) Order 1987 The Secretary of State for the Environment (1987) http://www.legislation.gov.uk/uksi/1987/764/made (as amended) puts uses of land and buildings into various categories known as Use Classes.). However, differences between categories may be unclear. For example, restaurants and cafes (classified as A3), may also sell food to takeaway and hot food takeaways may have small seating areas. Classifications of interest are in Table 1. An increasing number of local authorities are using supplementary planning documents (SPD) to control fast food outlet proliferation (SPD provide detail to support policy in higher level development plan documents. SPD are a material consideration in the assessment and determination of any planning application. (reference from http://www.newcastle.gov.uk/planning-and-buildings/planning-policy/supplementary-planning-documents))⁽Reference Lake, Townshend, Burgoine, Buttriss, Welch, Kearney and Lanham-New¹⁴⁾. It is one of the mechanisms suggested by Public Health England for local government to influence the out-of-home food environment⁽²²⁾, alongside the use of the local plans, joint strategic needs assessments, joint health and wellbeing strategies sustainability and transformation plans and the use of Health in all Policies.

Table 1. The Town and Country Planning order description of land and buildings in categories known as Use Classes

However, not all takeaway outlets are necessarily unhealthy. Moreover, where seating is provided food outlets are classified as restaurants and therefore are exempt from takeaway restrictions. This system of classification, in relation to food outlets, requires an overhaul, but the planning system is probably never going to be nuanced enough to differentiate between healthy and unhealthy outlets.

Barking and Dagenham was one of the first planning authorities to produce an SPD aimed at restricting permission for hot-food takeaways (A5 use)⁽⁵⁹⁾. It has been reported that in 7 years, this London borough has reduced fast food outlets by 15% from 187 to 160⁽Reference Butler⁶⁰⁾. The mayor of London has recently announced a double-pronged approach to limit new takeaways opening within 400 m of schools and health standard to boost baked or grilled food rather than fried⁽⁶¹⁾.

A review by Dr Foster Intelligence in 2011⁽⁶²⁾ found four main approaches taken by local planning authorities to curb fast food proliferation:

1. 1. Only allowing takeaway outlets in specified areas. By defining only certain locations, for example existing shopping areas, where further takeaway outlets are deemed acceptable; however, this may introduce issues of concentration and clustering.

2. 2. Restricting concentration and clustering. If a location is suitable, local authorities can seek to restrict the number of takeaway outlets in a row (e.g. to two, or three) or the percentage of frontage (ground floor use facing the street) given over to takeaways (e.g. 5% has been used).

3. 3. Restricting proximity to other uses. This means setting out buffer zones (e.g. 400 m) around land uses such as schools, parks and children's centres, where the development of takeaway outlets is forbidden.

4. 4. Clamping down on ‘back door’ applications. Fast food outlets have their own classification in UK planning terms, referred to as A5, hot food takeaways. Developers can sometimes try to circumvent takeaway restrictions by opening outlets under different classifications, primarily A3, restaurants and cafes, where the intention is that food will be primarily consumed on the premises. For example, this might be done by adding in a nominal seating area on plans submitted to the local planning authority.

Some planning authorities have sought to charge a levy, or fee, where planning permission is granted for a new takeaway, with funds raised going to initiatives to tackle childhood obesity, for example, improving green spaces to encourage physical activity⁽Reference Lake, Townshend, Burgoine, Buttriss, Welch, Kearney and Lanham-New¹⁴⁾.

Since this review (2011), the number of local authorities producing SPD to tackle fast food outlet proliferation has increased and guidelines have been developed to meet local situations. Estimates are that there are at least twenty SPD in place to tackle hot food takeaways⁽Reference Butler⁶⁰⁾. Gateshead Council's SPD⁽⁶³⁾, prevents new permissions for A5 use in wards where more than 10% of year six pupils are obese. This has effectively barred any new hot food takeaways in the borough.

While authorities may object to food outlets being added to their foodscape, civic society also has a perspective. Also in England, Spence et al.⁽Reference Spence, Blundred and Chaffer⁶⁴⁾ analysed local residents’ opposition to a multinational fast food company submitted to the planning enquiry. While there were many health reasons to oppose this particular development, including its close proximity to a school (within 400 m) the main concerns by local residents included traffic, the effect on the environment (littering, etc.) there were also concerns about noise and safety. This research highlights the importance of cross-sectorial working in local governments, and to consider engaging with residents about health impacts of proposed new food outlets⁽Reference Spence, Blundred and Chaffer⁶⁴⁾.

Our recent qualitative work⁽Reference Lake, Henderson and Townshend⁵²⁾ sought to understand the views of individuals working in public health and those working in spatial planning within local government on their respective responsibilities for addressing obesity through spatial planning. One-to-one interviews aimed to explore respondents’ perceptions concerning the wider issue of their role in public health before asking them about tackling issues of obesity, community nutrition and increasing levels of physical activity. Eight interviews were conducted with three Directors of Public Health, one Deputy Director and four planners with a range of seniority from across five local authorities within the North East of England. The findings, from this relatively small study, illustrate what tackling a global problem is like on the front line of local government. It identified a range of barriers to engaging with planners, including an insufficient understanding of the causes of obesity and the primacy of addressing obesity via multiagency approaches, fragmentation in the health system and conflicting priorities. Our research indicates that planners could be better engaged in the obesity agenda via formal incentives (e.g. written within planners’ job descriptions or regulations), and aligning priorities via ‘soft approaches’ (e.g. public health leadership roles)⁽Reference Lake, Henderson and Townshend⁵²⁾. Since this research, the Local Government Association and Public Health England have published a briefing document about a Whole Systems approach to obesity; this practical guide will be published in 2019⁽⁶⁵⁾.