Food marketing works and is an important target for public health action

In high-income countries, we obtain the vast majority of our food through commercial interactions. This marketplace leads to marketing activities where producers and retailers create, communicate, deliver and exchange food products with consumers⁽¹⁾. Sometimes referred to in terms of the four Ps of product, promotion, placement and price, food marketing captures not just what foods are advertised (promotion) but also which outlets they are stocked in (placement), how much they cost (price) and even what is in them (product).

Much research^{(Reference Boyland, Nolan and Kelly2–Reference Coleman, Hanson and van Rens4)}, including in the current special collection^{(Reference Boyland, Muc and Coates5)}, has demonstrated that food marketing ‘works’ to influence food choices, consumption and purchasing and that it is concentrated on less healthy foods high in fat, salt or sugar (HFSS). In the UK, just 1 % of food advertising spend goes on fruit and vegetables^{(Reference Goudie6)}. At the same time, healthier food is consistently more expensive per unit of energy than less healthy food^{(Reference Goudie6)}.

There have been substantial efforts to restrict the harmful effects of food marketing both through voluntary approaches and government legislation^{(Reference Boyland, McGale and Maden7)}. The UK has been at the forefront of some of these efforts: it was one of the first countries to introduce mandatory restrictions on TV food advertising to children in 2007 and led a successful programme of voluntary targets to reduce salt in packaged food^{(Reference Alonso, Tan and Wang8)}. More recent actions include voluntary sugar and energy reduction programmes^{(Reference Tedstone, Targett and Mackinlay9,Reference Tedstone, Targett and Owtram10)} , a tiered soft drinks industry levy (SDIL)⁽¹¹⁾ and restrictions on placement of HFSS foods at aisle ends, checkouts and storefronts⁽¹²⁾. Although further legislation has also been passed to restrict price promotions on HFSS foods and further limit advertising on TV and online⁽¹³⁾, implementation of these actions has been delayed until at least 2025⁽¹⁴⁾.

Additional evidence on the impacts of food marketing and associated public health action

Papers in the current special collection of BJN add to our understanding of the influence of food marketing on diet, as well as the potential for government regulation of food marketing to support healthier diets. In an updated systematic review and meta-analysis, Boyland et al. find a small but significant effect of exposure to food marketing on consumption^{(Reference Boyland, McGale and Maden7)}. In a qualitative interview study with marketing experts, Harris et al. report that whilst the UK’s new restrictions on online HFSS advertising are considered ‘groundbreaking’ in their ambition, experts also think they are likely to be ‘entirely ineffective’ in reducing total exposure to HFSS advertising^{(Reference Harris, Boyland and Muc15)}. A key reason given for delay of these restrictions, as well as those on price promotions, is inflationary pressure and the ‘cost-of-living’ crisis.

Reformulation is a potential outcome of many policies that restrict marketing – by limiting marketing to foods that meet a definition of ‘healthier’, companies are incentivised to reformulate their products and portfolios to be ‘healthier’ and so avoid the restrictions. Gressier et al.’s updated review finds that mandatory reformulation policies are more effective than voluntary ones for both changing the food supply and intakes^{(Reference Gressier, Frost and Hill16)}. In a second paper, Gressier et al. compared the effects of the SDIL and the sugar reduction programme, finding that the SDIL had greater impacts on total sugar purchased. Whilst both strategies triggered reformulation, the SDIL additionally encouraged consumer switching – perhaps because it received much greater publicity and so had a signalling as well as price effect^{(Reference Gressier, Frost and Hill17)}. Finally, Li et al. extend the concept of the SDIL to foods, demonstrating the additional potential food taxes could have in the UK^{(Reference Li, Gressier and Hill18)}.

Focusing research where it is likely to make the most difference

Alongside the important contributions of the papers in this special collection, they also raise a number of important questions. Boyland et al.’s headline findings replicate those of numerous previous reviews and reviews of reviews^{(Reference Boyland, Nolan and Kelly2–Reference Coleman, Hanson and van Rens4)}. Whilst Boyland et al. show that research is now beginning to evolve, as food marketing has already done, into digital spaces^{(Reference Boyland, Muc and Coates5)}, we have known for some decades that food marketing ‘works’. Similarly, Gressier et al.’s finding that mandatory approaches to reformulation are more effective than voluntary ones replicates many previous findings documenting the limitations of voluntary and self-regulatory approaches to supporting dietary public health^{(Reference Coleman, Hanson and van Rens4,Reference Erzse, Karim and Foley19,Reference Ronit and Jensen20)} . Confirmatory research and replication are important. But we must be alert when replication and confirmation turn into research redundancy.

As a community of evidence commissioners, funders, creators and users, we must focus more clearly on reducing ‘critical uncertainties’ and generating the additional evidence that is likely to make the most difference to decision-making^{(Reference Ogilvie, Adams and Bauman21)}. This requires careful judgements about when we have enough evidence of the right sort to support action, as well as recognition of the limits of scientific evidence in different contexts. Whilst the ‘hierarchy of evidence’ privileges the type of meta-analyses of randomised controlled trials that Boyland et al. have adeptly delivered, not all research questions are amenable to randomised controlled trials. Randomised controlled trials are an ideal method for assessing, for example, whether acute exposure to TV food advertising influences preferences, choices and consumption. But they are not necessarily well placed to help understand the impacts of restrictions on TV food advertising on exposure to food marketing as a whole. Addressing such a question would require randomised implementation of restrictions on TV food advertising across a number of different jurisdictions – a possible but, in practice, unlikely study design to achieve. In these contexts, natural experimental methods are likely to be more useful^{(Reference Ogilvie, Adams and Bauman22,Reference Craig, Cooper and Gunnell23)} .

Food marketing is part of a complex adaptive system

Randomised controlled trials of the acute impacts of food marketing exposure are also not necessarily predictive of the impacts of restrictions on food advertising. As Harris et al. beautifully illustrate, food marketing is part of a complex system which adapts to events with newly emergent behaviours^{(Reference White, Aguirre and Finegood24)}. Harris et al. report that restricting online product-based marketing (e.g. of specific food items) is likely to lead to increased brand marketing (e.g. of logos closely associated with particular items) leaving total exposure to food marketing unchanged^{(Reference Harris, Boyland and Muc15)}. This has been referred to as the ‘balloon theory’ of marketing – where using restrictions to squeeze one type of marketing leads to compensatory expansion of other types^{(Reference Mora25)}.

Thinking about food marketing as a complex adaptive system of interacting parts is one way to help predict, and hence act to avert, unintended consequences of restrictions. In the case of the UK’s online restrictions, an obvious conclusion is that brand advertising should also be included. This has been resisted on the basis that it would be difficult to define an ‘unhealthy’ brand (which tends to represent a range of both more and less healthy products). An alternative framing suggested by Harris et al. is that all brand and product food advertising – irrespective of ‘healthfulness’ – infringes children’s right not to be economically exploited and so should be restricted^{(Reference Harris, Boyland and Muc15)}.

Commercial food interests do not necessarily align with public health interests

A clear understanding of the value of additional research to support decision-making may be particularly important with respect to commercial systems such as the food industry. The interests of commercial systems focus on growth and profit. In the context of less healthy food, these interests do not necessarily align well with public health interests of improved health and well-being and reduced inequalities^{(Reference White, Aguirre and Finegood24)}.

Claims of the absence of exactly the right sort of evidence are a well-documented tactic of harmful commodity industries to delay the implementation of restrictive legislation^{(Reference Lacy-Nichols, Marten and Crosbie26)}. Allowing industry to demand more and more evidence on what we already know wastes scarce public resources. Similarly, allowing industry to demand that action is delayed until ‘perfect’ evidence of effect is available distorts the nature of public health policymaking, which relies on iterative cycles of taking action based on the best available evidence of what is likely to work, followed by evaluation, and refinement and development of further action^{(Reference Ogilvie, Adams and Bauman21)}. Although it is true that the food industry will have to be ‘part of the solution’ in terms of implementing public health policy to support healthier eating, this does not mean that they should necessarily be allowed to play a role in determining what policy should be implemented, when and based on what evidence^{(Reference Lacy-Nichols and Williams27)}.

Ineffective interventions are not necessarily harmless

Despite their finding that experts expect the UK’s planned ban on online HFSS advertising to be ineffective, Harris et al. conclude that is a good ‘first step’, indicating to other countries that such action is possible and challenging them to build on the UK’s experience^{(Reference Harris, Boyland and Muc15)}. Whilst it is certainly the case that implementing one action can make further, additional, actions ‘possible’^{(Reference Keeble, Burgoine and White28)}, it is not necessarily the case that ineffective public health action is harmless^{(Reference Macintyre and Petticrew29)}.

Passing any legislation requires political capital to be used and the delays in implementing restrictions on price promotions and the online HFSS advertising ban indicate that such political capital is not always forthcoming. Like any resource, political capital is not unlimited and expending it on one thing may also undermine the potential for other, possibly more effective, policies being implemented. Critics of ineffective policies can easily frame them as indicative of an ill-informed research, advocacy and policy community who are unable to make good policy suggestions.

Voluntary agreements with the food industry are a good example of the delaying impact of ineffective policy. In 2011, the UK government proposed a range of voluntary agreements with the food, and other, sectors to support public health under the ‘Public Health Responsibility Deal’. Messaging around the Deal indicated that if it was not successful, mandatory regulation would follow. Reflecting Gressier et al.’s findings that voluntary actions tend to be less effective than mandatory^{(Reference Gressier, Frost and Hill16)}, evaluation of food sector actions within the Public Health Responsibility Deal indicates that these focused on actions that prioritised industry over public health interests, were already underway and had known limited effectiveness^{(Reference Knai, Petticrew and Durand30,Reference Knai, Petticrew and Douglas31)} . Despite the limitations of the Public Health Responsibility Deal, mandatory action did not begin to occur until the 2018 implementation of the SDIL.

Whole-food reformulation is the next big challenge

Restricting the marketing of less healthy foods can incentivise manufacturers to reformulate existing products or rebalance their portfolios towards healthier alternatives^{(Reference Reyes, Smith Taillie and Popkin32,Reference Saavedra-Garcia, Meza-Hernández and Diez-Canseco33)} . In the UK, a nutrient profiling model is used to distinguish healthier from less healthy products based on whether or not they are HFSS⁽³⁴⁾. As Gressier et al. and Li et al. both describe, there is substantial potential for policies such as taxes on food and drinks and the sugar reduction programme to trigger reformulation and achieve changes in population purchasing and consumption of, in these cases, sugar^{(Reference Gressier, Frost and Hill17,Reference Li, Gressier and Hill18)} . Nutrient-based epidemiology indicates that this would be likely to be beneficial for public health⁽³⁵⁾.

However, the great majority of products impacted by these policies are likely to be ultraprocessed^{(Reference Monteiro, Cannon and Levy36)}. Reformulated ultraprocessed foods and soft drinks remain ultraprocessed. If we take a food-based, rather than nutrient-based, approach to dietary public health policy, the aim of policy should be to encourage consumption of less processed foods, rather than to reduce consumption of particular nutrients^{(Reference Adams, Hofman and Moubarac37)}. Both nutrient- and food-based goals could be achieved by manufacturers replacing HFSS ultraprocessed products with non-HFSS less processed alternatives. This has been termed ‘whole-food’, in contrast to ‘nutrients-to-limit’, reformulation^{(Reference Scrinis and Monteiro38)}.

Whole-food reformulation providing large-scale access to affordable non-HFSS less processed foods poses significant challenges across the food system. It demands that different foods are grown, different products manufactured in very different ways and may require substantial changes in food practices among citizens. How this could be achieved on a large scale in an affordable way that minimises social inequalities and maintains business viability is the next big challenge for our dietary public health research, advocacy and policy communities.