• Accessibility
• Email alerts
• RSS feeds
• Contact us

Recipe for health: a plan to fix our broken food system Contents

---

Chapter 4: Ultra-processed foods

Context and definitions

130.In view of the intensifying public and scientific debate about their role and importance, a core element of our remit has been to investigate the role that ultra-processed foods (UPFs) play in relation to diet and obesity.

131.Professor Carlos Monteiro and his colleagues at the University of São Paulo proposed the NOVA classification of food groups in 2009, as set out in Box 4.187

Box 4: The NOVA classification of food groups

NOVA classifies commercially available foods and drinks into four groups based on their preparation method and the degree of processing they have undergone. They are as follows: Group 1: Unprocessed and minimally processed foods. Unprocessed foods are described as the “edible parts of plants” (for example, fruits, leaves, roots, seeds, stems and tubers) or “from animals” (such as muscle, offal, eggs and milk), and also “fungi, algae and water, after separation from nature”. Minimally processed foods is used to describe unprocessed foods that have been altered by industrial processes, such as chilling, freezing and powdering. For example, frozen vegetables, herbs and spices. Group 2: Processed culinary ingredients. These are defined as “substances obtained directly from Group 1 [unprocessed and minimally processed] foods or from nature by industrial processes”, such as extracting, pressing and refining. Examples of processed culinary ingredients include honey and vegetable oils. Group 3: Processed foods. This describes products made by adding oil, salt, sugar or other processed culinary ingredients to unprocessed and minimally processed foods, using preservation methods. For example, bottled or canned vegetables in brine, salted nuts, and smoked fish and meats. Group 4: Ultra-processed foods (UPFs). These are defined as “formulations of ingredients, mostly of exclusive industrial use, typically created by [a] series of industrial techniques and processes”. UPFs include baby formula, sweetened breakfast cereals, carbonated soft drinks, mass-produced packaged breads, margarine and other spreads, and reconstituted meat products, as well as some meat substitute products and non-dairy milks. The production of UPFs involves multiple stages. It starts with the breakdown of whole foods into components like sugars, oils, fats, proteins, starches and dietary fibre. These components are frequently derived from a select group of high-yield plant sources (such as beet, corn, soy, sugarcane or wheat) and from the processing of animal carcasses, typically sourced from intensive animal farming.

Source: House of Lords Library, Impact of ultra-processed food on children’s health, InFocus, 9 October 2023—citing Food and Agriculture Organisation (FAO) of the United Nations (UN), Ultra-processed foods, diet quality and human health using the NOVA classification system (2019), pp 13–14: https//openknowledge.fao.org/server/api/core/bitstreams/5277b379-0acb-4d97-a6a3-602774104629/content#page=13 [accessed 20 September 2024]

132.Professor Monteiro said that “the purpose of making ultra-processed foods is different [from other foods]: it is to create affordable novel products to replace NOVA groups 1, 2 and 3, while at the same time maximising profits and making irresistible and overconsumed products.” He added that a key aspect of the definition of ultra-processed foods is therefore that they are designed to be affordable, to be overconsumed, and to maximise profits, explaining the worldwide commercial success of ultra-processed foods.188 We explore the role of the food industry in relation to UPFs further below.

133.Dr Chris van Tulleken stressed that the definition of ultra-processed foods was not developed as a consumer tool or a policy instrument, but rather to test the hypothesis that an influx of industrially processed foods was causing an epidemic of diet-related disease in Brazil. Dr van Tulleken conceded that the NOVA definition was not perfect, and was developed initially to describe foods available in Brazil rather than the UK. It was nevertheless a helpful “way of describing, in the tightest way possible, the category of industrial food” and “a very powerful way of describing our terrible diet.”189

134.Dr van Tulleken added that “a good working definition would be that, if a food is wrapped in plastic, and has at least one ingredient that you do not typically find in a domestic kitchen … then it is probably an ultra-processed food.” He argued that “the evidence applies to the category and to the dietary pattern, not to the individual product. … The degree to which ‘worse’ or ‘better’ products can be disentangled is missing the main issue of UPFs.”190

135.Henry Dimbleby likewise warned against a hyper-literal approach to UPFs focusing on individual products such as wholemeal bread.191 Dr Vicky Sibson, Director of First Steps Nutrition Trust, agreed that focusing on applying the UPFs definition to individual products was missing the point: “We just need to turn it on its head and … consider the fact that we know that our children from the earliest age are not eating enough whole fruits, vegetables and fibre.”192

136.Professor Ian Young told us that, in the July 2023 SACN statement on processed foods and health,193 SACN concluded that, of the various definitions of UPFs put forward, only the NOVA classification had potential at the moment to be applied in the UK. However, SACN noted that not every food in NOVA category 4 is equally unhealthy, and some may be beneficial as part of a healthy and balanced diet. Similarly, SACN had expressed significant concerns about the health consequences of some foodstuffs in the other categories, such as processed red meats (which are in NOVA Group 3).194

137.DHSC stated that the NOVA UPFs definition is controversial because it captures a wide range of foods, including many which are typically considered ‘less healthy’, such as sugary drinks, salty snacks and confectionery. However, UPF also captures products that are generally considered ‘healthier’ such as some fortified foods, sliced wholemeal bread, baked beans, low fat yogurts, vegetable sauces and higher fibre breakfast cereals.195

138.Professor Janet Cade argued that the UPF definition was never designed for consumer use, is very nebulous and is difficult to define.196 Dr Charlotte Mills and Professor Gunter Kuhnle, Department of Food and Nutritional Sciences, University of Reading, suggested that the NOVA definition was ambiguous and inconsistent, in particular regarding additives.197 A number of studies have examined the extent to which academics agree on which foods should be classified as UPFs for research purposes. One recent French survey asked approximately 200 nutrition specialists to code independently two lists of foods (generic and marketed) using the NOVA classification. The authors found relatively low levels of agreement between specialists for both lists.198 However, others have indicated a high level of concordance (>80%) between independent coding of foods as UPFs or not in research studies.199

139.Jenny Chapman argued that, since the UPF definition included healthy and unhealthy food and was never intended to group foods on the basis of nutrition, it was unfit for use in nutrition policy.200 Professor Susan Jebb said that “ultra-processed food is a binary system, very yes or no. That means we can spend an awful lot of time arguing which side of the line a particular food falls on.”201

140.The food and drink industry were particularly vocal in their criticisms. The British Soft Drinks Association said that the NOVA classification criteria had been “critiqued for ambiguity and inconsistency, hindering consistent application and comparison of research findings.”202 The Food and Drink Federation made similar comments.203

Levels of consumption of UPFs

141.Population intakes of ultra-processed foods have been estimated using dietary survey data. SACN stated that estimates of UPFs intake in the UK based on the National Diet and Nutrition Survey (NDNS) ranged from 51% of energy intake in adults aged over 19 years to 68% among 12 to 18 year-olds.204 Professor Franco Sassi, Chair in International Health Policy and Economics, and Dr Eszter Vamos, Senior Clinical Lecturer in Public Health Medicine, cited Imperial College London research drawn from NDNS data between 2008 and 2017, which found that, among primary school-aged children, 61% of calories in school lunch come from UPFs, rising to 70% among secondary school-aged children.205 Dr Vicky Sibson said that NDNS data likewise suggested that 61% of the total mean energy intake of two to five year-olds in the UK comes from UPFs.206 Dr Chris van Tulleken claimed that, in the UK, 57% of calories are from ultra-processed food, and it is very normal to eat 80%.207

142.Dr Mathilde Touvier cited statistics showing that globally the UK and the US had the highest UPFs consumption rate, at nearly 60% of energy intake (see Figure 8). In this respect, the UK and the US appear to be significant outliers.208

Figure 8: Mean proportion of ultra-processed foods in adults’ diet across countries (% energy intake) from nationally representative surveys

Source: Supplementary written evidence from Dr Mathilde Touvier (Research Director at Inserm) (FDO0143)

143.The Soil Association noted that UPF consumption is consistently higher in more deprived groups in the UK, while consumption of fresh and minimally processed foods is consistently lower.209 There is evidence in a study using the National Diet and Nutrition Survey 2008–19 that adolescents from lower socio-economic backgrounds, of white ethnicity and from the north of England consumed more ultra-processed foods than average.210 A recent analysis of NDNS data for both children and adults, including the years 2008–18, found consistently higher UPF consumption in more socio-economically deprived groups (assessed using household income, educational attainment, and the index of multiple deprivation (IMD)).211 Inequalities were wider among younger compared with older participants.

Comparing the definitions of UPFs and HFSS foods

144.One of the key questions in assessing the significance of these statistics is the extent of overlap between UPFs and food already considered less healthy according to UK dietary guidelines and regulation.

145.Government guidelines on healthy diet are based on evidence that excess consumption of calories, saturated fat, salt and sugar are associated with higher obesity and diet-related disease.212 SACN advice is that eating too much sugar increases risk of obesity and tooth decay, while eating too much salt raises blood pressure, increasing risk of heart disease and stroke. Cutting down on saturated fat generally reduces the risk of cardiovascular disease and coronary heart disease, lowers cholesterol and improves glycaemic control.213

146.Government dietary advice (reflecting SACN’s advice) is encapsulated in the UK’s national food model, the Eatwell Guide, which “shows how much of what we eat overall should come from each food group to achieve a healthy, balanced diet.” The Eatwell Guide encourages a balance of fruit and vegetables, starchy carbohydrates, sources of protein and dairy/dairy alternatives, accompanied by healthy fluids. It encourages the choice of unsaturated oils and spreads eaten in small amounts, and states that foods high in fat, salt and sugar (HFSS) should be eaten less often and in small amounts.214

147.HFSS foods are defined for regulatory purposes by the UK Nutrient Profiling Model (NPM). DHSC stated that the UK NPM 2004/5 was developed as a tool to enable Ofcom to identify ‘less healthy’ foods and drinks that are subject to restrictions during children’s television programming. Points are allocated on the basis of the nutrient content of 100g of a food or drink. Points are awarded for ‘A’ nutrients (energy, saturated fat, total sugar and sodium) and for ‘C’ nutrients (fruit, vegetables and nut content, fibre and protein). The score for ‘C’ nutrients is then subtracted from the score for ‘A’ nutrients to give the final nutrient profile score. In most cases, the score is calculated for products as sold; however, where products are reconstituted before being consumed (e.g. a cake mix) the score is calculated based on the reconstituted product. A score over four is defined as HFSS. Ofcom has been using this model for broadcast media since the restrictions came into force in April 2007 and for non-broadcast media since July 2017.215

148.The NPM is now also used to define HFSS foods for the purposes of restrictions on location promotions in retail settings. It will similarly be used to define HFSS foods for the purposes of new regulation on the advertising of less healthy foods on television and online and on volume-based price promotions from 1 October 2025.216 Specific categories of HFSS products (such as breakfast cereals, savoury snacks and confectionery) are subject to these restrictions on the basis that they are of most concern in relation to childhood obesity.217

149.DHSC stated that, therefore, “there is already an established method of identifying and defining HFSS foods and drinks that is used by government, businesses and academia.” They noted that “a number of foods that are considered as UPF are also HFSS, which means there is some overlap between the two definitions.”218

150.Some witnesses discussed the quantitative overlap between UPFs and HFSS foods. Dr Mathilde Touvier reported an analysis of more than 130,000 products on the French market, and said that when UPFs were ranked according to the French Nutri-Score system (based on an algorithm derived from the Nutrient Profiling Model219), 63.2% of ultra-processed foods received the lowest nutritional quality marks, whereas only 5.3% of ultra-processed food had the highest score.220 Dr Chris van Tulleken estimated “there would be an overlap of some 80% to 90% between the HFSS Ofcom definition and UPFs”.221

151.To provide data relevant to the UK, an analysis of the overlap between HFSS foods and UPFs was conducted by University of Cambridge academics222 to explore the extent to which policies based on identifying HFSS foods (such as the Ofcom food advertising restrictions) would also apply to UPFs. This analysis considers all HFSS foods. As such, it is not limited to the specified categories of HFSS food subject to current restrictions on location promotions, and planned restrictions on volume promotions and television and online advertising. Data from the annual UK National Diet and Nutrition Surveys from 2008–2019 was analysed:

(a)using the number of each food consumed (represented as percentages);

(b)using the number of calories in the foods consumed (also represented as percentages); and

(c)using the weight of foods consumed.

The analysis included 1.7 million products (excluding alcohol and supplements) representing 4,555 unique food names that were consumed by 15,655 individuals over the study period. We have summarised the findings for numbers of foods and calories (energy) in foods in Figure 9.

Figure 9: Venn diagrams showing the foods classified as UPF, HFSS, both and neither by (a) percentage of foods, and (b) energy (kilocalorie, kcal) content

Source: MedRxiv—the preprint server for Health Sciences, Overlap between ultra-processed food and food that is high in fat, salt or sugar: analysis of 11 annual waves of the UK National Diet and Nutrition Survey 2008/09-2018/19 (28 August 2024): https://www.medrxiv.org/content/10.1101/2024.08.27.24312650v1.full-text [accessed 22 September 2024]

152.The analysis found that, in terms of the food allocated to each category, around 56% of UPFs would be identified using the classification of HFSS foods. However, in terms of energy consumed, the overlap is slightly higher at around 59%. In terms of number of foods, around 50% are neither HFSS nor UPFs, but when assessed using energy content, this is much lower at around 28%. This is because many of the foods that are neither UPFs nor HFSS are whole foods such as fruits and vegetables, which are relatively low in calories. When assessing the overlap around 16% of foods are UPFs, but not HFSS; this proportion is around 25% when assessed using energy content. These are the foods that would be missed by policies that are driven solely by the 2004/05 Nutrient Profiling Model.

153.Drawing on this analysis, Table 1 shows the kinds of foods that fall into this category, as well as the kinds of foods that are both UPFs and HFSS, and neither. The table shows the food groups that make significant contributions to each of these categories. It does not imply that all products in each food group fall into any one category. In that context, it is of note that foods that are UPFs but not HFSS, and hence would not be covered by current regulations, include various manufactured breads; manufactured potato products; yoghurt, fromage frais, and dairy desserts; pre-packed cooked vegetables; and low calorie soft drinks.

Table 1: Food groups contributing to the categories ultra-processed food only, high in fat, salt and sugar only, both and neither

UPF only	HFSS and UPF	HFSS only	Neither HFSS nor UPF
Manufactured white, brown, granary, wheatgerm and wholemeal bread	Manufactured biscuits	Sugars, preserves and sweet spreads	Semi-skimmed milk
Manufactured chips, fried and roast potatoes, and potato products	Soft drinks, not low calorie	Butter and cheeses	Unprocessed fruit
Yoghurt, fromage frais and dairy desserts	Crisps and savoury snacks	Whole milk	Pasta, rice and other cereals
Pre-packed cooked vegetables	Chocolate confectionery	Bacon and ham	Other potatoes, potato salads and dishes
Soft drinks, low calorie	Manufactured buns, cakes, pastries and fruit pies	Nuts and seeds	Pure fruit juices

Source: This table has been derived from tables 1–3 in MedRxiv—the preprint server for Health Sciences, Overlap between ultra-processed food and food that is high in fat, salt or sugar: analysis of 11 annual waves of the UK National Diet and Nutrition Survey 2008/09-2018/19 (August 28 2024): https://www.medrxiv.org/content/10.1101/2024.08.27.24312650v1.full-text [accessed 22 September 2024]. Because the lists of commonly consumed foods in each column by number, energy content and weight differ slightly, only foods common to the same column in all three tables in the paper have been used to populate Table 1 in this report. The labels for food groups used in the three tables in the paper have been slightly modified for this Table for clarity.

154.Artificial sweeteners (of which the FSA states there are 22 licenced for use in the UK223) replace sugars and offer a non-caloric alternative. They are mostly used in liquids (e.g. soft drinks) and semi-liquid foods (e.g. desserts), but also some confectionery. Reducing sugar reduces calorie intake and glycaemic load, which reduces risk of obesity and type 2 diabetes. Non-nutritive sweeteners, however, have also been shown to have effects on appetite and glycaemic control, so some experts consider them to be problematic and are concerned about their increasing use (e.g. as a result of the Soft Drinks Industry Levy (SDIL)).224 A 2022 WHO systematic review has concluded that while results of randomised controlled trials have generally suggested non-sugar sweeteners may have little impact on glucose metabolism and result in lower body weight when coupled with energy restriction in the short-term, there is no clear consensus on whether sweeteners are effective for long-term weight loss or maintenance, or if they are linked to other long-term health effects at intakes within the acceptable daily intake.225

155.The number of milk and meat alternatives that are not of animal origin has grown considerably over the last decade, in part due to health concerns (e.g. lactose intolerance), and in part due to the rise of vegetarianism and veganism and increased awareness of the environmental implications of meat and dairy consumption. Thus, these replacement foods, while largely UPFs, offer important alternatives for those choosing to avoid eating dairy and meat. We explore this further below.

156.Current dietary advice recommends the consumption of whole wheat or whole grain breakfast cereals and breads rather than low fibre or sugary cereals and white bread. While most of these products are UPFs, they are generally considered to form part of a healthier dietary pattern in terms of the range of foods routinely affordable in the UK grocery market.

UPFs and the food industry

157.In Chapter 3 we explored the role and responsibility of the food industry in relation to the obesogenic food environment, and in particular how the ‘junk food cycle’ incentivises the food industry to produce and sell unhealthy products. In the context of UPFs, Henry Dimbleby argued that this creates significant systemic obstacles to the creation of a healthier food environment:

“After the Second World War we created a new form of farming that meant that the cheapest forms of calories were refined vegetable oil, refined wheat and refined sugar. … We can see ultra-processed food as a way of taking those cheap calories and making it palatable by adding as little other food as possible. In some senses, that was a huge success: we do not have the malnutrition that that form of farming was invented to stop. … But it is impossible not to draw the conclusion that the way in which that has developed is now causing huge harms.”226

158.Dr Rob Ralston said that over the past four decades—the period over which obesity has risen rapidly—the ultra-processed food industry has doubled in size, and is now worth over $1.1 trillion. He said that it was dominated by key corporations with familiar consumer brands, including Nestlé, Pepsi, Coca-Cola, Mondelez, Kraft Heinz and Danone. He said that UPFs were important in their portfolios because they are very palatable, have very long shelf lives, and are highly marketable. He said that multinational corporations devote huge sums of money to marketing ultra-processed foods. He said that higher profits come from processed foods in comparison to fresh foods or whole foods, meaning it is often not in the interests of corporations or their shareholders to promote healthier food products.227

159.Professor Franco Sassi agreed that products whose price per calorie has been reduced the most are UPFs. This has been achieved through large-scale production, standardisation of products and standardised ingredients, which has allowed companies to market their brands and products on a large scale. He said that this has made highly industrially processed products dominant in diets and the food market.228

160.Dr Chris van Tulleken argued that diet-related diseases are fundamentally “commerciogenic”: caused by the marketing and consumption of industrially processed foods.229 He asserted that, as with the approach of the tobacco industry, “almost all the doubt, the anxiety and the uncertainty” over the impact of UPFs “comes from industry-funded voices.”230 However, Professor Eric Robinson said that this was a deceptively simple narrative, and that there was widespread uncertainty in the academic community around whether the level of processing was the problem.231

161.The food industry also sought to refute these arguments. Sam Fulton, Group Director of Corporate Affairs and Sustainability at Nomad Foods, warned against using processing as a proxy for good or bad nutrition.232 The Food and Drink Federation argued that food processing provided for increased choice and availability of foods throughout the year; increased shelf life and stability of foods; more convenience; reformulation to reduce fat, salt and sugar in foods; and fortification to help the population reach nutritional recommendations. They said that a focus on UPFs conflicts with Government policies to support healthier choices, including on reformulation and fortification.233

162.The British Retail Consortium likewise stressed the benefits of processing.234 Andrew Opie also argued that the retail sector’s concern about UPFs was that “we need to take customers with us on this journey. … Are we going to confuse customers by talking too much about UPF and then losing the benefits with them that we have started to build up around the HFSS issue?”235

Public attitudes

163.The FSA’s Consumer Insights Tracker showed that UPFs have been in the top three issues of concern to consumers since August 2023, with approximately 75% of those surveyed expressing concern. The FSA’s Advisory Committee on Social Science has stated that it will conduct a further rapid evidence review on consumer understanding, knowledge, behaviours and information needs with respect to UPFs.236

164.Professor Robin May noted that, while ultra-processed foods generally have risen considerably in people’s concern, there has also been a rise in public confusion. Professor May stressed the FSA’s role through its consumer information work “to set out very clearly what the facts are, what we know and, in particular, what we do not know.”237

165.The Food, Farming and Countryside Commission cited the findings of its citizen engagement project, The Food Conversation. It showed that citizens want tighter controls on the availability and marketing of UPFs and they were not confused by the definition of UPFs—“for many it seemed intuitive”. There was support for transparency on the ingredients within UPFs, more information for citizens on the harms of UPF consumption, targets to reduce substantially UPFs in shops, and greater intervention to protect children from unhealthy food and UPFs. As we have seen, the Food, Farming and Countryside Commission argued that businesses should pay for the damage UPFs cause through a polluter pays type policy.238

166.The British Dietetic Association cited data from the Institute of Grocery Distribution (IGD) showing that only two in five claimed to understand what UPFs means.239 Several food industry representatives argued that introducing the concept of UPFs in policymaking would be confusing for consumers.240

167.Nevertheless, Professor Susan Jebb acknowledged that “UPF has caught the public imagination in a way that our talk of HFSS, frankly, never did. We might want to stop and consider how to use that opportunity and this moment in time to reconnect people with the fundamentals of food.”241

Health outcomes

168.The debate over the health outcomes of consumption of UPFs has focused on the strength of the evidence that different types of research study provide.242 Appendix 5 summarises the different types of research study cited in this report, including laboratory studies, observational studies (including cross-sectional and longitudinal or cohort studies), experimental studies (including trials and natural experiments), ways of synthesising evidence (including systematic reviews and meta-analyses), and modelling studies.

SACN’s findings

169.In its scoping review of the evidence on food processing and health, SACN concluded that the association between higher consumption of (ultra-) processed foods and adverse health outcomes is concerning, but it is unclear whether these foods are inherently unhealthy due to processing, due to their nutritional content or due to a combination of factors. SACN concluded that:

• “Systematic reviews consistently reported that increased consumption of UPF was associated with increased risks of adverse health outcomes (including obesity type 2 diabetes, cardiovascular disease, depression, mortality and a range of maternal and child health outcomes).” However, “the available evidence was almost exclusively observational (and therefore unable to show causation.”
• “Consumption of (ultra-) processed foods may be an indicator of other unhealthy dietary patterns and lifestyle behaviours.”
• “It is unclear to what extent observed associations between (ultra-) processed foods and adverse health outcomes were explained by established relationships between nutritional factors and health outcomes”.243

170.Professor Ian Young said that it was possible that there was “an added problem with processing above and beyond nutritional content … but at the moment we cannot find sufficient basis to say that there is a separate and additional effect of ultra-processed foods once their nutritional content was considered.”244 SACN has added the topic of processed foods to its watching brief and considered it at its meeting in June 2024, and in a subsequent horizon scanning meeting in October 2024.245

171.Some disputed the findings of the SACN review, and, as we discussed in Chapter 3, argued that it demonstrates the need to ensure that SACN is free from industry influence. Rob Percival argued that SACN’s conclusions “may have been skewed by industry ties, conflicted financial interests, and a narrow framing of the science. … These declared interests do not imply corruption or bias on the part of SACN members, but they illustrate how pervasive are industry ties at the interface of science and policy.”246

172.Others defended the integrity of SACN’s work. Dr Alison Tedstone argued that in upholding standards of scientific review SACN had not been able to include some sources of information such as “small statistically under-powered RCTs, in-vitro and unpublished studies.” She stressed that SACN was clear that increasing consumption of UPF is consistently associated with worsening health outcomes. In her view, “SACN has made a fact-based statement on evidence on UPF, rather than giving a biased opinion.”247

173.DHSC cited SACN’s work, and stated that “based on their nutritional content, most people are likely to benefit from reducing their consumption of many foods classified as UPF, which … are captured by existing government policies.”248 Dame Andrea Leadsom MP, said that “the problem is that we do not know whether it is ultra-processed foods themselves that are the danger, or whether the danger is that so many foods high in fat, sugar and salt are themselves ultra processed.”249 Preet Kaur Gill MP expressed similar caution.250

Further evidence to the Committee

174.Witnesses agreed there was a correlation between consumption of UPFs and adverse health outcomes. However, there were divergent views on whether those adverse health outcomes were the effects of the high energy density or high levels of fat, salt, or sugar in many UPFs, or whether other properties of UPFs were responsible.

175.Charis Bridger Staatz, Senior Research Fellow, and Rachel Seabrook, Senior Policy and Public Affairs Manager at the Centre for Longitudinal Studies, University College London, argued that there was a correlation between increased consumption of UPFs and increased rates of obesity:

“Over the 1980s there was mass expansion of industrialised food production, and more ready introduction of ultra-processed foods into our diets. As younger people experience additional exposure to obesogenic environments, it is likely that these trends in BMI development will increase.”251

176.Dr Chris van Tulleken argued that “we have satisfied the Bradford Hill criteria”252 showing UPFs are causally associated with negative health outcomes, including obesity, type 2 diabetes, cancer, cardiovascular disease, non-alcoholic fatty liver disease253, dementia and all-cause mortality. He drew attention to 32 systematic reviews and meta-analyses of cohort studies linking UPFs and negative health outcomes, arguing that they showed an association that was “specific (in the sense of there have been large sudden population exposures in Latin America following the North American Free Trade Agreement), temporal (population health changes following exposure) and there is a dose response relationship254 seen between UPF and harm.” He added that “we never had that with tobacco.” Dr van Tulleken cited Professor Kevin Hall’s randomised control trial,255 as well as more than 1000 clinical and pre-clinical studies on properties and additives unique to UPF linking them to negative health outcomes in humans and animals.256 We note that randomised control trials provide the strongest evidence of causality, but are not always feasible and may have their own limitations (e.g. small sample sizes and limited length of follow-up) (see Appendix 5).

177.Rob Percival cited the British Medical Journal’s publication of an umbrella review of 45 meta-analyses encompassing studies involving almost 10 million people, which he said found a direct association between UPFs and multiple health outcomes.257 We note that the study rated 22 studies as low quality and 19 as very low quality.258 Professor Camila Corvalán accepted this, but stated “that is how epidemiology judges observational studies”. She argued that the BMJ evidence was nevertheless “very consistent and reliable”, and “stands when you adjust for nutrient content, suggesting that the mechanisms are related not only to the sugar, salt and fat content.”259

178.Dr Mathilde Touvier said that the work with the NutriNet-Santé cohort260 in France since 2018 had shown associations with an increased risk of cancer, cardiovascular disease, type-2 diabetes, mortality, depressive symptoms, obesity and gastrointestinal disorders. She asserted that extant epidemiological studies adjusted for nutrient intakes, energy intake, salt, sugar, fibre and vitamin intake still observed associations between ultra-processed food and health outcomes: “we have many things yet to discover, but, still, it is time for action, even with what we know now.”261

179.On the other hand, Professor Janet Cade said that most evidence on UPFs comes from large-scale long-term cohort studies that were never designed to look at UPFs, but rather at nutrients. She also said that a recent systematic review on UPFs and overweight showed that studies tended not to adjust for total energy, and where they did adjust, the effect was dramatically reduced.262 Dr Keren Papier, Senior Nutritional Epidemiologist at the University of Oxford, said that there was limited information from studies on the pattern of consumption of UPFs, and in particular what processing had taken place.263 Dr Charlotte Mills and Professor Gunter Kuhnle argued that there is a paucity of reliable data from observational, clinical and in vitro studies to support the claims of an adverse effect of ultra-processed foods beyond their nutrient profile and possibly their texture.264

180.Professor Robin May said that ultra-processed foods are such a broad category that it is not scientifically valid to look at them versus everything else: “You need to break it down into the particular combinations and types of ultra-processed foods.”265 Professor Susan Jebb said that while research was starting to do this, “we are a long way off being able to link individual foods to particular health outcomes.”266 Professor Jebb accepted that “there is plenty of observational evidence linking UPF to adverse health outcomes”, but stressed that “we have to think about whether the adverse health outcome associated with UPF is being driven by something that you are eating or something that you are not eating instead.”267

181.Several food industry representatives argued that the scientific evidence linking UPFs to poor health was limited. Danone UK and Ireland argued that there is insufficient evidence to demonstrate any causal link between the degree of food processing and negative health outcomes: “based on the currently available evidence, we assert that what makes a product unhealthy or otherwise is its nutritional qualities, not its level of processing.”268 Nomad Foods269 and Ella’s Kitchen270 made similar arguments.

182.Some witnesses drew attention to evidence suggesting that not all ultra-processed foods are associated with adverse health outcomes. Professor Mike Rayner, Professor of Population Health at the University of Oxford, pointed to recent cohort research showing an association between higher UPF consumption and an increased risk of multimorbidity but only for certain categories of UPFs, i.e. animal-based processed products and artificially and sugar-sweetened beverages, but not ultra-processed breads and cereals or plant-based meat alternatives.271

183.Marlow Foods Ltd, trading as Quorn Foods, warned against classing all highly processed foods as unhealthy regardless of their nutritional value. They said that their portfolio of vegetarian and vegan mycoprotein containing products that are high in fibre and protein, while falling into the NOVA 4 category, had been shown by research to provide health benefits including a lowering effect on total and LDL272 cholesterol and an emerging benefit on gut health.273 The Good Food Institute Europe,274 the Alternative Proteins Association,275 Vegetarian Society,276 Vegan Society,277 Sarah Nájera Espinosa et al, London School of Hygiene and Tropical Medicine278 and Pro Veg UK279 made similar arguments.

184.The Vegetarian Society also stated that such meat and dairy alternatives provide core nutritional value to specific demographic sub-groups including allergy sufferers, vegetarians, vegans, and those seeking to reduce meat and dairy by switching to plant-based alternatives. They argued that for those sections of society, plant-based meat and dairy alternatives represent accessible sources of fibre and key nutrients including protein, calcium and vitamin B12.280

Possible causal mechanisms

185.Professor Eric Robinson said that while observational studies show that high consumption of UPF as a dietary pattern is predictive of worse health, “we do not really know why that association exists.”281 Professor Susan Jebb stressed that “understanding the mechanisms will help policy”.282

186.In seeking to identify potential causal mechanisms linking UPF to poor health outcomes, several witnesses cited the randomised control trial published in 2019 and led by Professor Kevin Hall and colleagues into whether ultra-processed foods affect energy intake.283 In this trial, 20 participants were randomised to be fed on unprocessed or ultra-processed diets, each for a period of two weeks, and then swapped to the other diet for a further two weeks, while remaining confined in a research facility. Diets were matched for total calories, macronutrients, fibre, sugars, and sodium. The ultra-processed and unprocessed meals were matched for energy density when including beverages, but the food in the ultra-processed diet was more energy dense.284 The ultra-processed diet had more added sugar compared to total sugars, insoluble fibre compared to total fibre, and saturated fat compared to total fat. Participants could eat as much or as little as they wished. On the ultra-processed diet, participants ate more and gained an average of 0.9kg over two weeks.

187.Professor Hall explained that the study:

“was not designed specifically to look at mechanisms but simply to understand whether, if you took two diets that varied to a huge extent in their composition of ultra-processed foods but matched for various nutrients—such as salt, sugar, fat and fibre—overall glycaemic load, overall energy density and things like that, we would see differences in how many calories people chose to eat when they were asked to eat as much or as little as they wanted.”285

188.The study had a number of limitations, some of which are acknowledged in the paper. These include the limited control for previous dietary habits of adult participants—all were paid volunteers who were able to spend four weeks in an in-patient clinical research facility. The study had a limited length of follow-up and thus could not determine the effects of long-term exposure to either of the diets tested. Meals were provided to participants without choice, which does not reflect most people’s usual daily experience with food. In many other respects the study was unable to replicate the eating experience of free living subjects.286

189.Notwithstanding these limitations, Professor Hall summarised its findings as follows:

“The surprise to me was that, despite matching for the various nutrients of concern that we match for, people spontaneously chose to eat many more calories on the ultra-processed diet, gaining weight and body fat, whereas, when the same people were exposed to a food environment that had no ultra-processed foods but were matched for various nutrients of concern, they spontaneously lost weight and lost body fat.”

190.Professor Hall said that his team had identified two factors which they think are likely most important for driving excess calorie intake: energy density and hyperpalatability.287

Energy density

191.Energy density is defined as the number of calories per gram of solid food. Henry Dimbleby said that UPFs have an energy density that is not normally found in foods cooked from scratch. At the same time, insoluble fibre tends to be taken out.288 Professor Susan Jebb said that her own research with Professor Andrew Prentice had demonstrated the importance of the energy density of food, and the significant impact of water content upon this.289 Dr Chris van Tulleken suggested that the soft texture of many UPFs drives excess consumption.290

192.Professor Ian Young said that SACN had found that energy density was an important plausible mechanism for adverse health associations, notwithstanding the need for further research. SACN had also considered whether the texture of UPFs promoted a faster eating rate.291

Hyperpalatability

193.Professor Hall noted that, in his study, despite the overall diets being matched for sodium, carbohydrates, fat and sugar, “it turned out that a number of individual foods that we presented to people exceeded pairs of thresholds—sodium plus fat, or fat plus sugar, or carbs plus sodium. When pairs of thresholds are exceeded, they are so-called hyperpalatable foods.”292

194.Professor Tim Spector argued that many ultra-processed foods “are designed by the industry to make us overeat—that is the definition of hyperpalatability—through the combination of fats, sugars and particularly salt in exactly that right combination”.293 Professor Carlos Monteiro said that ultra-processed foods are engineered to be overconsumed “by combining pre-tested proportions of fat, sugar and salt, by making products that are consumed very quickly … and by using flavours and flavour enhancers that make products irresistible.”294

195.In a 2023 study published in Nature Food, Dr Tera Fazzino reported that energy density, eating rate and hyperpalatability were consistently associated with meal energy intake across all diets.295 On the other hand, the University of Bristol Nutrition and Behaviour Unit said that it was remarkable that “the often-reported claim that UPFs are more palatable than unprocessed offerings” was widely held, when only one study published in 2024 directly tested this idea. They said that their research found no evidence that level of processing or food energy density predicts food liking and food reward.296

196.Professor Ian Young said that SACN found that palatability was a plausible causal mechanism of adverse health outcomes, but again stressed that more research was needed.297 From the point of view of the food industry, Sam Fulton said that “palatability and the taste of food needs to be taken into consideration by food manufacturers, because there is an enjoyment factor. … Food is not meant to just be something that you endure.”298

Additives

197.Food additives are substances added to processed foods, or other foods produced on an industrial scale, for technical purposes, such as to improve safety, to increase shelf life, or to modify sensory properties.299 They include emulsifiers, which are used to bind ingredients to support the structure, texture and flavour of foods.300 Professor Carlos Monteiro said that cosmetic additives—the flavours, colourants, sweeteners, thickeners and emulsifiers that are necessary to create these products to mimic the properties of whole foods—are a key feature of UPFs.301

198.Some witnesses said that additives cause over-consumption. Professor Barry Smith said that “irresistible savoury snacks” contain additives that trick the body to expect to receive nutrients and then to over-consume when they do not arrive. He also cited research showing that when consuming artificial sweeteners “our bodies are thinking that we will get sugar and no sugar comes. … the body will start storing those calories more because it is getting a signal that it was going to get a lot of that energy and did not get it, so it starts trying to find and chase the calories.”302 He warned that additives eaten every day in large volumes are subject to very little scrutiny.303

199.We previously referred to arguments that the softness of UPFs causes over-consumption. Henry Dimbleby referred to “the extraordinary softness that emulsifiers bring, which makes us eat more.”304

200.Others pointed to evidence linking additives to poor health outcomes. The World Health Organization (WHO) has found that non-sugar sweeteners do not reduce body fat in the long term and may be linked to long-term increased risk of type 2 diabetes, cardiovascular disease, and adult mortality. A WHO guideline recommends against the use of non-sugar sweeteners to replace free sugars for weight control.305

201.Dr Mathilde Touvier said that her research306 had shown associations between a higher intake of several artificial sweeteners, preservatives and food emulsifiers, and a higher risk of hypertension, cardiovascular diseases, type-2 diabetes, and some cancers. She said that randomised trials in humans and animal studies also showed the adverse effects of food additives on metabolism, microbiota and inflammation.307 On the other hand, Dr Charlotte Mills and Professor Gunter Kuhnle questioned this evidence.308

202.Dr Chris van Tulleken said that, on the basis of such evidence, the threshold for use of these additives should be much higher. He argued that if there is any suspicion of synthetic emulsifiers causing harm, they should be subject to stringent regulations.309 He also said that it was misleading to describe such products as UPF wholemeal bread (which he described as “emulsified foam”) as healthy compared to “actual wholemeal bread”.310 Professor Tim Spector argued that the FSA was falling down on the job in terms of highlighting the problems with emulsifiers. He called on the FSA to adopt the precautionary principle in relation to their regulation.311

203.However, Professor Robin May said that new additives have to be approved by the FSA, and they are extensively assessed for their individual impact: “We look at possible DNA damage, liver and kidney impacts. I think I can say, hand on heart, that individual components are very carefully assessed and safe.” He said that the FSA would not hesitate to act in response to compelling evidence of harm. However, he conceded that considering combinations of additives was much more difficult: “In the United Kingdom, we have something in the order of 320 licensed additives. Even if you had only two additives always at the same concentration in the food, you are looking at something like 100,000 combinations that would need testing. The scale is quite overwhelming.” Professor May also said that there was no system worldwide that he was aware of for monitoring the effect of long-term exposure to additives. He said that finding a way to collect such data in a proportionate way would be welcome.312

204.Professor Ian Young said that SACN had considered the possible effects of specific additives or groups of additives, such as emulsifiers and low-calorie or non-calorie sweeteners. However, it concluded overall that there was insufficient evidence that UPFs have a greater impact than HFSS foods.313 In June 2023, SACN said it would scrutinise the WHO guideline on non-sugar sweeteners to see if additional assessment was needed.314

Other issues relating to the make-up of UPFs

205.Witnesses raised other issues related to the make-up of UPFs. Katharine Jenner, Director at the Obesity Health Alliance, said that the low levels of fibre in many UPFs meant they lacked nutritive value:

“As you process food, nutrition gets degraded. … The biggest and most telling thing that … gets taken out, or at least processed to such an extent that it has no benefit, is the fibre. These products all score very poorly in terms of nutrition because they are very low in fibre.”315

206.Cinnamon Days Community Interest Company cited evidence that micronutrient intake declines with higher consumption of ultra-processed foods. They cited NDNS data highlighting micronutrient deficiencies in young people and adults (in particular girls and women) in iron, selenium, magnesium and potassium.316 Henry Dimbleby also cited evidence that the micronutrient content of UPFs may differ from other foods, although he stressed that “we are in the foothills of that science”, and “we do not yet know what that means causally”.317

207.Dr Mathilde Touvier also argued that changes to the food matrix during processing may have a variety of effects on the body’s absorption of nutrients, including a faster rate of energy intake, and a reduction on the bioavailability of nutrients. She also argued that food processing can produce potentially toxic compounds, and that there may also be a migration of contaminants from contact packaging.318

Convenience and affordability

208.Think Through Nutrition said:

“UPFs and HFSS foods are typically more affordable than minimally processed foods due to the lower cost of their artificial, mass-produced ingredients and the economies of scale in processing. These ingredients not only reduce production costs but also extend the shelf life of products, ensuring their availability and reducing retail prices. This pricing structure makes UPFs more accessible, particularly to those in lower income brackets, who might rely on these cheaper, more accessible food options. The constant availability and lower prices drive up consumption rates of these less healthy options.”319

209.The Food Foundation’s Broken Plate report highlights that the most economically disadvantaged groups would need to spend a disproportionate amount of their disposable income (50%) to meet dietary guidelines, compared to just 11% for the least deprived.320 By way of illustration, Food Foundation ambassador Dominic Watters told us that, where he lived in Kent, “the shop on the estate only sells the lowest quality of processed food, making it a food desert in the garden of England.”321 We discuss this further in Chapter 7.

The need for further research

210.Many witnesses stressed the necessity of further research into UPFs. Dr Tazeem Bhatia, Chief Nutritionist at DHSC, stressed that the “really concerning” association between processed foods and poor health outcomes needs further research, in particular to work out what components are of concern and breaking it down into food groups.322

211.SACN called for:

• “further assessment and development of a (ultra-) processed foods classification system that can reliably be applied to estimate consumption of processed foods in the UK.”
• “Further evidence exploring relationships between (ultra-) processed foods and health outcomes” including:
  • “Good quality randomised controlled trials (RCTs) that may help to identify potential mechanisms and establish whether they are independent of energy density or other dietary factors.”
  • “Good quality prospective cohort studies that can address concerns relating to confounding and reverse causality for observed associations between (ultra-) processed foods and health outcomes.”
  • “Good quality studies that consider the benefits of consuming products with minimal processing in comparison with existing UK dietary recommendations and/or other dietary patterns for which there is evidence of beneficial health outcomes.”
  • “Assessment of any role of food additives or other processing methods in observed associations between (ultra-) processed foods and health.”
• “Further assessment and refinement of NDNS methodology to better estimate and monitor processed food consumption, while minimising impact on participant burden.”

212.SACN noted that the Office for Health Improvement and Disparities (OHID) is engaging with UK funders of research, including the National Institute for Health and Care Research (NIHR), to identify priority areas.323 In that context, we welcome the fact that in July 2024, NIHR released a call for research on the following question: “What are the health and health inequality impacts of interventions that affect consumption of ultra-processed foods?”324

213.Professor Robin May said that in work with the Chief Medical Officer the FSA had identified a large evidence gap, in particular around carefully controlled human studies such as that conducted by Professor Kevin Hall. He said that there is now a very active conversation with the wider research community, and in particular with UK Research and Innovation (UKRI) as the biggest spender on research, about how best to target research to fill some of the evidence gaps.325 In that context, we note the roundtables facilitated by the Government Office for Science in November 2023.326

214.UKRI stressed the need for:

• An improved mechanistic understanding of the links between processed foods and health across the life course.
• Innovation in food manufacturing, to improve nutritional content and understand factors underpinning food composition, texture and palatability including understanding any unintended consequences of processing on health and the environment.
• A better understanding of the biological, social, cultural and environmental drivers that influence diet choice, including UPF.
• Strategies to facilitate the translation of fundamental research to applied nutrition research to realise the potential health impacts for different population groups.327

215.Henry Dimbleby said that given the harms that our diet is causing us, it was an “absolute scandal” that many more studies of the type undertaken by Professor Kevin Hall had not been undertaken.328 Professor Susan Jebb agreed that “we have plenty of observational epidemiology, but now we need to understand the mechanisms” through experimental research.329 Professor Hall said that his team is undertaking a new trial with the goal of teasing apart the relative effects of energy density and hyperpalatable foods to better understand what aspects of the ultra-processed food environment drive excess calorie intake.330 We note that a number of other registered trials are either under way or planned.331

The policy response

216.Against this backdrop, witnesses debated the merits of a range of policy responses.

Dietary guidelines and the Eatwell Guide

217.DHSC stated that modelling indicates that if everyone ate in line with existing UK dietary recommendations as illustrated by the Eatwell Guide, population life expectancy could be increased by eight years. DHSC argued:

“Based on their nutritional content, most people are likely to benefit from reducing their consumption of many foods classified as UPF, which the Eatwell Guide already shows are not part of a healthy, balanced diet. … Further research is needed however before any firm conclusions can be drawn about the extent to which processing itself is the problem and therefore before any significant updates to the Eatwell Guide and government dietary recommendations would be considered.”332

218.DHSC stated that “given the broadness of the current classification, a blanket statement to limit all UPF foods is unlikely to be helpful and may be particularly difficult to achieve for those on lower incomes or less able to find time to cook at home.”333 Dr Tazeem Bhatia, Chief Nutritionist at DHSC, emphasised that wholemeal sliced bread, fortified flours and fortified cereals, which all come under NOVA Group 4, are part of the Eatwell Guide’s advice for a healthier diet: “we do not know what the unintended consequence might be of asking people to consume a lot less of these foods.”334

219.Professor Ian Young said that SACN was particularly focused on nutritional intakes “because that is where the vast majority of high-quality evidence currently lies”. He noted that the Eatwell Guide, which seeks to interpret SACN’s nutritional advice, “is effectively a food-based dietary guideline”.335

220.Some witnesses argued that the Eatwell Guide did not need to take account of UPFs. Professor Janet Cade said that, compared to the UPFs definition, it was a useful tool that offered “good, clear guidance” on the balance of foods that we should be consuming.336 The British Nutrition Foundation noted that foods such as some processed meats, butter, cakes and biscuits made at home from minimally processed foods are excluded from what is considered to be ‘ultra-processed’, but the Eatwell Guide states that consumption of them should be limited.337

221.Sam Fulton said that by focusing on processing “you therefore end up demonising some perfectly healthy foods and healthy food products”, which could lead to perverse outcomes in terms of guidance.338 Dairy UK likewise argued that using UPF definitions to shape policies could potentially have a detrimental impact on the consumption of nutrient-rich foods or beneficial ingredients.339

222.On the other hand, Professor Tim Spector said that the sole focus on nutrients in dietary guidance was “hopelessly out of date” and “behind the times”, noting that “there is nothing in them about eating whole foods … with real structure and fibre … rather than heavily processed foods.”340

223.Rob Percival said that while the Eatwell Guide already encourages eating lots of fruit, veg, pulses, and wholegrains, “there is now a rationale for making it explicit that we are talking about less highly processed foods.”341 The Obesity Health Alliance342 and Intelligent Health343 made similar recommendations.

224.Professor Lucilla Poston, Professor of Maternal and Fetal Health at King’s College London, argued that “saying ‘we prefer our children to eat a healthy diet that’s composed of fresh fruit and vegetables and fresh food, unprocessed’, is the right, commonsensical approach at the moment, because to unravel all this and to come up with a [UPF] definition will take years.” She called for a public health message to recommend fresh food.344 Kristin Bash agreed, arguing that “when more than half of your dietary calories are coming from ultra-processed foods … we know that the healthier foods are being crowded out of the plate, if you will.”345

225.Professor Tim Spector said that it is a misconception that fresh vegetables are healthier than tinned or frozen ones. He pointed to findings showing that most frozen vegetables have as many or sometimes more nutrients than fresh ones, and said that cans of vegetables are often extremely nutritious. He stressed that “families on real budgets should be eating these things rather than lots of white bread and ready meals.”346

226.The Food Foundation noted that the term “ultra-processed” is featured in the national dietary guidelines or nutrient profiling models of Belgium, Brazil, Ecuador, Israel, Maldives, Peru, and Uruguay. In each of these cases, the consumption of UPFs is discouraged and diets rich in unprocessed and minimally processed alternatives which are low in fat, salt, and/or sugar content are promoted.347

227.Although he said that it was not appropriate to use the definition of UPFs to regulate individual foods, Dr Chris van Tulleken did argue that the evidence has reached a threshold where it would be safe and sensible to advise reducing the proportion of UPF in the diet.348 He also drew attention to a number of international examples of dietary guidelines, as set out in Box 5:

Box 5: International examples of dietary guidelines

Belgium: Choose as few ultra-processed products as possible. UPFs have no real added value in a healthy and environmentally responsible diet. Brazil: Avoid UPFs. Always prefer natural or minimally processed foods and freshly made dishes and meals. Do not offer UPFs to children. Canada: Limit highly processed foods and drinks because they are not part of a healthy eating pattern. If you eat highly processed foods, try to: eat them less often; eat them in small amounts; replace them with healthier options. Ecuador: Avoid the consumption of UPFs. France: Limit sugary drinks, fatty, sweet, salty and ultra-processed foods. Avoid giving ultra-processed products to under-threes. Avoid giving commercial baby foods and ready meals to under-threes. Israel: Reduce the consumption of UPFs as much as possible. Avoid processed, industrialised and packaged foods as much as possible. There is no need to buy food made especially for children and babies. Mexico: Avoid UPFs. UPFs promote preference for very sweet or salty flavours, and increases the risk of both obesity and malnutrition. Peru: Avoid the consumption of UPFs. Uruguay: Base your diet on natural foods and avoid the regular consumption of ultra-processed products with excessive contents of fat, sugar and salt.

Source: Supplementary written evidence from Dr Chris van Tulleken (FDO0145)

228.Still others acknowledged wider deficiencies with the Eatwell Guide. The British Dietetic Association and others cited Food Foundation findings that the cost of eating in line with the Eatwell Guide in households with children where annual income is less than £15,860 would require 42% of disposable income.349 The British Retail Consortium said that the guidance was not flexible enough to take account of naturally HFSS products such as sunflower oil which nevertheless bring health benefits.350 Nomad Foods said that the Eatwell Guide was last reviewed eight years ago, and should be updated to include a stronger focus on sustainable diets and cultural diets.351

229.More broadly, witnesses highlighted that changing dietary guidelines does not by itself translate to improving nutrition. The Food and Drink Federation pointed to evidence that less than 0.1% of the UK population meet all the underpinning nutrient recommendations in the Eatwell Guide.352 Professor Christina Vogel said that dietary trends in the UK at the moment do not match the Eatwell Guide. She highlighted consumption of too many foods that are high in sugar, with sugar intakes being more than double current recommendations; eating too few portions of fruit and vegetables, with only a third of adults and 12% of children meeting the five-a-day recommendations; and eating far too few wholegrain foods, meaning that fibre intake is very low.353

230.Preet Kaur Gill MP stressed that the Eatwell Guide needed to be made more useful and accessible to citizens if it was to have any impact: “Why is it not linked to the NHS app? Why is it not digital? Why is information not available to people who have type 2 diabetes?”354 Nomad Foods called for simpler and more practical explanation of the Guide, underpinned by a nationwide UK promotional campaign.355

Updating the Nutrient Profiling Model

231.Our witnesses discussed whether the Nutrient Profiling Model should be updated to take account of UPFs. DHSC argued that “there is currently no objective, universally agreed method to determine UPF products in the same way there is for HFSS products. Any method for determining UPF products would need to be robust enough to apply to policy decision making.”356

232.Preet Kaur Gill MP said that the Nutrient Profiling Model can be effective in reducing obesity, and that more research and science was needed on the impact of UPFs.357

233.The George Institute for Global Health noted growing interest in combining UPF and HFSS definitions.358 Professor Barry Popkin, Professor of Nutrition at the University of North Carolina, proposed combining HFSS and UPF criteria to develop a definition of less healthy foods for regulatory purposes, by starting with HFSS criteria and adding elements of UPF criteria, including colours and flavours.359 The Soil Association called for them to be employed in tandem, and said that the Government should update the Nutrient Profiling Model to incorporate markers of processing.360 Food Active361 and the Obesity Health Alliance362 made similar recommendations.

234.Priya Sinha Kenward called for the Government to consult on revisions to the Nutrient Profiling Model as part of an overall Nutrition Index, showing in a single, objective score the healthiness of a food and how frequently it should be consumed. This would include the nutrient density of the food and the level of processing.363 Think Through Nutrition364 and Dr Mathilde Touvier365 made similar proposals.

235.Food industry representatives argued for retention of the existing NPM. Morrisons argued that the 2004/05 Nutrient Profiling Model allows businesses to work towards a level playing field, and is the basis of promotions and advertising restrictions. However, they acknowledged that the model is far from perfect.366 The British Retail Consortium367, British Soft Drinks Association368, Food and Drink Federation369 and James Mayer370 made similar points.

236.Dr Charlotte Mills and Professor Gunter Kuhnle argued against incorporating UPFs into the NPM, as nutrient profiling tools use an objective measure to characterise foods, and are used by many countries. On the other hand, “UPF requires the interpretation of a complex, and sometimes contradictory, set of rules. … NOVA does not take food composition into consideration and therefore does not make any a priori claims about impact on health.”371

237.However, the NPM is not a perfect system, and underpins policies that present opportunities for unhelpful reformulation. Dr Chris van Tulleken noted that a product could be high in salt and sugar and yet not be defined as HFSS if its fruit and vegetable content, fibre or protein improved the score. He said that the debate between the relative role of nutrients and processing in driving health harms “misunderstands that using high levels, and particular ratios, of fat, salt and sugar are part of the industrial processing that drives excess consumption and negative health outcomes.”372 He said that the Nutrient Profiling Model was “a complex, subjective definition of food. … there is no list of HFSS products that you can access as a member of the public. … if we just used our thresholds for fat, salt, sugar and saturated fat, and calories, from our national dietary recommendations, we would be limiting ultra-processed foods.”373

238.Professor Susan Jebb said that the NPM can be used with a binary cut-off point as with advertising, but as it gives a continuous score “you have the potential to demonstrate that there is a spectrum of healthiness rather than a black and white situation.” Professor Jebb was concerned that throwing the NPM out in favour of attempting to define UPFs would lead to delays in necessary action. She preferred to see the Nutrient Profiling Model evolve if it needs to do so.374

239.Diabetes UK warned that “the misplaced tension between these two approaches provides fertile ground for exploitation by the food and drink industry to delay, cast doubt, and derail current policy. Nutrient-based and processing-based approaches are complementary and overlapping.”375 Dr Alison Tedstone said that the HFSS classification system is established in UK legislation and could not easily be replaced by a more complicated UPFs system. That said, she acknowledged that as a nation we are being sold, and eating, too much UPF and HFSS foods.376

240.Beyond UPFs specifically, several witnesses expressed concern that the NPM had not been updated. Anna Taylor said that the process for updating the NPM was “unforgivably slow.”377 Cancer Research UK observed that the NPM is now nearly 20 years old and outdated. They noted that the proposed 2018 modification reflects updated SACN dietary guidance on free sugars intake and on fibre intake,378 but had yet to be adopted by the UK Government partly due to industry lobbying.379 Professor Susan Jebb380 and Professor Christina Vogel381 also said that the NPM needed to be updated to reflect lower dietary recommendations for free sugars. The Alcohol Health Alliance noted that alcoholic drinks are not included in the definition of HFSS foods for regulatory purposes, even though they make an important contribution to energy intakes.382

241.DHSC stated that an updated NPM was consulted on in 2018 but has not yet been finalised.383 DHSC Director of Health Improvement Natasha Burgon said that the existing NPM had been deliberately chosen as the basis for current regulations to avoid unnecessary delay, but that the department was looking to update this as soon as it can.384

Additional policy and regulatory measures

242.We heard that the Government must act now to drive down consumption of unhealthy foods, without waiting for more evidence to emerge on ultra-processed foods. Dr Alison Tedstone warned that an HFSS versus UPFs debate could be used as an excuse to delay taking action because “not even experts agree”.385 Professor Camila Corvalán warned that the need for further research on ultra-processed foods must not become an excuse for inaction: “if we keep waiting to have all the evidence—all the RCTs or all the mechanisms involved—that will probably take decades. … The health of our population will probably be at a chaotic level.”386

243.DHSC argued that that many foods classified as UPFs are already captured by existing and forthcoming measures restricting HFSS product placement, advertising and price promotions. They said that such products are also likely to be the focus of efforts to reformulate products high in calories, sugar and salt, the Soft Drinks Industry Levy, and calorie labelling regulations for food sold in large out-of-home businesses.387 However, as we noted above, such a regulatory approach will not capture any of the potential independent effects of UPFs beyond the nutritional content of the subset that are also HFSS.

244.Professor Susan Jebb argued that existing policy interventions could be adapted to take emerging evidence on ultra-processed foods into account.: “if we understood the mechanism then our policymaking could become more sophisticated.” However, she was not convinced that the arguments on UPFs had opened up new policy options. She argued that any emerging evidence on the harmful effects of UPFs could be dealt with through existing tools, such as product reformulation to tackle nutritional concerns; the FSA risk assessment process regarding the safety of additives; advertising restrictions to tackle the food industry’s large promotion budget; and restricting availability or portion control to tackle hyperpalatability.388

245.However, Professor Carlos Monteiro argued that product reformulation was at best a partial, incomplete solution to the problem of UPFs, as it only dealt with the nutrient content of a product, rather than other factors such as hyperpalatability. Furthermore, UPF products remain UPF even when reformulated.389 We discuss reformulation further in Chapter 5.

246.Dr Chris van Tulleken said that, while evidence around UPFs should help inform policy, as it tells us about both the dietary pattern and the companies that need regulation:

“The definition UPF is not a regulatory tool. To define harmful food that should be subject to regulation we can look to official UK Dietary Recommendations. Since most UPF has excessive amounts of sugar, saturated fat, total fat or salt according to these recommendations, they can be used to regulate non-UPF or so called ‘clean label’ products high in nutrients of concern.”390

247.Witnesses discussed whether the precautionary principle now justified firmer action in relation to UPFs. Professor Tim Spector asserted that, while the mechanisms were not yet fully understood, there was enough evidence of harm to adopt the precautionary principle and start to warn people to avoid UPFs.391

248.However, Professor Robin May said that he saw no current scientific justification to apply the precautionary principle for ultra-processed foods, and warned against a “knee-jerk reaction”: “When you apply the precautionary approach, it needs to be done in balance with the risk associated with it.”392

249.In terms of additional regulatory responses, some witnesses discussed labelling of UPF products. Dr Chris van Tulleken said that he did not “know anyone credible who is talking about using that [NOVA] definition to slap labels on things”.393 Instead, he advocated a new labelling requirement based on a nutrient profiling model with a maximum of five warning labels for sugar, total fat, saturated fat, trans fat394 and/or sodium. He said that using this, 95.1% of ultra-processed products would be required to feature at least one warning label.395 Intelligent Health396 and Think Through Nutrition397 argued that there was a strong case for including specific indicators for UPFs as part of a mandatory front of pack labelling scheme.

250.Industry representatives warned against using a UPFs definition in labelling, citing uncertainties over the definition and quality of evidence.398 The British Soft Drinks Association said that there were not aware of any global labelling approach that considers the level of processing or uses a UPFs definition. They said that the often cited Chilean ‘warning label’ scheme focuses on labelling HFSS products, using a nutrient-based approach.399 We consider labelling more generally in Chapter 5.

Conclusions and recommendations

251.The concept of ultra-processed foods (UPFs) is a matter of intensive public discussion and scientific debate. The evidence we have heard suggests that the concept of UPFs (based on the NOVA classification) can be a useful tool for describing the way in which the food system incentivises the production and marketing of cheaply produced, highly palatable, energy dense and nutritionally poor foods, and drives unhealthy diets and obesity. However, it is widely considered that the NOVA classification of UPFs lacks sufficient precision to be suitable for the characterisation or regulation of individual foods. There therefore remains debate over the extent to which it should be used in policymaking.

252.Notwithstanding these limitations, the rapidly growing epidemiological evidence showing a correlation between consumption of UPFs defined using the NOVA classification and poor health outcomes is alarming. There is a significant overlap between UPFs and foods high in calories, saturated fat, salt and sugar and low in fruit, vegetables, nuts, fibre and protein (HFSS). Many UPFs are therefore already considered less healthy according to the UK’s dietary guidelines and the regulations they underpin. Beyond energy and nutrient content, causal links between other properties of UPFs and poor health outcomes have not thus far been clearly demonstrated and the existence of such links remains scientifically uncertain. There are strong views on both sides of the argument over whether industry has influenced the scientific research on UPFs.

253.We welcome the commitment of the scientific community in the UK and beyond to undertake further research into the links between UPFs and poor diet and health. Working with bodies such as UK Research and Innovation (UKRI) and the National Institute for Health and Care Research (NIHR), the Government must commission further research independent of industry into these links, and in particular to explore:

• Potential mechanisms of causation linking UPFs and poor health outcomes;
• The health benefits of consumption of products with no or minimal processing on both overall dietary patterns and on health outcomes;
• The impact of the hyperpalatability and energy density of foods on dietary consumption;
• The long-term health effects of consumption of specific additives and combinations of additives; and
• The cost-effectiveness of interventions that can reduce consumption of less healthy, often highly processed foods, and increase the availability and affordability of healthier foods.

The Government and SACN must monitor and respond urgently to any emerging evidence.

254.The need for further research into ultra-processed foods must not be an excuse for inaction. The food industry must be held to account now for the rise in unhealthy diets and obesity. The Government has already taken some steps to regulate HFSS products, many of which are also UPFs. It must now go much further. We set out how in Chapter 5.

255.Furthermore, the Government should commit as part of its new comprehensive food strategy to tackling the over-consumption of such less healthy foods. The Government should also commit to increasing consumption of healthier, largely unprocessed and minimally processed foods, ensuring a healthy and affordable diet for all.

256.This will not be easy. Since less healthy, highly processed foods often represent the affordable and convenient option, they can crowd out healthier options from the diet, particularly for those facing food insecurity. Furthermore, many households have neither the time, skills nor facilities to prepare meals from scratch. The Government should therefore take action to make healthier, largely unprocessed and minimally processed foods more accessible and affordable. It should also explore how best to enable the preparation of meals and snacks at home from scratch as often as personal circumstances allow. In the following chapters we set out proposals on what elements such a strategy should include.

257.In addition, the Government must within two months publish a detailed response to the July 2023 SACN statement on processed foods and health, and any subsequent findings published by SACN. This should set out the Government’s current position on UPFs, and in particular: whether and how national dietary guidelines should reflect the need for caution about eating a diet containing a high proportion of UPFs; and the current evidence on their impact on public health. Any guidance on consumption of UPFs should be accompanied by an operational definition for individual foods that is robust and easily useable both by consumers and by the food industry.

258.The Eatwell Guide published in 2016 should be reviewed to ensure it reflects updated evidence and advice, and further reviewed as required in the future if further evidence on the public health impact of UPFs emerges. The Government should review and implement the most cost-effective ways of making the Eatwell Guide useful and accessible to consumers (for instance via interactive digital platforms).

259.The Government must implement the 2018 Nutrient Profiling Model without further delay. Going forward, the NPM must also be regularly reviewed to ensure it reflects emerging scientific evidence and dietary guidelines, including any emerging research evidence on the potential harms of UPFs or their properties, and used to identify both healthier and less healthy foods for the purposes of regulation.

---

• A-Z index
• Glossary
• Contact us
• Freedom of Information
• Jobs
• Using this website
• Copyright
• Privacy notice
• Cookie policy
• Cookie Manager