Unpacking Ultra-Processed Food (UPF)

Key Takeaways

Understanding Ultra-Processed Foods (UPFs) and the Nova Classification System
UPFs are characterised by high levels of processing, affordability, convenience, and marketing. The Nova classification system categorises foods by their level of processing, but challenges remain, particularly when applied to consumption data due to limited detail. Inter-rater variability and inconsistent interpretations further complicate its use.

Health Impacts of UPFs and the Evidence Base
High UPF consumption is consistently associated with negative health outcomes, including obesity, cardiovascular disease, and mental health issues. Potential mechanisms include poor nutrient quality, altered food matrix, and exposure to harmful additives and contaminants. However, stronger evidence, particularly from RCTs, is needed to clarify causal pathways.

Burden of UPFs on European Populations
UPF consumption is highest among adolescents, young people, males, and individuals from lower socioeconomic backgrounds. Economic models suggest that reducing UPF intake could result in major healthcare savings, as some countries like France and Belgium begin to incorporate UPF reduction into national guidelines.

Marketing of UPFs and Youth Exposure
Transnational food corporations use sophisticated digital strategies to target children and adolescents with UPF advertising. This exposure undermines healthy eating behaviours and reinforces diet-related inequalities. The WHO emphasises that the power and reach of marketing demand stronger public health responses.

Food Marketing and Policy Gaps
Evidence shows that children’s food choices, preferences, and consumption are influenced by UPF marketing. WHO guidelines advocate for mandatory marketing restrictions over voluntary industry pledges. Current nutrient profile models may not capture all unhealthy foods, highlighting the need to integrate broader indicators of food healthiness into policy frameworks.

Equity Considerations and Barriers to Healthy Eating
Reducing UPF intake across different income groups requires more than education – it needs structural change. Minimally processed foods often lack the marketing appeal of UPFs and may be less accessible in low-income areas. A multifaceted policy approach is required to reduce barriers and promote healthier dietary environments.

Future Directions and Next Steps

• Strengthen the evidence base by prioritising RCTs investigating causal links between UPF consumption and health outcomes
• Support WHO-aligned, government-enforced restrictions on UPF marketing, especially those targeting children and adolescents
• Adapt existing nutrient profiling models to better capture the full range of unhealthy UPFs, including brand marketing and ultra-processed characteristics
• Design equity-focused intervention/policies that specifically reduce UPF consumption in vulnerable groups by improving access to minimally processed foods
• Promote cross-sector collaboration by aligning efforts across public health, policy, and research to address the commercial determinants of diet and reshape retail environments

Transcripts are auto generated, if you spot an error, please email enquiries@easo.org

Speaker 1
00:00 – 03:11
Hello, everyone, dear ECN members. It’s my pleasure to welcome you again on another e-learning hub event for this month. And today we have a very interesting subject. I’m sure that will be very attractive to all of you. We are going to talk about ultra-processed foods have become a central focus in discussions on public health due to their widespread consumption and potential effects on health. This webinar will explore various aspects of ultra-processed food consumption, including its socio-economic patterns, classification systems like NOVA, and the impact on healthcare costs. I would like to remind again that the NOVA Nordisk Foundation has provided support to EYASO for early career network development activities, including this webinar series. Zanova Nordisk Foundation has had no influence over the content. I would like to introduce our board members. I will start from myself. My name is Dr. Emil Hidayatli. I’m a medical doctor, obesity care professional, practitioner and I also want to introduce my colleagues Dr. Eugenio Romano, Lisa Heiji and would like to remind you again that this webinar is being recorded and at the end you will get the link in order to if you miss some points you can get the link to see the videos. Also, I would like to remind that this is the e-learning hub online events that usually held every month. And ECN is free to join. So please, we are welcome all of you. Also, tell your colleagues, your earlier network colleagues to join ECN and to join the webinars every month. So again, the webinars are held every month, and you have a great chance to be the part and to get the knowledge, to share the knowledge. Also, I would like to remind that we will have today two presentations, each by 20 minutes. And then we will have question and answer 20-day sessions. So feel free to start to write the questions to the chat. close to the end of our one hour meeting. And please don’t forget that we would like to have some feedback from you. It’s very important to us to take feedbacks in order to be support and develop the future webinars. Now I would like to hand over to my colleague from ECN board to give some news. So please Eugenia, hand over to you.

Speaker 2
03:12 – 05:08
Thank you, Emil. Welcome, everyone. I’m Eugenia Romano. I’m a researcher at King’s College London. And as Emil said, I’m also a member of the ECN board. So there’s a few opportunities that I would like to share. And today I’m going to focus on ICO, because in May 11th to 14th, we have the European Congress on Obesity in Mallorca, Spain. So we hope to see most of you, if not all of you there. And like every ECO, DCN will gather to share research and promote and participate in networking and career development opportunities. So keep an eye for updates. The office of submission is closed, but you may still register to join ECO as a delegate. And ECN has organized some activities during lunch times. On the first day, we’ll have a speed date networking. So we’ll basically give everyone an opportunity develop an elevator speech to quickly introduce yourself, which hopefully will help you all to network during the Congress. On the second day, you’ll have a chance to meet the winners of the Nobel Laureate Foundation New Investigator Award, and you will be able to ask about their experience in an informal setting, which is going to be a more expanded version of what they will do later on in the update session for during the programme. And on third day you will connect and collaborate with your peers in your field so basically you’ll have an opportunity to connect with anyone that works in say basic science, mental health and to have discussion groups according to your research area. We also have awards and prizes at ICO so we have two new prizes that will be announced and are open to early career researchers presenting their

Speaker 3
05:08 – 05:09
abstracts ECO.

Speaker 2
05:10 – 06:54
One is the ECO 2025 prize for industry impact and the other one is the ECO 2025 prize for public engagement and science communication. So the winner will receive a €3,000 grant for career development and a travel grant for the next ECO in 2026. And as I mentioned before there’s also the EAS and Novo Nordisk Foundation New Investigator Awards. So there will be the 2025 award session where you can hear about this year’s winner and there will be a 2024 update session in the ECN lounge as well. So there’ll be interviews with individual winners and senior researchers in their field so if you want to come and hear about their experience in the past year since winning the award that’s a good chance to do it. Of course check the ECN website for all other opportunities because we have many. We have exchange program opportunities, we have the ECN masterclass, we have travel grants, we have spotlights, so if you want to be spotlighted and be promoted, the ECN can do it for you on social media. And also, I want to remind everyone that we have a WhatsApp group, which is no spam. So if you want to be in touch and know anything new going on about ECN, please join us. And also, if you’re interested about becoming an active role in organizing ECN activities, remember that there’s applications to be a member of the board, and they’re open until, well, the 4th of April. So it’s passed, but maybe there’ll be more opportunities in the future. So hope to see most of you at ICO, and I’ll pass it over once again to Emil to introduce the speakers for today.

Speaker 1
06:55 – 07:40
Thank you very much, Eugenia. So today we have three brilliant speakers. I would like to introduce them, the Dr. Zoe Colombe from University of Liverpool and Dr. Janina Chavez Ugalde from University of London. Together, they will have the common presentation, Unpacking Ultra-Processed Foods, Conception Trends, Potential Health Risks and Healthcare Costs. And another speaker, Professor Emma Boyland from University of Liverpool with ultra process food marketing and policy approach. So first, I would like to hand over to the couple of Dr. Zoe and Dr. Janina, please hand over to you.

Speaker 4
07:47 – 07:55
Thank you. I tried to put that as full screen full screen. Yeah.

Speaker 1
07:56 – 07:56
Yes,

Speaker 4
07:56 – 07:57
seems to work now.

Speaker 1
07:58 – 07:58
Yes.

Speaker 4
07:59 – 08:23
Now, perfect. Thank you for the introductions and thank you for having us today. So we’ll do a presentation on UPF as a whole. And also we will present some of our work. So I’m Zoe, I’m a researcher in epidemiology and public health at the University of Liverpool. And I will do this presentation as a duet with my colleagues.

Speaker 5
08:24 – 08:45
Fantastic. Thank you for having us. So I’m I’m Jenna Ena from City, St. George’s University. And previously, I was a postdoctoral researcher at University of Cambridge where I started working with Zoe. So we hope you enjoy this duet presentation. All right, so UPF who like why why are we

Speaker 3
08:45 – 08:45
here?

Speaker 5
08:45 – 09:20
So ultra processed products, as we would like to call them rather Southern foods have been very much in the media. And there is a difference between processed and ultra processed foods, because processing has been key for civilization. And thanks to processing, we have been able to be here and be alive through millennia. And not all industrial products are UPFs. But there is debates around the definition, of course, depending who you ask, a tomato is a fruit or a

Speaker 3
09:20 – 09:20
vegetable,

Speaker 5
09:21 – 10:47
right? And UPF is not the exception. So next slide, please. There are several classification systems to include the level of processing. However, the NOVA classification system has been the one that has been mostly used in research and the WHO recognizes it as a sound classification system to track diets and dietary transitions basically across the globe. So in a nutshell, ultra-processed foods are formulations of ingredients derived from foods with additives, including colorings, flavorings, sweeteners, and emulsifiers. Next slide, please. So something that a lot of people forget about the classification system and the NOVA classification system per se is that we often forget to talk about the purpose of ultra processed food. And this is that it’s not only that there are more nutrients or less nutrients, it’s a lot about the purpose of the processing. And they tend to be highly palatable, dare I say addictive for some people. They’re highly convenient, they’re ready to consume, they’re shelf stable. Most of them are very affordable and usually cheaper than their non ultra processed food counterparts. They’re very highly marketed and I’m sure Emma is going to touch on this.

Speaker 3
10:47 – 10:48
And yeah,

Speaker 5
10:48 – 14:16
some of them have an excess of nutrients of fat, sugar, salt, energy, density, and some of those also have a deficiency of nutrients, particularly lack of fibre. And there have been a lot of associations in research saying that high consumption in ultra processed food has a very poor dietary quality overall. Next slide, please. So we did a research recently, we published it a couple of weeks ago, about what is the overlap of these foods that are high in fat, sugar and salt, but also ultra processed food. So based on weight, 60% of ultra processed foods are left out if we consider only nutrients, so high in fat, sugar and salt. And if we consider energy, 40% of ultra processed foods are left out. Our current policies in the UK are based on being high in fat, sugar and salt. So this kind of flags a wave of we have to pay attention about something about the level of processing. And usually something that’s not accounted for in high levels of fat, sugar and salt are sweeteners that are non-caloric sweeteners, some emulsifiers, flavors and colors. Next slide please. And of course the classification system, the NOVA classification system has four groups. Group one, next slide please, are foods basically how you find them in nature, removing the parts that you don’t want to eat or that might be harmful or minimally processed that you can squeeze them, crush them, grind them, but it’s basically how you find them in nature. So an example is corn, corn on the cob, just as it is. Then we have group two of the Nova classification system and these foods are extracted from group one, from minimally processed and are called culinary processed ingredients. So to follow on the example of corn, group two would be corn oil. And then we have group three, processed foods, that are made by combining group one and two, minimally processed and culinary ingredients, but also you can process them to a point to be canned, bottled, even in the case of breads and cheeses using non-alcoholic fermentation. And even processing can enhance some of the nutrients in those foods. Now group one, two, and three have not been linked to ill health. But then we have ultra processed food, which is group four of the Nova classification system. And to follow through from the corn example, we have Doritos, a corn based snack that are usually have a lot of ingredients, such as flavor enhancers, colors, emulsifiers, sweeteners, thickeners, gelling agents, things that you wouldn’t have in your kitchen cupboard to cook at home, and Zoe will give a good example of this as well. So if we go to the next slide please, here Zoe will follow on to give you some examples of processed things that we can find in a supermarket that there’s quite a distinction between those.

Speaker 4
14:18 – 16:36
Thank you, so you have seen that the Nova classification seems pretty simple with four groups and it seems pretty easy to classify products. But the thing you need to know is that Nova classification was designed for purchase data. So it’s pretty easy when you are going to see this product, it’s not pretty easy, but looking at the ingredients, it will be pretty easy to classify that according to the Nova classification. So for example, here you have example that in the same supermarket for processed food, you can have ultra processed food or process it. It’s pretty easy. But one thing we want to underline today is that most of the studies we are using to see the impact on health are consumption data. And sometimes we don’t have the level of detail we add in supermarkets for the consumption data. So for example, in some consumption data you will have a line which said “strawberry ice lolly” without any more information than that. And then you have multiple options there. It can be something branded or it can be something homemade. And in this case, you will not classify the same. It’s the same title of product, but it will not be the same classification in Nova. Just to say that sometimes when you are working with consumption data, you will need to make some decision or assumption based on the line you are. And because we don’t have any standardized way of doing the classification, we only have the guideline from Montero and his team and some decisions that have been published in other papers. And it will depend on the information you have in your data sets. But it will also then depend on your interpretation of Montero paper and your culture. For example, as a French, I will often classify bread as processed foods, not ultra-processed food, which is not maybe the case in some countries like the UK. So depending on your knowledge and your culture and everything, you may classify things differently. Just to underline that the Nova system may have low consistency between people or can be interpreted differently, but also to underline that some work has been done and sometimes for the consumer, the concept of UPF is also a bit vague. So we might have a bit of

Speaker 1
16:36 – 16:36
work

Speaker 4
16:36 – 17:57
to do around this. When we first met with Janaina, we all are working on NOVA classification on NDNS without knowing each other. And so we decided, like, let’s compare the classification we made. And we arrived at an 87% agreement without ever talking about the classification, which means that in NDNS, depending on the conception you have of Nova classification and UPF, you still have a good agreement. We discussed and we arrived around a 100% agreement and we published our decision book and code online. You can find it on GitHub. And now we are working with a larger team of people coding Nova for the new years of NDNS. And we are working with the NDNS team, Polypage, Burden and David to provide a Nova code for the New Year’s or NDMS with a decision book. Hopefully it will be out soon. Just to give an example, this number are not representative of the diet in UK. It was just a simple sample of NDMS to show you something. Remember when I was talking about strawberry ice lolly, you can do it as three or as four, depending on how you are interpreting the line. So I did that for every food. Each time I was not sure about something, I was doing the low classification,

Speaker 3
17:57 – 17:58
the lowest

Speaker 4
17:58 – 18:26
group of NOVA classification possible, or the highest. And you can see that it’s not changing that much to conception of process or ultra-processed foods. But I did the same thing in another data sets that don’t have the same level of information as in DNS. For example, we don’t have a lot of brand information. So it will almost always be ice lolly, not Mr. Freeze’s eye lolly or made eye lolly. And then here you can see that the low or high

Speaker 3
18:26 – 18:28
classification is very

Speaker 1
18:28 – 18:29
different. And we

Speaker 4
18:29 – 20:22
can estimate, depending on the way you are coding the NOVA, you can double the UPF consumption. Just to underline that, it might be a limitation. And we may question the comparability of study. But you need to remember still that all the signals we get from the NOVA classification on health and everything we will discuss now, still are pretty consistent, independently of the study design, the definition and the people that code NOVA classification. So it seems that it’s still really, really, really, sorry. And also to underline that I show you some differences in the French West Indies, and still the socio economic association remained the same, even if we were using the high or the low classification. So still, it seems pretty boring. And maybe we should still improve better the way we are collecting data or recording the data and conceptions to take into account this UPF concept that is emerging and well studied now. So for example, we might include systematically the brands when we are recording the food and we can also imagine a data composition with more information that we have right now. And I want just to underline that even if we are today advocating for a change or a next change around the Nova classification, the way we are using the classification and we are open to debate, we are not involved in the Nova Nordisk Foundation project of doing a new classification, Nova, especially because the leader of this new Nova classification is well funded by the food industry.

Speaker 5
20:24 – 25:30
And we’re very happy of the disclosure at the beginning saying that Nova Nordisk supports EASO and I think, well, resources are limited, especially in our field of research, but it is important to be very clear on the separation of receiving funding and how much that influences the findings. And yes, the Nova classification system has its faults, but again we need to remember that it’s more about the specific thing that affects health, it’s about dietary patterns and this follows through what I’m going to present next. Thank Thank you, Zoe. So why are we concerned about ultra-processed food? So research, the latest review that was published last year about the association between a high UPF consumption and health shows that there is convincing evidence about having an effect over type two diabetes, adverse sleep and anxiety, mental health outcomes, an increase in the risk of and some cardiovascular diseases. Now, the thing is that a high consumption in ultra processed food is like a balloon. If you consume a lot of ultra processed food, you will be leaving out of your diet, minimally processed foods that actually hold the magic of the nutrients in them. And there are some mechanisms suggested of how this works. Next slide, please. So is it a single nutrient or is it the dietary pattern Research published by Samuel Dickin a couple of years ago, they controlled for nutrients and they still found that the associations of high ultra processed food consumption and ill health were held even after adjusting for nutrients. So I think this signals that there’s something about the nutritional quality, but also the level of processing they’re having an interplay and that the processing is correlated and complementary to how an unhealthy dietary pattern impacts health. Some of the mechanisms that have been suggested, it could be a lower nutrient quality, the previous one? Lower nutrient quality, changes in the food matrix, which was what I was saying, it’s different to have corn on the cob and having the fibre, the fat, all the nutrients that are held within that, than crushing it, separating the oil, the fat, the fiber, and then reconstituting it to make a corn-based snack. We still don’t understand exactly how all the nutrients and active compounds react in our bodies, so that could be one of the mechanisms. Or some contaminants from the processing, and also in like the biggest picture is the displacement of minimally processed foods. Next slide, please. And some of the ingredients that are there’s a lot of research about this and the impact on gut microbiota is some emulsifiers, particularly carboxymethylcellulose, polysorbate 80 and carrageenan’s, some nitrates in some products, food additives, some of the the e-numbers that we find commonly in some of the labels of purchased food. So keep tuned with that because there’s some more research emerging. And on the field of to try and see what’s the cause, the direct cause, I’m sure you’re all aware of Kevin Hall and his team in the US and the only published randomized control trial to date. It’s his and said, is a randomized controlled trial with two groups, one group eating a high diet in ultra processed food and another one low ultra processed food. And after two weeks, they discovered that the ones eating a lot of ultra processed food were consuming 500 calories more per day and were gaining weight versus the other group that was consuming 500 calories less and losing weight and their cardiometabolic markers were much better. Now there’s a follow-up trial. These are very preliminary findings trying to uncover what’s the causal pathway. And NIH removed their funding with the Trump administration and pending to see what happens with that. But it seems that most of the ultra-processed food effects on intake are due to the energy density of the food and a little bit about hyper-palatability. So this mix of fat and sugar, salt and fat. So again, let’s see what happens with this trial but there are more randomized control trials on the pipeline. Next slide, please. So yeah, Zoe. – Yes, thank you.

Speaker 4
25:30 – 25:49
So we have seen that UPF seems linked to health and this is why some people are advocating now for public health action toward this UPF intakes. The reason beyond that is partially because as Janina just underlined, it’s linked to health. And

Speaker 3
25:49 – 25:50
we know,

Speaker 4
25:50 – 25:57
I will not teach you anything today, but the obesity and cardiovascular disease burden are

Speaker 3
25:57 – 25:58
increasing,

Speaker 4
25:58 – 29:12
especially in Europe, and it’s not going to stop near. So we need to find solution to curb this increasing burden. And so UPF seems to be a good possible target if it’s really linked to health, as it seems to be, especially because the UPF consumption is very high in some countries. So here, for example, on the right is the consumption as part of the energy in the countries, and on the left is the percentage of purchase that are UPF. And it’s interesting to see that the UK is ahead of the other countries here, And this is a country where we don’t have any policy or recommendation looking at UPF. Just to underline something, it’s from our work with Janaina on UK, just to underline here, that the consumption is already super high in UK in adults, but it’s even worse in adolescents. And it tends to be even worse in the kids now, which means that the burden of UPF intake will increase in the future if we are not doing anything, leading to the think that maybe it will contribute to increase in obesity and cardiovascular disease in the coming years. Also to underline that it’s something we are observing in every country in Europe, but it seems that UPF intakes are not the same in every group of the population. We can see for example in adults, a higher consumption in males. The larger determinants in age, with a decrease in consumption with age, not meaning that when you are ageing, you are decreasing, just meaning that we seem to have a higher consumption in younger and that will persist over time. But also we have higher consumption in low socioeconomic groups. So maybe some policy can specifically target these groups. With my team, we are specialists in doing estimation of the policy at the national level, and try to estimate the policy impacts. And it was interesting to see that a couple of years ago, we were not able to look at the direct effects of UPF because we didn’t have consistent data. So we used at the time change in micronutrients or nutrients within the UPF. And we still see that it was if we were reducing the UPF, then reducing the micronutrients inside, then we will reduce cardiovascular disease mortality. And we are now working with different country to see the direct effects using the umbrella review Janaina just talked about. And it seems that reducing UPF intake will prevent or postpone some premature death worldwide. So it should be maybe something we should look in the future. Also, interestingly, in the UK, if we are reducing the UPF consumption, we can save up to £22 billion a year, just looking at cardiovascular disease, diabetes

Speaker 3
29:12 – 29:13
and common

Speaker 4
29:13 – 29:16
mental disorder, which means that we may have a bit of

Speaker 3
29:16 – 29:17
money

Speaker 4
29:17 – 29:29
to put some intervention or policy in place, but we’ll save money elsewhere. Just to underline that some countries already have implemented

Speaker 1
29:29 – 29:29
UPF

Speaker 4
29:29 – 29:55
as part of a national recommendation. In Europe, it’s the case for France and Belgium, saying that we should limit the UPF consumption. And just to underline a nice example from Uruguay, where they are putting in their recommendation now the different NOVA classification categories, passages, to quickly and easily explain to people the like what is non-process

Speaker 3
29:55 – 29:56
and what is virtual

Speaker 4
29:56 – 29:59
process. So we found this example quite interesting.

Speaker 5
30:02 – 30:37
Yeah and just to wrap up there are different levels that we can act on on the product and the branding nutrient reformulation but we can also have product deformulation removing those emulsifiers that are harmful incentivizing better healthier portfolios within brands acting on the consumer and Emma will develop a lot more on the marketing side of things but also remember that poverty, cost of living crisis is here so let’s not stigmatize but the most important thing how can we up the good minimally processed food and reduce the consumption of the ultra processed

Speaker 3
30:37 – 30:38
food

Speaker 5
30:38 – 30:54
and acting on the environment you know companies we can make them pay for the external costs of the foods that we’re consuming and subsidizing acting on trade and and agriculture. So I will leave it there, but I’m sure that Emma will leave us with more food for thought. Thank you very much.

Speaker 1
30:57 – 31:11
Thank you, Zoe. Thank you, Janaina. And my pleasure now to introduce Professor Emma Boylan from University of Liverpool, UPF Marketing and Policy Approach topic. Emma, hand over to you.

Speaker 6
31:12 – 52:04
Thanks very much, Emil. So yes, I’m going to be talking about UPF food marketing and what we might do in terms of policy directions. And as mentioned, I’m a professor of food marketing and child’s health from the University of Liverpool. And please feel free to reach out by email or Blue Sky if you want to get in touch afterwards. So as has already been outlined in much more detail by the previous two excellent speakers, clearly there’s a public health concern here to do with obesity and also the poor health outcomes and premature mortality that are risk, you know, that poor diet is a risk for. And we’re now moving from that to thinking about what are one of the main drivers of consumption of UPFs, and that is the food marketing of UPFs. I’ve put the definition of marketing at the bottom there from the World Health Organization. And I think that’s an important one because it’s quite broad. It speaks to the fact that marketing is any kind of act that seeks to promote a product or service or its related brand and anything that increases the appeal and recognition or likely consumption of any of those products. So it’s open to a number of the different actions of the food industry that push us towards greater consumption of UPFs. And if we think about it in a really kind of big picture level, something that the food industry sometimes called as big food. One thing that they do is to establish these really global networks. So we know that the food system is dominated by these huge transnational corporations that produce economies of scale in terms of production and distribution of their products. And we can see at the bottom here that for each of the brands that we’re familiar with, they have a number of sub-brands that sit below them. So you’ll be familiar with some of them, but they operate across this full range of product portfolios. And each of those often are well-placed internationally and hold big market shares in many countries worldwide. And that brings with it the ability and the funding to kind of flood the market with all of these products as well. So particularly through proliferation of things like supermarkets and local convenience stores, they’re able to get to the point where all of these products are really readily available to consumers in most countries at all different income levels and are often available sort of 24/7. And within that then, because they are competing with so many other processed foods in the same food environment, and because they rely on this kind of business model of generating increasing profit, even in markets where there’s such competition for what we’re going to buy. They have to scale up their marketing then to differentiate themselves from other products and to ensure that they stay top of mind for consumers and that that drives sales of their products. And one thing that we’ve seen particular investment in recently, of course, is digital marketing. That’s really the growth and innovation area in marketing and both in terms of actual spend and projected spend in the coming years. It’s all predicted to just keep rising year on year in terms of how much is spent promoting these sorts of products to us online. And it’s really digital marketing that is that area of innovation. And one thing that’s particularly concerning when we’re thinking about what was said in the previous talk about how the diets of children and adolescents are made up of a lot of UPF products, is the fact that those children and adolescents spend a lot of time online, as we all do, but the marketing that they receive that’s delivered to them is done so in a targeted and personalized way using data that has been obtained from those children as they spend time online. So they have a digital footprint that identifies them in terms of their kind of demographic profile and also their behavior and the context as well. So the types of online content that they tend to engage in, the sorts of products that they have clicked on and purchased before. And all that information is held by the kind of big players in the digital marketing ecosystem and used to ensure that marketing that’s delivered to them is done so in ways that is marketing that they’re most likely to engage with at times when they’re most vulnerable as well. And that’s why it’s particularly effective at changing consumer behavior. The WHO framework also sets out a couple of really important things. So the impact of marketing for UPFs on behavior, so anything from choice, purchase, consumption, the impact is a function of both exposure and power. So by exposure, we mean the reach and frequency of marketing messages. So essentially how much marketing you see, what it’s for and how many people are exposed. And then the power is essentially all the kind of creative content of the message, the themes, the appeals, the techniques and all of those things that are used to make the marketing grab your attention, hold your attention and persuade you to change your behavior as a result. Obviously, ideally in their case, in the direction of purchasing and consuming the UPF products. Now, you’ll all be familiar with where we see food marketing and where young people also see food marketing. This is not all of them, but it’s just to give a sense of the scale, you know, basically in any kind of setting or in certainly any use of media, there will be a commercial component to that. So we see advertising outdoors, we see it on every screen that we use, and we see it in settings like sports arenas and sports broadcasting as well. And with digital marketing, as I mentioned, it’s a very innovative area just in the, say, last 20 years. In terms of food marketing, It’s gone from sort of public health concerns around marketing on specific websites, so food brand websites, and things like simple online games that feature brands and food cues, all the way through to what we have now are, you know, sort of immersive environments that can be brand driven. So, you know, a young person could spend essentially an unlimited amount of time in something like a UPF branded world, as a Pringles world, as McDonald’s world. And you can actually be immersed in these environments using VR technology, as well as things like, you know, food delivery apps that obviously promote and market the least healthy options as a priority. And to accompany that, you can find a paper or a report that will document the types of food marketing found in each of those spaces. So whichever media or setting you’re interested in, there will be somewhere a report or a paper talking about monitoring activity that has looked at that particular media or setting and quantified the volume and nature of the food marketing that is going on there. Most of the data we have until now doesn’t necessarily use a UPF definition. So often because these studies are done at scale, they’ll use a sort of very simple way of allocating foods into maybe core and non-core or sort of broadly healthy and unhealthy using research of judgment that doesn’t require us to look up every single food product and classify it with any nutrient-based system or with an over system. But there are some studies increasingly so now given the prominence of the UPF concept, starting to look at what it means specifically in terms of how many foods at UPF. So I’ve got an example study there from a few years ago looking at TV advertising in Brazil. And that study found that almost 91% of the advertised foods were ultra processed according to the NOVA classification. So if you think about the contrast between that and what young people in particular are told about, good dietary health, what we should be eating a lot of, what they see in terms of the advertised diet is very, very different. And whichever country you look at, generally you find that the majority, somewhere between 50% and almost 95, 100% of advertised foods would fall into an unhealthy bracket by whichever definition you use. And it’s dominated by these five, so-called big five product categories of sugary, sweetened beverages, increasingly including energy drinks, fast food, breakfast cereals, snacks and confectionery. And it was mentioned in the previous talk about the issue of inequalities and that’s absolutely huge here as well because we do see that there are differences in advertising exposure by a number of socioeconomic variables. In terms of sex or gender, we often see that males receive more food marketing than females do. And it’s often a male dominated sort of marketing environment. We also see differences by socioeconomic status such that people living in more deprived communities will often have greater unhealthy food marketing exposure. And that’s sometimes to do with the physical environment outside. So there’ll often be more sort of streetscape food advertising in more deprived communities and also greater screen use in those households as well. And there’s also some evidence of differences by ethnicity. So as a package, individuals of non-white ethnicity who are in a deprived household or community will often receive more marketing. And given the sort of breakdown of the types of foods that are being marketed, you can see how this can contribute to diet-related inequalities. There’s not much evidence that there’s a difference in kind of behavioral impact of that marketing. The inequalities are largely driven by exposure. When we look at the persuasive power of marketing and the techniques used to influence us to purchase more and consume more UPF foods, you see there are lots of techniques I’m sure most of you are familiar with in terms of celebrity endorsement and those sorts of techniques. Also a lot of kind of monetary based offers and competitions, things like the ability to download content is really popular with young people. And of course, anything in terms of price or vouchers or those sorts of things tend to get a lot of engagement from young people in particular. And increasingly it’s being recognized that marketing is not seeking to engage with us or persuade us in a sort of conscious, rational manner. It’s not trying to convince us in a sensible decision-making process that this product is better. It’s trying to capture us at an emotional level. So there will often be emotion-based claims in there that essentially suggest to young people and all of us that this product will be associated with greater fun, positive feelings, and essentially all emotionally salient vibes depending on who they’re targeting. So popularity might be super important to a teenager, happy family and happy engaged children might be really important to a parent. And so, you know, it will target accordingly. And that’s often the case, no matter which country’s advertising we’re looking at. And that’s really super important because we have found in this recent meta-analysis that when you look at sort of neurological responding to food marketing, both children and adults, when people are exposed to food marketing in an fMRI scanner, a number of consistent areas of the brain is activated including things like visual perception processing, which you would expect because the adverts tend to be brightly colored and so on, but also areas of the brain associated with taste and emotional processing as well. So it’s not possible to think in sort of intervention terms that we can train people to get better at resisting food marketing. Our emotions are very difficult to control and difficult to train in that way. And so it really lends itself to a situation where there’s an onus on governments to act and to restrict exposure, rather than expecting this to be something we can rationally decision-make our way out of. In terms of the impact of food marketing, there’s this model that I was part of the team to produce this a few years ago, based largely on evidence from TV advertising, essentially to suggest that exposure to food marketing has a number of outcomes, some more immediate and some longer term, but essentially that food marketing will create more positive feelings about brands as well as greater recognition. And that will follow through to when somebody is in a position to either purchase a product or ask another person such as a parent to purchase a product. And then if something’s purchased, it’s highly likely to be consumed. And there’s been increasing data recently to show that when we are exposed to food marketing, not only do we consume foods, but we also don’t compensate for that later such that we’re over consuming energy as a result of marketing driven intake. And this was a systematic review of meta-analysis that I undertook a few years ago that underpins new WHO global guidelines on food marketing restrictions. So this is the most up-to-date evidence synthesis to show that food marketing affects children’s intake as well as choice and preference. And given what we know about the types of foods that are marketed, it’s influencing preference and choice and intake towards unhealthy UPF products. So based on all of this evidence of the impact of food marketing and the sort of public health concern, as I mentioned, the World Health Organization updated their global guidelines in 2023 encourage countries to act and to enact restrictions to reduce particularly children’s exposure to food marketing. And within that, based on the evidence that was available at the time, it specifically recommends that a government-led nutrient profile model is used to classify which foods should and shouldn’t be restricted from marketing. And that could either, you know, there are many models that have been developed by governments or the WHO regional models that have been adopted by governments They suggested it should be mandatory policies and that’s building on a lot of evidence from previous reviews and studies that have shown that self-regulatory action by the food industry is ineffective and that’s something we found in the systematic review underpinning this particular, these new global guidelines as well. and that’s the fact that if you implement mandatory policies, you tend to get favourable outcomes, so it tends to be effective in achieving reductions in things like exposure and power and have good outcomes for public health. When voluntary measures from the food industry are used, the effects of these studies tend to cluster around the not achieving public health goals or not finding a significant effect in a positive direction. So mandatory regulations are clearly the way to go to bring about effective public health change. The reason it said a nutrient profile model on the restrictions and the reason that’s where we’re at in most countries is because the evidence to date suggests that the most effective bans either have a total ban on commercial activity or have product-based restrictions that use a nutrient model. The industry initiatives that don’t tend to work are the ones where they have specified their own classification system. So, for example, the sugary sweetened beverage companies come together and say we won’t advertise any products that have more sugar than this per 100 grams, but they tend to set very lenient thresholds for that amount of sugar that just happen to allow them to continue advertising all their main products. So nutrient profile models are evidence-based and tend to be more stringent than the sorts of things that industry come up with. But as we already heard, we’ve already seen some of this data that using a nutrient profile model that determines what’s HFSS or not doesn’t necessarily capture all the food products that we’re concerned about in terms of what we heard earlier about the dietary impacts of UPFs. So just to highlight that data again, you know, a majority of UPF products are also HFSS, but there’s a non-negligent proportion of foods that are not captured by an HFSS definition, and that we also need to be considering in terms of dietary quality and population health outcomes. One of the concerns that I’ve heard in a lot of debates around food marketing policies and in any policy is this sort of like this binary decision making between using something like HFSS and using UPF as a definition. But I think it’s disingenuous to suggest that it needs to be one or the other. We’ve seen success in tobacco regulation that’s come from not taking one line and sort of sticking with it and refusing to budge but actually integrating a number of different over time and building on the original concepts to make, you know, increase the strength of regulations and then keep adding to and adding to until we get to a point where we’ve achieved the population health goal that we were looking for. So I think here there’s a lesson to be learned that we should be considering how can we add additional value and make an incremental improvement. And one way we can do that is to think about can we adjust nutrient profile models to incorporate elements of UPF definitions. You may have seen that the WHO regional office for Europe have updated their nutrient profile model recently to address kind of difference or greater knowledge around some of the nutrients that are involved in that. So it is adaptable, it’s not a fixed document, it can be adapted and developed over time as new evidence becomes available. And that might be the case that it could be adapted to incorporate UPF in the future. And studies have looked into how we can close that gap between what’s captured by an HFSS and a UPF definition. And by adding things in, so having a starting point of HFSS and then adding things like colourings and flavourings, adding things like non-nutritive sweeteners and adding things like additives into that definition really closes that gap to the extent that all of those products would be captured by a kind of updated definition if we considered it as HFSS plus additives that really would cover essentially all of the food products that we’re concerned about the dietary implications on. But just another note to say that in terms of food marketing policies there are a lot of question marks remaining around brand marketing because whether you use a nutrient approach or a processing approach that assumes that you’re talking about a particular product when increasingly we’re seeing brands marketing just elements of the brand without showing a product that would be able to classify in that way. And also there’s a need to consider the role for reformulation. It was mentioned in the previous talk with a lot more sort of nuance, so I won’t go into it again, but just to say that a lot of governments favour making allowances for reformulation in policies because it’s a good incentive for industry to keep looking at how they can improve and use some of of that profit that they’ve gained from

Speaker 3
52:04 – 52:05
the dominance

Speaker 6
52:05 – 52:18
of the global food system to actually improve dietary outcomes without the need for behavior change. And I will stop there. But thank you very much for your time and attention. And I’m happy to take questions.

Speaker 1
52:20 – 52:57
Thank you very much, Emma. And I would like to go for questions. We have them. So the The first questions for both talks, actually, are there examples of targeted interventions, for example, subsidies for minimally processed food, as an example, that have successfully launched so far? Zoe, Janaina, Emma, this is a question for all of you. OK, Janaina, you can start.

Speaker 5
52:57 – 54:09
So there is a program in London, in some boroughs of London, trying to subsidize minimally processed food for people that are struggling financially. And it’s just showing that there’s a lot of appetite for good minimally processed food and that people eat them and that people want them and that people like them. Although it’s just a pilot and it’s heavily subsidized, It’s an organization very localized, but it is targeted to target, you know, the people that are struggling the most that are trying to eat a healthier diet. However, at governmental national level, I have not seen an example that it’s targeting that aspect, but I think Chile’s approach to having an array of policies that are incremental and then start with octagon labeling and then those foods cannot be introduced into institutions like schools, hospitals and the military. So it’s having an array, a package of policies and interventions that try to increase minimally processed food consumption.

Speaker 1
54:12 – 54:16
Do you want to add something Zoe or Emma regarding this?

Speaker 6
54:19 – 54:26
It’s not really something that’s done in food marketing because the food marketers are all promoting UPS. So I’ll leave it

Speaker 3
54:26 – 54:27
to

Speaker 6
54:27 – 54:29
Zoe to supplement that point if she has anything.

Speaker 4
54:30 – 54:39
– I don’t have any clear example, but we did in our evaluation of potential policies. So it’s not a real policy, but it

Speaker 3
54:39 – 54:40
seems that if we

Speaker 4
54:40 – 54:58
are increasing the email process, it could be a good idea to reduce CBD, especially for fiber and protein and veg. So it’s not per se targeting minimally processed, but for example, the fact that you cannot have VAT for food and veg can be like a good

Speaker 3
54:59 – 55:00
subsidy,

Speaker 4
55:01 – 55:01
we’ll say.

Speaker 1
55:02 – 55:12
– So thank you. Here’s another interesting question. Are there any studies about the impact of processing organic and non-organic foods?

Speaker 4
55:18 – 55:38
I don’t think so, not published at least. What I know though, is that in the NutriNet Sante study in France, they are doing a lot of work on your behalf, but also a lot of work on organic food. So maybe they will publishing something in the coming years, but not I’m aware of.

Speaker 1
55:40 – 55:42
– Okay, another, yeah, Janaina.

Speaker 5
55:42 – 56:05
– May I just, like defining what’s organic and non-organic, it’s also highly debatable, You can have a parcel of apples here that are labeled as organic. And the one next to it is spraying with fertilizers or whatever. And the winds can take it there. So it’s, yeah, the framework’s on what’s organic, what isn’t. But I think even if an apple is not organic,

Speaker 3
56:07 – 56:07
it might

Speaker 5
56:07 – 56:16
be better than having it as an ultra processed food pureed apple with sweeteners and potential emulsifiers. That’s all I’m going to say.

Speaker 1
56:17 – 56:38
Okay, the question for Emma, I guess, or you can also join. What policy measures do you believe would be most effective in supporting low income or socio-economically disadvantaged populations who, according to data, consume higher amounts of UPFs?

Speaker 6
56:40 – 58:11
Thank you. I mean in terms of food marketing policies, the restrictions that I was mentioning there, the mandatory government restrictions have tended to be effective in every country regardless of income level. The issue becomes moving towards what we’ve just been talking about in terms of promoting the minimally processed foods. So I think there’s obviously the issues around accessibility and affordability. In terms of food marketing, promotional activities that promote the minimally processed foods do not tend to be as effective as promotions that promote the ultra-processed foods. So we can advertise all of these apples and broccoli if we can find the commercial imperative to do so, but you know intrinsically the hyper-palatability of UPS wins out and you would need to spend an awful lot more money to promote the healthier stuff to achieve the same sort of price impact. So from my perspective it’s just absolutely crucial as part of a package of measures and you know I’m mentioning there about you know what’s so effective about the impressive budget is their multifaceted approach and a real systems approach to this that’s integrated. I think that what we need to do in terms of marketing is to take away all the UPF marketing rather than invest in increasing

Speaker 3
58:11 – 58:12
the minimally

Speaker 6
58:12 – 58:24
processed food marketing and then there are other interventions in terms of you know the retail environment and affordability that need to also come into play in order to make those foods affordable for families.

Speaker 1
58:26 – 58:28
Thank you. Yeah, Janaina.

Speaker 6
58:29 – 58:30
I totally

Speaker 5
58:30 – 59:06
second what Emmett said and in a nutshell, remove barriers to access. Because if you cannot access it, you cannot eat it. You cannot taste it. You cannot build the preference for it. And the intergenerational legacy of food preferences and taste and etc. that happens from the womb. it just trickles down generations. So reducing the barriers to access, making it the healthier choice, the easiest choice, including all the marketing promotion, banning of that, and just making it easier to have those foods.

Speaker 1
59:08 – 59:23
– Thank you. Question to Zoe and Janayina, I guess. Do you have any specific tips on how researchers can more accurately categorize UPF from longitudinal dietary intake data based on your experience?

Speaker 4
59:27 – 59:52
First, I will say it’s a good idea to do that not alone. Like having different coders, a bit like we are doing a lot in research when we are doing systematic review or this type of thing. We always have two people coding and then we are comparing it. And I think it’s important there to have consistency. Yeah, I think it would be my tip.

Speaker 5
59:53 – 01:00:16
– And because classifying foods is not perfect, make it open source, you know, list your assumptions, publish it online, say what the assumptions were and just build better. Don’t start from zero, see what Zoe and I have done, see what other people are done, share that and just make it open so that people can see what if the strawberry lolly was classified as UPF or not UPF and then let’s have a discussion about it.

Speaker 1
01:00:19 – 01:00:39
Another question to Zoe. Zoe made the point about the fact that despite varying interpretations of the NOVA classification across studies, they still provide sufficient or reasonable evidence to show that UPFs are harmful. Could you elaborate on that?

Speaker 4
01:00:40 – 01:01:59
Yeah, it was just to make a point that it seems when I was doing the presentation, I like, Nova classification cannot be relied on. And I just wanted to make the point that even if people are classifying differently around the world, every study or a lot of study looking at UPF, high UPF intakes impact on health are still finding consistent effects from high consumption of UPF on health. Independently, if it was longitudinal, cross sectional, we are seeing the same determinant of UPF consumption also in different study. So my point was just to say that even if the people are not classifying exactly the same way NOVA, it seems to be pretty consistent that UPF have an impact on health and the socioeconomic status associated with UPF intake are the same. And just throughout that, it doesn’t mean that it’s UPF per se, it can also be what inside and every mechanisms Janaina just talked about in the presentation, just that. Overall, it seems to have an effect. We don’t know which one, but it seems to have an effect. – Okay,

Speaker 1
01:01:59 – 01:02:34
okay. And the last question from Ellen, she also thanks for the great talks. Is there any research on how usable the general public find UPF classification when navigating the current food environment in order to find the healthy food. As a clinician, it is this classification that seems to help people the most practically compared to low calorie, low fat, low carb, for example, in navigating this environment.

Speaker 5
01:02:36 – 01:03:32
– I agree. I really think that the UPF classification should be left to the policy level and to reducing its prevalence and ubiquity in our food environments. And again, reduce the barriers to having access to the good foods. I think it can be confusing. and also part of the strategy of a few organizations is to make this classification even more confusing and instigate doubt about the evidence of the link between ultra-processed food consumption. But I come back to the displacement of foods, displacing the minimally processed food because the other ones are more convenient and usually cheaper, just tips the balance into having a preference for the other foods. So yeah, in terms of classification, maybe having the octagons and just having an octagon that says this is ultra processed and you know, letting people decide if they have a choice.

Speaker 1
01:03:36 – 01:04:00
Thank you. Thank you, Janaina. Thank you, Emma, Zoe for brilliant talks. Thank you for your time. Thank you all members, ECN members. And please don’t forget that we have usually monthly webinars. And also don’t forget to give feedback. Lisa, do you want to add something before?

Speaker 3
01:04:01 – 01:04:16
I’ll just say thank you, everyone, for coming. And hope to see you at other ECN events. And thanks to the speakers for their nice talks. Thank you very much. Thank you, and see you on our next webinars. Thank you.