Do ultra-processed foods really shorten our lifespan? Maybe...some of them...

As we see yet more headlines about the latest edible bogeyman (sorry about that mental image), what did the study actually show? And what does ultra-processed even mean?

Ahh another week, another terrifying headline about ultra processed foods. Today, they increase the risk of early death, according to the Guardian who are reporting on what they call an international study of the foodstuffs.

We’ll get on to what the study found in a minute, but first, what does ultra-processed actually mean? Given the number of scary headlines about them, the term is surprisingly non-specific. Broadly, foods can be unprocessed, processed or ultra processed, according to something called the NOVA food classification, published in 2016 by a team from Brazil. The NOVA guidelines can be found here, and if you click on that link, you’ll notice that they state in the very first sentence that this classification system has nothing to do with the nutrient levels of the foods.

Processing refers to what’s done to the food. Unprocessed foods are basically raw ingredients. Foods where nothing has been added to them or done to them. Most food we eat is processed in some way – cooking, seasoning, combining foods and generally altering their state in some way. All the cooking we do at home – baking bread or cakes, making a pasta sauce from scratch, can be considered processing foods.

Ultra-processed food is the next step up. It’s often described as food that contains many ingredients (some say more than 5), and particularly ingredients rarely found in a home kitchen. And this is where it can get vague. This could mean the addition of sweeteners, emulsifiers (used to combine foods that don’t normally mix well – often oils and water), flavourings or colourings.

Lots of what we’d consider ‘bad foods’ (aka delicious foods) fall in to this category – ice-cream, shop bought cake, processed meats like ham and sausages, biscuits, fizzy drinks and some types of alcohol (in particular spirits like whisky and rum). High fat, high sugar, high calorie, low nutritional value. The problem is, lots of foods that aren’t necessarily so bad also do. Things like breakfast cereals or shop bought brown bread. These do have added ingredients that you wouldn’t use in home cooking, but some of these are actively good for us – breakfast cereals for example are often fortified with extra vitamins.

The concept of ultra-processed food has got a life of its own now, and method of preparation is now seen by some as more important than the nutritional content – something this term was never meant to convey. This means when a study talks about ultra-processed food, the types of food they’re lumping together vary from absolutely-no-doubt-about-it-bad-for-you sugary jelly sweets to billed-by-many-as-health-food plant based meat alternatives (most vegan food like this is automatically ultra-processed because of how meat substitutes are made), and lots of food in between. Which can make research hard to interpret.

And so, back to today’s headlines. Which say eating these types of foods puts us at risk of earlier death. And not a small amount, according to the Guardian, the paper found that “UPF… is implicated in as many as one in seven of all premature deaths that occur in some countries”.

Yikes.

Time to look at the actual paper. It’s a meta-analysis of observational studies from seven different countries. This means these individual studies recruited a group of people, and then watched what they chose to do (in this case, chose to eat) over a period of time, and then looked at how many of them went on to die prematurely, and when. Crucially, none of the studies could control what people ate – or anything else. And this is often an issue in these types of study. The people who chose to eat more UPFs could be different in lots of other ways from those who didn’t. It means we can’t say for certain that the UPFs caused the early deaths. As UPF consumption goes up, risk of earlier death also increases. But this could be for any number of reasons, as UPF consumption is likely to be associated with other things too, like lifestyle, exercise, income, age, or any number of other things.

A strength of the study is that they looked in a number of different countries round the world, where the pattern of who is likely to be eating UPFs might look different. For example, in some countries like Chile and Brazil, UPF consumption is linked to higher income. In other countries like the USA, it’s more strongly linked to deprivation or food insecurity (shown in this research here). If (as they were in the current study) associations between UPFs and earlier mortality are seen in both these contexts, its suggests something else is going on other than just UPFs being a marker of food insecurity or deprivation. Because if that was the case you’d expect the association to be different in the different contexts.

However, there could be other factors influencing health too. And maybe the UPFs being consumed in different countries are different. It might be that in the USA, UPFs are much more likely to be sugary sweets, while in Brazil they are bread – it’s impossible to tell from these studies.

(As a little nerdy stats aside, which seem to be becoming common in these posts – it’s really unclear how or why they picked the papers they did to include in their study. Usually studies for meta-analysis would be identified via systematic review of the literature, and a detailed description of how this was done in the methods section of the paper. That hasn’t been done here, which is a shame, because it makes you wonder why not).

Which brings us back to this ‘one in seven early deaths’ figure. This has been calculated using something called a Population Attributable Fraction. This uses statistics to estimate how many of the early deaths seen would have been prevented if no-one had consumed any UPFs, assuming that UPFs cause the deaths. Meaning this is a hypothetical number for many reasons. First as we’ve already established, these observational studies can’t prove causality – we still don’t know the deaths were related to eating or drinking UPFs. Secondly, in order to work out the number of excess deaths, we need to estimate what the mortality rate would look like in a population who’ve never eaten UPFs at all! Such a population doesn’t exist, so again this is hypothetical.

Many people, including the authors of this current study, argue that guidance about eating fewer UPFs should be issued. I’m not so sure. Guidance, certainly in the UK, is already clear that foods high in sugar and fat, and low in nutritional value are unhealthy and should only be eaten in moderation. Many UPFs are already covered by this, and those that aren’t…well…I’m not yet convinced by the evidence that they are as harmful as the ‘traditionally’ unhealthy ones. If you gave up shop bought brown bread and replaced it with home-made cake slathered in jam, I doubt you’d feel better.

Do we know that eating high fat high sugar high salt foods, particularly empty calories without nutritional value is likely to be bad for us in lots of ways? I believe yes, we’ve known this for a long time.

Does the term ultra-processed food help us understand this? I’m less convinced by this. I’m also concerned that this, particularly in countries like the UK and the USA, becomes very middle-class aspirational, and doesn’t actually help people who live with food insecurity and often have to rely on foods that would be classed as ultra-processed for price and convenience reasons. Instead of thinking about guidance, maybe we should be looking at how to make healthier diets more achievable for all.