One of the most common questions I get in nutrition is about UPFs.
My thoughts in general are that categorising foods as the Nova classification does is scientifically meaningless.
I am sort of stunned that “how processed is it?” has become the dominant narrative in nutrition in the media, based on very woolly evidence which I cover in my previous post. (I think Chris might have read it too because he’s gradually softening his stance HAHAHA*).
Essentially, I don’t think the evidence shows that processing per se or number of ingredients etc tells us anything about the healthfulness or not of foods:
Sugar is not a UPF (it’s in category 2), yet we are certain enough about the negative impact on health, such that dietary guidelines everywhere recommend its consumption is limited.
Butter likewise.
Mostly, I strongly suspect that the reason diets high in UPFs are associated with ill health is because diets high in UPFs are high in energy dense foods like crisps, pastries, sausage rolls, chocolates, energy drinks, takeaways and high calorie ready meals, and low in whole fruits and veges.
However, there are some grey areas to consider and the science on some aspects of UPFs is ongoing. Amelia Tait writes concisely and clearly about some of these areas in this excellent Guardian article.
The purpose of this post is how to practically navigate these grey areas.
The grey areas:
It looks like a key reason diets high in UPFs might cause people to over-consume calories and thus gain weight is because they tend to be energy dense (they have a large number of calories per gram).
UPFs might also cause over-eating because they are more likely to be soft, and people eat soft foods quicker than than they do hard foods.
There might (but might not) be some additives in UPFs which somehow hijack appetite centres and cause us to eat more or might somehow negatively affect the gut.
How to deal with energy density
Foods which are energy dense tend to be:
high in fat.
low in water.
low in fruits, veg and legumes (FVL).
So the likelihood is that UPFs (or meals which include UPFs) that are not high in fat and/or low in water, fruit, veg and legumes do not cause overconsumption of calories.
Practical tips:
Supermarket breads and high-fibre cereals are very likely absolutely fine and probably have net health benefits- I eat these and don’t worry about it whatsoever.
If you get pizzas or ready meals, try to choose lower-fat versions and add loads of veges on them:
Extra green pepper, halved cherry tomatoes or courguette on a pizza before you pop in the oven.
Microwaveable carrots/broccoli on eg microwaveble spag bol ready meal.
Stir-fry beansprouts or a prepared stir-fry mix with eg chicken chow mein ready meal.
Veges like frozen carrots, peas and sweetcorn (which you can microwave in 3-4 mins) on processed protein like chicken nuggets or plant-based meat**.
And remember!:
If energy density is the property of foods which drives overconsumption of calories than many unprocessed foods will too***:
Crisps with 3 ingredients (potato, fat, salt).
Homemade chips.
Homemade sweet potato fries.
Homemade ice-cream.
Homemade bread with loads of butter on it.
Homemade/artesanal pizza with loads of delicious cheese.
How to deal with hardness vs softness
Some good research actually shows that whether a food is soft vs hard is a bigger factor in whether people eat more of it than the degree of processing. A proposed mechanism is that soft foods take less time to chew so people eat them quicker, which leads to increased energy intake.
UPFs might be more likely to be soft than unprocessed foods.
Practical tips:
Supermarket breads don’t seem to be any softer than artesenal or homemade so don’t worry about it.
If you do have a soft UPF, I am pretty sure (I couldn’t find a study on this) that if you eat it with something that requires more chewing it would negate any impact of the soft UPF on rate of chewing and thus food intake. (Also note how the below suggestions are healthy additions anyway), eg:
Nuts and seeds on cereals.
Gherkins, portabello mushroom, tomato on soft burger patties.
And remember!:
If softness is the property of foods which drives overconsumption of calories than many soft unprocessed foods will too:
Homemade ice-cream.
Any delicious homemade dessert, yes tiramisu I am looking at you.
Homemade bread (do you know how soft that stuff is when it comes out of the oven???).
Homemade moussaka, lasagne.
How to deal with the question of additives
People freak out about additives because they’re chemicals. Remember, everything we eat is made up of chemicals.
There are thousands of additives in food and many have been around for decades. It is true that we have tended to focus on toxicology outcomes in studies to test how “safe” they are. However, some research has been done on the roles of additives like sweeteners and emulsifiers on outcomes including appetite, glucose homeostasis and type 2 diabetes. Some studies seem to suggest a negative effect of some additives, some might have beneficial properties and some don’t do anything.
However, I am skeptical that additives are playing a significant role in driving the obesity epidemic and cardiometabolic disease.
If we look at Kevin’s Halls study, the difference in energy intake between the UPF and unprocessed diets was about 500kcal/day, and this aligns quite closely and consistently with the literature on energy density, and suggests that the overconsumption of calories and associated weight gain in Kevin’s study is predominantly explained by the energy density of the two diets. So even if (and it’s a big if) some additives/or a combination of additives have an independent effect on weight gain, the additional effect is likely to be very, very small. And smaller still when we consider the contribution of additives against very high-sugar, high-fat products we consume too much of. (Remember, Kevin’s study had controlled for these).
If we consider excellent (bearing in mind the challenges of doing this type of analysis) observational research studies which have tested >20 emulsifiers, only a few are associated with things like cardiovascular disease, and even for those few, there are enough methodological weaknesses that the jury is still out****. (It’s also worth emphasising than even in a very large dataset which is able to pick up even tiny increases in risk, the majority seemed fine).
There is also a possibility that some additives could disrupt the microbiome (bearing in mind we don’t even know what this means), and/or cause inflammation in the bowel. I don’t think any of the research to date is conclusive on this area but the proposed mechanism for some emulsifiers makes biological sense***** and good trials are being done which will be more informative. Note that this is not my area, and if you’re interested in the effect of additives on gut health I would just read everything Kevin Whelan writes (this is great) because he is very objective, a proper scientist (while other people are writing books, he’s actually testing hypotheses in controlled trials) and he’s not on instagram flogging supplements or microbiome testing to the masses. At the moment his message seems to be “wait for results of trials”.
So how do we deal with all the unknown, especially when there are lots of scare stories in the media?
My approach is based on the fact I have been in research a long-time and there are always things that look like they help or harm human health in pre-clinical/underpowered single-arm or experimental trials/epidemiology studies****** and then when they’re tested properly such as in a powered randomised trial……they don’t.
So I don’t worry until there is good evidence I need to (and if we see genuine harm I expect regulators to step in anyway). Hence, I don’t plan my diet based on exposure to additives.
Instead, I try to follow a diet for which there is evidence for health benefit, (and by doing so I reduce my exposure to additives anyway!). For example, in this epidemiological study, about 60% of total carageenans came from dairy desserts (I’m presuming mostly ice cream-type stuff) cakes, confectionary and biscuits). Well, I limit my intake of ice cream, cakes etc because they’re not nutritious foods anyway.
It’s also important to consider trade offs. Is there a possibility that a/some additive might do something we haven’t previously measured before? Yes. But there are thousands of them. So if you want to absolutely avoid any risk, you can grow your own food/make things from scratch etc. But my guess would be you’d spend an awful lot of time preparing food, and this might inadvertently affect your health because you’d miss out on stuff like having fun with your friends, eating nachos and watching the Champions League final. (And presumably you’ve got to go to work at some point. Oh and you might inadvertently develop a micronutrient deficiency).
So I think the most sensible thing to do is ensure we are getting a lot of whole, unprocessed food in our diets, and then use foods which have additives in them where these products offer us some benefit, whether nutritional (eg fortified foods), taste, price, convenience or anything else. This way, we are going with evidence about which we have a lot of certainty (having loads of fresh fruit and veg is good for us), while enjoying the benefits of processed (and yes haters there are lots), and by chance we’ll reduce our exposure to many additives anyway.
Practical tips (see image above)
Eat whatever cereal you like, ideally one that is high in fibre and doesn’t have added sugar/salt, and then add whole fruit, nuts and seeds to it.
I grate an apple, then add blueberries or strawberries and then whatever seeds or nuts I have around.
Make sure half your lunch and dinner plate is fresh veg, salad or legumes. (Don’t worry about the presence of calcium chloride or sulphites if you like the convenience of canned legumes because the benefits of the legumes far outweigh any theoretical risk of the additives which based on current data is non-existent).
If you enjoy a snack:
Fruit, nuts and dried fruit are obviously good choices.
Vege sticks and guacamole, hummus etc are good. (Again don’t worry about the additives in this dips - they have a good fat profile, are made from whole plants, and half your intake here is veges anyway).
Crackers, and a bit of cheese and something like cucumber.
Finally
Here is a list of UPFs which I eat daily or multiple times a week and don’t worry about it at all:
Mayonnaise (it’s delicious, helps me get enough calories and is a really healthy fat source).
Protein powder (just because I exercise a lot and this helps me meet protein requirements, and I usually have with a banana).
Plant-based milks, yogurt and meat (because I am a tree hugger).
Ready made sauces (I might not eat this daily but when I want to whip up a tofu, blackbean and vege stirfry sauces are quick and delicious).
Supermarket bread. (I mainly have for breakfast, often with eggs/plant based sausage and then loads of avocado, tomatoes, mushrooms, asparagus or other veges).
Sorry got lots of footnotes for this one:
*Frustratingly what I see with media/instagram doctors is they go “all in” on one single aspect of nutrition or health (keto, metformin, UPFs), and then after a while they read a bit more and then soften, then recant, then move onto the next shiny thing. Obviously by this point they’ve freaked lots of people out and very often made quite a bit of money.
**I’m a paid scientific advisor to Beyond Meat so you can ignore this bit if you think I am biased haha.
***The combinations of carbohydrate plus salt and fat plus sugar seem to be particularly enticing flavour combinations to humans - it’s not all about processing!
****There are some great comments here especially from Prof Kevin McConway who is brilliant at explaining statistics in an easy to understand way.
***** just because something makes sense biologically doesn’t mean it will pan out to be true. (also see ******)
****** leptin, ghrelin, irisin, “inflammatory” omega-6, any inflammatory marker and any disease (haha). The list is long.
Disclaimer
As a dietitian, and not a medical doctor, I am not qualified to diagnose any health conditions.
I love how you correctly recognize that mayo isn't evil. For years I've been confused why mayo has been denigrated as being "unhealthy" because it's so low in saturated fats and relatively high in health unsaturated fats. I don't like to eat too much of it because of the calories, but it's not inherently bad and it can replace so many less healthy condiments!
I smiled at CvT being uneasy about profit driven health services like Zoe. Did he publish his book as a charitable endeavour?