The Science of Food

Anyone with even a cursory interest in food and nutrition must have noticed that ultra-processed foods are having a big moment in the media. Barely a day goes by why without some article like “Every Bite of Ultra-Processed Foods May Increase Risk of Early Death, Study Says” or “Ultra-processed foods are silently altering your metabolism, scientists warn” popping up in your feed. It’s all a bit intimidating. Especially so when you consider that a bread roll is an ultra-processed food. Am I killing myself and my kids by eating bread rolls? Honestly, I don’t think this is the case. I’d suggest that the media’s poor interpretation and hysterical reporting of observational studies filled with confounding factors is causing far more harm than packaged bread rolls.

I am not a ‘supporter’ of ultra-processed food (I’m going to call them UPFs from now on); high in salt, fats and sugar and low in dietary fibre these are not the food stuffs that are the foundation of a happy, healthy and long life. I don’t need science to tell me that. But this is not a cigarette company or Sackler-type moment. Ultra-processed foods are not addictive, they are sometimes useful and they are sometimes delicious; we can’t just ban all UPFs. But if we assume that food companies are more concerned with our money than our health then it is really up to us to educate ourselves on what we should eat. If you only eat potato chips you’re going to die young, but surely you can have a packet of crisps every now and again? The problem is that if all we get is reporting like I highlighted above it is very hard to answer this question.

So lets say we are hungry non-specialists trying to make sense of what we have been told about what we should eat. The first question we probably have is: what is a UPF anyway? Almost all recent studies of UPFs use a classification system published in 2009 by a Brazilian researcher who wanted to categorise the amount of processing a food has undergone. The system is called the Nova classification scheme and it has four different categories of processed food:

1. Unprocessed or minimally processed foods that may have been,
   for example, cleaned, fermented, ground, pasteurised or in general
   have had things removed but no oils, fats or sugars have been
   added. Examples: fresh or frozen vegetables, nuts, grains, pasta,
   couscous, eggs, lentils, tea and coffee
2. Processed Culinary Ingredients are products that are extracted from natural foods by pressing, grinding, crushing, pulverising and refining. Examples: oils, lard, butter, honey, molasses, starches and sugar
3. Processed foods are foods that are a combination of groups 1 and 2. Examples: Salt or vinegar preserved vegetables, tomato paste, beef jerky, bacon, sugared nuts, canned fish, fresh bread (not packaged), beer, cider and wine
4. Ultra-processed foods are made entirely from substances extracted from other foods, food constituents (hydrogenated fats, modified starches) or things entirely synthesised in a laboratory. Examples: ice cream, biscuits, cookies, confectionery, potato crisps, sweetened and flavoured yoghurts, packaged bread, prepared burgers and sausages, breakfast cereals, ‘ready-to-eat’ meals, chicken nuggets, infant formulas, cakes, cake mixtures, distilled alcoholic beverages (rum, whisky etc).

Generally, when people are talking about UPFs they are talking about the fourth category of the Nova classification scheme (there is a cheat sheet here if you want the full classification).

Armed with this classification system scientists have been able to subset dietary data according to the Nova classification scheme and conduct, so far mostly observational, studies on the effects of UPFs on our health. Many of these studies have shown associations with some detrimental effects on our health and helped by some inept media reporting a new dietary controversy has been born. TikToks have been posted, books have been published, industry-sponsored research has been conducted, blog posts have been written and influencers have stuck their beaks in. So you know, the usual, and suddenly ‘ultra-processed food’ is being bandied around with no one really knowing what it means, and more importantly, the limitations of both the categorising system it is based on and the studies they are citing.

One of the big criticisms of the way the Nova system is used is that it focuses on processing and not nutritional content. Because of this a UPF can be one of many different types of food and they can differ markedly in nutritional content. Just based on some internet searching, baked beans (but I’d look out for salt content), some cereals, fish fingers, hummus, plant milks, yoghurt, whole grain bread and dried apricots are all UPF foods that may actually be good for you, or at least not bad. The point is that by just lumping all these different foods into one bucket we are ignoring their nutritional content. Intuitively I would say that a wholemeal bread roll should not be in the same category as KFC but that’s exactly what happens when an observational study uses the Nova system. Now, because of these studies, everyone is scared of supermarket bread but should we be scared? We know KFC is not great for us we want to know about bread (here is a good article about supermarket bread).

The next problem is that most of the sensational headlines we see are reporting on observational studies. I recently did a whole blog post on how preliminary scientific data, in the form of observational studies, is the engine room for the craziness that erupts around diet and dietary fads and UPFs are no different. I am going to pick on a UPF study from the SUN cohort, not because it’s bad but because it’s representative of most of these observational studies. This study found a a relative increase in risk “of all-cause mortality of 18% for every serving of UPFs”. So if you have two servings a day versus one your relative risk of death would increase by 18%. This may sound scary but it is a relative measure (see my post on observational studies for an explanation of this). What it amounted to was, after 14 years, a mortality rate of ~2% for the lowest consumers and ~3.5% for the highest consumers of UPFs, a ~1.5% difference in mortality.

The point is that reporting of this study said things like “with each additional serving of UPF, all-cause mortality increased by 18%” (see here and here for example). Most people, who haven’t had the misfortune of sitting through an advanced stats class, interpret that statement a lot differently than they should. Many UPF studies are the same old story of lack of randomisation, confounding factors, results explained in terms of relative risk and hysterical media reporting. This study, for example, found an association between UPFs and accidental death. A counter-intuitive finding that could be explained by considering that people with mental health issues consume more UPFs and also are more likely to die from accidental causes. But mental health is very hard to measure, it wasn’t controlled for in the SUN study and it could easily be an alternative explanation for the observed difference in mortality (or UPFs could cause mental health issues but this is why we need a better understanding – we just don’t know).

OK, enough epidemiologist bashing. I’m not trying to prove UPFs are good for us. I don’t think anyone doubts at this stage that some UPFs are harmful but I think we need a better way of understanding the risk. What we want to know is how bad, what foods in particular and how exactly do they cause harm? We want to be able to do some type of risk management for what we are eating. How may packets of crisps can I eat without having to worry about having a stroke? To understand this we really need to go beyond observational studies and look at what actually happens, we need to find some mechanisms for these purportedly bad effects of specific UPFs. So lets do that now, lets have a look at what we know about how UPFs may be harmful .

Firstly, the no-brainer. Too much salt will cause heart disease and increase the risk of stroke and some classes of UPFs are high in salt. So be careful about high salt foods. Anything above 1.5 grams of salt per 100 grams is giving you a good blast of salt, so minimise these foods. This means not too many crisps unfortunately.

Some UPFs are also very energy dense foods. That is they provide more calories per gram then less processed foods, generally because of added fats and sugars and the lack of fibre and other nutrients. This means that for every gram of UPF you get more calories than if you had an equal amount of unprocessed food. So what? you might say. If you are happy not getting fibre (which you should not be, but anyway) what’s the problem? Just eat less grams. Well the problem is that UPFs are usually specifically designed to be very appealing to our tastes. Can anyone eat just one potato chip? So often we will eat more of a UPF because they taste really good and so get lots of extra calories.

This ability of some UPFs to cause hedonistic eating is exacerbated by their lack of fibre, which can mess with how full we feel. There is evidence that fibre increases satiety by taking up space in the stomach and slowing the emptying of the stomach, which makes us feel full for longer (see here for some research on that). A good example of this is a glass of pulp-free pure orange juice. It takes, roughly, three to four oranges to get 250 ml of juice so in a glass of orange juice you are getting the sugar from four oranges and little or none of the fibre. If you decided to eat oranges instead of having that glass of juice I bet you would stop eating before you got to four oranges. It’s probably worse than this because our household glasses often hold much more than 250 ml but I’ll leave you to do that maths.

The combination of tastiness, lack of fibre and calorie density means we can end up over-eating some types of UPFs and under-eating fresh fruit and vegetables; a good recipe for weight gain and obesity. This type of UPF is often easily identifiable: potato chips, confectionery, some breakfast cereals (the ones with added sugar), doughnuts and soft drinks. We all know we shouldn’t over-indulge in these things. But some of the more insidious UPFs are things like ‘ready-to-eat’ meals that look like a home-cooked meal but have high amounts of fat, sugar and salt. If you are buying these types of products you should really be checking the amount of sugar, fat and salt you are getting when you eat them.

If we are eating lots of UPFs, and studies show that people in developed countries can be getting half of their energy from UPFs (see here), it means we may be eating a lot of sugar. We all know too much sugar can make us overweight but it can also call cause other serious health problems like insulin resistance. Insulin is an especially important hormone in our bodies that plays a role in blood sugar levels and the regulation of hunger. When blood sugar levels are high the pancreas produces insulin which not only represses hunger signalling in the brain but also stimulates the uptake of glucose from the blood stream by muscle cells and adipose tissue (where it is converted to fat). Conversely, when blood sugar levels are low the pancreas produces glucagon which basically does the opposite. Muscle cells don’t take up oxygen, adipose tissue starts breaking down fat and hunger signalling is stimulated in the brain (this is a massively simplified version of what goes on but I plan to do a future post on hunger regulation).

If something goes wrong with our insulin signalling then bad stuff happens. For example, blood glucose levels can become high, a dangerous condition called hyperglycemia, as cells are unable to take up blood glucose and despite high blood glucose levels people can still be hungry if insulin signalling in the brain has been compromised. One way this unfortunate situation can occur is when the body is exposed to high levels of sugar over time, like if we are eating a lot of sugar-rich UPFs, and the effectiveness of insulin signalling diminishes. This is called insulin resistance and it can lead to type II diabetes if blood sugar levels continue to rise, especially so if the pancreas loses it’s ability to produce any insulin under the strain of compensating for the loss of signalling effectiveness.

Nothing I’ve talked about so far has been directly attributable to processing. We could all become obese and develop insulin resistance by eating a diet high in fats, salt and sugar and low in fibre without touching a UPF; Henry VIII managed it for example. But there is some evidence that processing itself is causing the problem. For example, it’s been suggested that processing can contribute to our tendency to overeat UPFs by increasing the softness of food. Making food softer reduces the time we need to process, i.e. chew, it and increases the amount of food that we can cram down before we start feeling full (see here for example and this was even noticed in the 1970s).

Another factor could be the effects of the various additives that are found in UPFs (an area that I found surprisingly understudied). There is some evidence that saccharin, and potentially other artificial sweeteners, can cause insulin resistance, as may carrageenan, a common thickening agent and emulsifier often found in dairy products (see here and here). Sweeteners may also caused metabolic dysregulation by ‘confusing’ our perception of sweetness and calorific content of food (see here for an example). Some other emulsifiers may induce metabolic perturbations and low-grade inflammation in mice (see here). We already know that trans-fats are bad for us and are often found in UPFs and, just to add to any dismay you may be feeling right now, the packaging of many UPFs can introduce a range of chemicals that may have deleterious effects on our health (see here for an example looking at phthalates and bisphenols).

Something that is also starting to get attention is the effects of UPF on our microbiota. This is yet another topic that I plan to do a post on but, if you don’t know, the microbiome is the collection of bacteria that live in our gut and intestine. Over the last couple of decades we’ve come to realise that the composition of this microbiome, the specific species and their relative abundance, has some profound impacts on our health. Our microbiome helps us digest some carbohydrates that we can’t digest on our own, it is crucial for good immune function, it provides a barrier against other harmful bacteria and it can affect what is known as the gut-brain axis that regulates hunger and food intake. It’s early days for microbiome research but dysbiosis (the term used to describe an imbalance in the composition and function of the microbiome) has been linked to mental health issues, inflammatory bowel disease, diabetes and some auto-immune diseases.

The microbiome does appear to be pretty important and evidence is mounting that UPFs can affect it in some detrimental ways. Mechanistic evidence is a bit light on the ground at the moment but evidence suggests that UPFs, particularly those low in fibre and high in fat, can change the composition of the microbiota in such a way that causes low grade, systemic inflammation and oxidative stress. Conditions that have been linked to several chronic diseases. There are also a bunch of studies showing that common emulsifiers and artificial sweeteners can also alter the diversity of the human microbiota, often in ways detrimental to our health. I’m going to do a post on this but if you can’t wait there is a good and recent review here.

So what do we do about UPFs? If we listen to the media we are doomed if a morsel of a UPF crosses our lips but I don’t think things are that dire. The first thing we need to do is understand what type of foods we are eating and try to eat ‘normal’ food when possible. Understand that some UPFs aren’t that bad, wholemeal packaged breads for example, so don’t be too guilty if you are eating those types of food. Understand the importance of looking for fat, salt and sugar content in UPFs and avoid the egregiously bad ones and be aware of how we can easiy over-indulge in UPFs. Maybe consider that it would be easier for us if food that was high in fats, salt and sugar were clearly and prominently labelled as such. Support more research into the additives that are going into our food and look after your microbiome (future post). Also realise that, yes, junk food and ready-made-meals are bad but they are also convenient sometimes and I’m comfortable, based on the current epidemiological data, that the odd junk or ready-to-eat meal is unlikely to take time off your life, particularly if you are eating plenty of other ‘real’ food. Ultimately I guess we all just need to take some agency, understand the issues with UPFs and try to eat as well as we can knowing what we know.