The True Cause of Ultra-Processed Foods' Negative Health Effects

whether I were to show you any foods, you would generally be able to tell whether they were highly processed or not. What do you know? Let us give it a go.

Oscar Meyer Weiner, the classic? Yes. extremely processed. It has many additions, including "mechanically separated meat," dextrose, glucose, and sodium lactate.

A handful of almonds, perhaps?

Again, exactly! Not too manipulated. A good bunch of almonds really scores a 1 on the conventional "NOVA" processing scale, which ranges from 1 for unprocessed or slightly processed foods to 4 (ultra-processed). Well done.

Doritos? extremely processed.

Tostitos? just "processed."

Having just maize, canola oil, and salt gives it a NOVA score of 3. California Olive Ranch Olive Oil is a nice example of a 2 on the NOVA scale; yes, processing is necessary to convert olives into oil, but that is about it.

All of this is similar to the Potter Stewart food classification system. You may be familiar with this tale. Stewart notably commented on a censorship issue while serving as a Supreme Court justice from 1958 to 1981. The issue was whether a French film that was very explicit might be prohibited due to its extreme pornographic content.

Although it was difficult to define hardcore pornography legally, he acknowledged that "I recognize it when I see it." The same is true with highly processed meals. We all have a broad idea of what they are, but it may be difficult to find a universal description. And you could be taken aback. For instance, most ordinary milk is not considered ultra-processed, but the majority of almond milk is. To put it another way, being highly processed does not automatically equate to being less healthful. However, compared to their non-ultra-processed counterparts, ultra-processed meals are often unhealthy. To be honest, determining which foods are ultra-processed has been pretty subjective up until recently.

However, a new technology has emerged that eliminates the element of guessing from this procedure. It will be important for consumers attempting to consume less processed foods as well as for researchers.

Because ultra-processed meals are often unhealthy. But maybe not for the exact reason you believe.

Allow me to demonstrate how an ultra-processed meal is categorized using the conventional NOVA system:

Oh no. It sounds less like something you would like to eat and more like building a vehicle.

However, many food items do not neatly fall within the group. In order to address this, Giulia Menichetti and her Harvard colleagues wrote a paper in Nature Food. They created a machine-learning model that can process any ingredient list, such as the type found on almost every packaged product you purchase at the grocery store, and produce a score. They call this score "FPro," or food processing score, which indicates the precise level of processing of a given food.

In order to construct the algorithm, they trained it using a database of meals that had been meticulously classified by humans into the four NOVA categories. However, machine learning models have the advantage of being able to be applied to previously untested data. And they did just that. Allow me to illustrate.

The score goes from 0 (essentially raw items) to 1 (the most processed item). Yes, I looked into their real database to find out which particular foods served as the research's bookends. Organic ground beef had a score of zero, whereas Wonder White hamburger buns and Ball Park Everything hamburger buns tied at 0.999056. Go without a bun on your next burger, I suppose.

Two cheesecakes are less severe instances.

Edwards Desserts Original Whipped Cheesecake is on the left, and Pearl River Mini No Sugar Added Cheesecake is on the right. Although the lengthy ingredient list on the left is a little concerning, Potter Stewart does not really assist me with this. I think they are both probably quite processed. However, we get a number—an real score—from the algorithm. The Pearl River sweets (14 ingredients, 5 additives) score 0.720, whereas the Edwards desserts (43 ingredients, 26 additives) score 0.953. According to the authors, giving customers rankings like this would enable them to make wiser decisions. However, there is a little flaw in that reasoning—pay attention.

We are not here to criticize specific foods. This new model's main selling feature is that it may be used with any food item you buy. It is also automatable.

The authors scored each and every food item that was sold at Whole Foods, Target, and Walmart using that automation.

Let us see how these three supermarket chains' scores are distributed. Although Whole Foods seems to have a much flatter curve, overall, you can see that there are often more goods at the upper end of the processing scale. It is interesting to see that Target has a higher percentage of ultra-processed items than Wal-Mart, which would not have been my first notion.

The range of processing scores within a certain category is shown in these boxes. Variability abounds throughout the world. Sure, it is difficult to locate less-processed cookies and biscuits, but if you search, you may discover yogurts, coffee beverages, and spreads that are comparatively less processed.

Is there a fee associated with all this processing? Not to your bank account. The researchers discovered that foods were often less expensive per calorie the more processed they were. That cheesecake that Edwards whipped? A serving costs $1.87 and has 260 calories, or 1.4% of total calories. The cheesecake from Pearl River? 260 calories, or 0.65 calories per %, at $4 a serving.

The more processed the food, the better if you are attempting to stretch your food budget. Of course, you will pay in other ways.

I was unable to resist. I looked through their database to find the cheapest meal per calorie. The victor? Super-Moist German Chocolate Cake Mix from Betty Crocker. 160 calories, or around 11 calories per cent, at 15 cents a serving. Amazing.

Ultra-processed, mmmm.

The "more processed goods are cheaper" conundrum was seen in almost every food category. I believe this happens because, although while processing costs money (people, machinery, raw materials, etc.), it is carried out on an industrial scale, which lowers individual prices.

Nowadays, a dish cannot be considered ultra-processed due to a single component. It is everything that goes into producing a food item. However, because every one of the more than 50,000 items in their database now had a score, the researchers were able to identify which individual elements are most often found in diets with high overall ratings. These elements, which I have labeled "red flags," may not even be highly processed in and of themselves, but they are indicators that the product has undergone significant processing. These are most likely the ones you want to be aware of while reading food labels.

The tale of additional oils is explained in this graphic.

Peanut and flaxseed oils are less reliable indicators of heavy processing than palm and maize oils. By limiting my search to components that appeared in at least ten items, I once again examined their raw data to determine which compounds were at the very top of the red flag list. The other ingredients were sorbic acid, spice oil (which I can only presume is from Arrakis), oat mix, crust grain oats, and "palm kernel oil with TBHQ for freshness."

As I said in the beginning, you have good cause to be concerned about highly processed foods, but you may be worried for the wrong reasons. You may be concerned that these additives are poisonous when you notice items like "sorbitan monostearate" or "palm kernel oil with TBHQ for freshness" on these ingredient lists. Indeed, there may be studies in rats that demonstrate that this color or that preservative causes cancer or anything if you search for it long enough on Google. But to be honest, this is not how eating a lot of manufactured food is killing us.

It is not that complicated.

Ultra-processed foods are particularly unhealthy since they are more convenient to consume. These additives? They add chemicals to prolong the food's shelf life, ensuring that it remains soft when intended and crunchy when intended. They add exquisite oils, spices, and salts. Grains that have had their fiber removed to improve their mouthfeel and chewability. All of them were created by very intelligent scientists to be really tasty and simple to consume.

Twenty participants participated in a historic 2019 research that required them to spend a month at the NIH. They were given a typical diet for two weeks, followed by a diet heavy in highly processed foods for another two weeks. The items they were shown were matched by macronutrients, energy density, calories, etc., and they were free to consume as much or as little as they desired. When randomly assigned to the ultra-processed diet, the typical participant consumed 500 more calories daily.

Simply put, we are killed by these foods because we are unable to stop eating them.

In the business world, this is even referred to as cravability. It essentially says, "We want this to be as near to addictive as possible," which is what they are aiming for.

And it is effective. I adore blueberries, but after I have had some, I do not really have any trouble quitting. However, have you ever made an effort to cut less on Doritos? It is difficult.

Here, I disagree with the authors, who propose that we may use FPro ratings to determine whether alternatives within a food category are better for us, such as less processed cheesecake versus more processed cheesecake. That concept has a flaw in that the healthier option, the one with the lower score, could not taste as nice.

And we have to accept that. In fact, this is how the cycle of highly processed meals is broken. It is necessary for us to learn to appreciate food that does not immediately hit our taste receptors or cause a dopamine spike from the scent alone. We must rediscover our enthusiasm for authentic cuisine.