UPFs are the new bad kid on the block when it comes to diet and health. But what are they and is it possible that the UPF concept could be doing more harm than good?

The fields of dietary and nutritional science have captured the public’s imagination. We are experts at eating after all, knowing how much and when to eat, at least sufficiently well to have survived long enough to be reading this piece. The flipside to all of this public interest is that nearly everyone has an opinion, and in this world of social media, most are unafraid to express it. It is therefore a very noisy field in which to work, full of waxing and waning dietary fads, with foods once bad now thought to be good, or good now bad, and even good and bad at the same time! Which brings us to the hottest topic du jour, “ultra-processed” foods or UPFs.

More extensive levels of processing shift foods into groups 3 and 4. Processed foods (group 3) can be made by adding sugar, oil, salt, or other group 2 substances to group 1 foods for instance. Processes include various preservation or cooking methods, and, in the case of breads and cheese, nonalcoholic fermentation. Everything else falls under the umbrella of ultra-processed food and drink products (group 4). As defined by Monteiro, UPFs are, in effect, industrial formulations, typically with 5 or usually many more ingredients, that most would find difficult, if not impossible, to replicate in a domestic situation.

What exactly is it about UPFs that makes it so apparently bad for us? First, because of the degree of processing, most UPFs are inherently lower in protein, fibre, or both. This makes UPFs very calorically available, meaning that they are easily digestible, and our body can extract a large proportion of the calories in the foods. Second, the high degree of processing also strips out much of the flavour, which then has to be replaced for the foods to be made palatable; thus, UPFs often contain higher amounts sugar, salt, and fat. Third, and most controversially, some argue that it is the processing that is inherently bad; the fact that many of the foods are so highly engineered. In truth, the evidence supporting this third reason is equivocal at best, and entirely absent at worst.

There are, however, a couple of points to reflect on. First, the umbrella review rated the quality of evidence as low to very low for nearly all outcomes, and thus its conclusions need to be considered in that light. Second, we need to seriously consider if all UPFs are created equal. For example, there are foods that appear to be group 1 or 3 foods, but because they contain additives, become classed as group 4. So plain yoghurt is a group 1 food; however, if flavourings, sugar, artificial sweeteners, or sweetened granola are added, it becomes ultra-processed and moves into group 4. Bread is a processed food, but if made with added emulsifiers and gluten, it becomes ultra-processed. Surely, there are many others who feel that the imprecision of the term UPF is an issue?!

The biggest methodological weakness in dietary and nutritional science is the difficulty to conduct truly blinded randomised control trials (RCTs). Unlike in trials of drugs and other purified compounds that allow for properly blinded placebo controls, it is impossible to do this with whole foods. Thus, the vast majority of dietary studies are observational. Ultimately, the success of such studies depends on power and a precise measure, both of intake and outcome. UPFs encompass such a broad range of foods, some of which we should certainly be eating much less of; but others, including the aforementioned supermarket bread that sneak into group 4 with just the addition of emulsifiers and few other additives. Thus, many studies that use the NOVA criteria are reporting that UPFs, on average, are linked to poorer health outcomes. Which of the literally thousands of types of UPFs are truly “bad” for you, and which are harmless foods that have hitchhiked along through some quirk of nomenclature?