Ultra Processed Foods: Are they the enemy they’ve been made out to be?
A review paper was published in the British Medical Journal in Feb 2024 leading to a plethora of terrifying headlines linking ultra-processed food (UPF) to harmful effects on health. No less than 32 effects were listed including cancer, heart disease and early death.
Before we throw all our sausages away, let’s put the brakes on for a minute and look at what this study involved, what they found and more importantly, what they didn’t find.
The study gathered a shed load of data from hundreds of past studies that had already been reviewed. This was a review of reviews, so it’s called an “umbrella review”. We’re talking about a lorra, lorra people being observed here - nearly 10 million, over 14 years…This has to make this a good study right?
Well, not so fast my friends, size isn’t everything…
They were observational studies. Essentially this means just collecting information; no experiments or interventions. In this case, the researchers were looking at how much UPF people ate, then see if there were any connections with adverse health effects. Different types of subjective dietary surveys would have been used to capture UPF intake; food frequency questionnaires (FFQ’s) and diet records that would have been completed by the study participants themselves.
The thing is, these collection methods have their own limitations:
They're not able to accurately quantify UPF eaten because this would mean too many questions and losing the will to live trying to complete the form, so you don’t. Most specify a frequency response relating to an average or medium portion, which is rather vague.
FFQ’s consist of specific foods lists which may not be culturally sensitive and can become outdated.
Self reported food intake is prone to inaccuracies, because even unconsciously, social norms and dietary expectations tend to influence what we write down. We can both under-report and over-report as a result which can alter study outcomes and conclusions.
They rely on generic memory rather than specific memory due to the nature of the questions asked and because the further back we try to remember, the more patchy we get. Some FFQ’s ask how often things have been eaten over the last year!
And there is an assumption the participants carried on eating the same until they were next followed up, which can be several years later.
With this in mind, how can they possibly establish any associations between UPF intake and adverse health outcomes with certainty? And to be fair the researchers admitted this, acknowledging there could be all sorts of other reasons that could explain the results.
Research must account and adjust for possible influencing (known as confounding) factors - like gender, level of education, alcohol consumption and physical activity, although it’s practically impossible to do so completely. Unfortunately, really important factors are often not considered. Social determinants of health like anti-fat bias, food insecurity, discrimination and access to affordable healthcare are key when trying to understand health outcomes. Not accounting and adjusting for these further reduces confidence in these research findings.
The associated increase in risk for the 32 adverse health outcomes and UPF consumption was not as serious as was implied in the news. If you look at the image below, the majority of concerns (under Grade) were either low or very low. And the trustworthiness (Credibility) rarely got above suggestive.
This review can’t tell us if the risks noted are seen with all UPF’s.
It can’t provide any causality between ultra-processed food and adverse health risk.
It is unable to tell us what in the UPF’s could go towards explaining the increased risks reported, meaning we’re a long way off from “how, why and when?” There’s no strong evidence that points to colours, emulsifiers or sweeteners being the “problem”. Or does it come back to these foods being more likely to contain higher levels of fat and sugar and less fibre? Although this is not always the case.
So despite the shit scary headlines, please know we are some way off from determining if UPF’s carry real risk, and why. The lack of conviction means we can already challenge the belief “all UPF’s are bad”. Join me next time when I talk about protective UFP’s and Joe Wicks’ recent cock up!
If you are worried your diet is all UPF’s and you’re feeling 'unhealthy' because of this, book in for a free 30 minute chat online.