‘Ultra-processed’ food – myth or a viable classification parameter?

In the wake of the obesity pandemic, the concept and demonisation of “ultra-processed foods” has moved into the popular – and academic/professional – mindset. Here one of SA’s leading food scientists, Nigel Sunley unpicks this thorny topic in an article based on a presentation he made at the SAAFoST Congress, Cape Town, in early September.

IT IS DIFFICULT to move around in the world of food activism and in some nutrition and government circles these days without hearing at some point that so-called ‘ultra-processed’ food is purportedly at the root of most of our obesity and other diet related problems.

At the outset it is important to note my use of both the phrase ‘so-called’ and the use of quotation marks around ‘ultra-processed’ itself.

This is because the term is emphatically not an established scientific or nutritional one, but is rather simply a concept propagated by one segment of the public health nutrition community that proposes it as a potentially more valuable classification parameter for assessing the nutritional quality of food than the classic quantitative measures of different nutrients. 

The concept is the brainchild of Carlos Monteiro, an academic public health nutritionist at the University of Sao Paulo in Brazil.

Professor Monteiro has a long established track record in public health nutrition but appears to have become obsessed with the ‘ultra-processed’ concept and first proposed it in a paper published in the journal Public Health Nutrition in 2009.

Since then he has acquired a set of like-minded followers, seemingly drawn from the left-wing fringes of the nutrition community along with an array of social scientists and miscellaneous academics, all of whom appear to share the common features of a paranoid dislike of the food industry, coupled with a spectacular lack of understanding of its workings.

Perhaps the most extreme proponent is the somewhat eccentric British academic, Geoffrey Cannon, who has the dubious privilege of being the only person I have ever heard being shouted down at a scientific conference!

This occurred at the IUNS Congress in Spain in 2013 where a long-suffering audience including myself in a lecture session eventually tired of his monotonous rambling interventions about what he claimed to be biased results from industry-sponsored nutrition research and simply told him to shut up!  

Be that as it may, the ‘ultra-processed’ concept has been the subject of numerous other publications in credible peer-reviewed journals since 2009. I was rather smugly and sanctimoniously informed by another of its followers at the World Public Health Nutrition Congress in Cape Town in 2016, where I had given a presentation strongly attacking it as a concept, that ‘we have published peer reviewed articles about it so it must be right’!

Disregarding the food scientists

What these same followers fail to mention is that at no stage does there appear to have been any attempt to consider food science principles or involve food scientists in developing or justifying the ‘ultra-processed’ concept – certainly there appears to be a complete absence of any food science related references in any of the published papers by Monteiro and his supporters.

Further, it is not unreasonable to speculate as to whether any food scientists have ever been asked to peer review the papers concerned – they after all are the food processing scientific experts and could reasonably be expected to have some say in the matter.         

So, it is perhaps equally reasonable to take look at the ‘ultra-processed’ concept from an unemotional perspective and ask the question as to whether it should be used, as its supporters propose, as a formal classification parameter for assessing nutritional quality of individual foods?

Perhaps the starting point should be setting out a proper and scientifically rigorous definition for ‘ultra-processed’ food. Best way of doing this is to refer to the various published papers.

Monteiro’s original 2009 paper divided foods into three categories, namely Group 1 – ‘unprocessed or minimally processed foods’, Group 2  – ‘processed culinary or food industry ingredients’ and Group 3 – ‘ultra-processed food products’.

‘Ultra-processed’ is then defined as ‘…processing of a mix of Group 2 ingredients and Group 1 foodstuffs in order to create durable, accessible, convenient and palatable ready-to-eat or ready-to-heat food products liable to be consumed as snacks or to replace home prepared dishes.

Specific processes include baking, frying, deep frying, use of additives and cosmetics, addition of vitamins and minerals, salting, canning and sophisticated forms of packaging’.

Quite a mouthful (pardon the pun) and one can already see the cracks in the definition in terms of its sweeping and vague generalisations, not to mention the lack of any form of quantitative measurement. Still, if that is his definition, that must in turn be considered our reference point.

Escaping definition

However, not a single subsequent paper on the subject uses this definition, but all of them resort to their own equally vague and rambling interpretations containing phrases such as ‘…they contain little or no whole food’, ‘…from substances extracted or refined from whole foods’ and ‘…to disguise undesirable properties of the final products’.

There are certainly some broad-based threads throughout the various definitions, but at no stage is there a single consistent and unambiguous definition and that alone puts the use of ‘ultra-processed’ as a classification parameter into serious doubt.  Furthermore, a classification parameter can only be considered in the context of its proposed usage in practice.

No prizes for guessing that those concerned wish to limit the manufacture, marketing and consumption of what they define as ‘ultra-processed’ food as part of their simplistic and vehemently anti-industry platform, although what they propose as a replacement is typically vague and appears to usually comprise some form of rather woolly, ‘back-to-nature’ approach.

Try telling that to a harassed lower-income consumer living on the 20th floor of an apartment block in Professor Monteiro’s home town!

For the ‘ultra-processed’ brigade however, regulation is the answer – real Big Brother stuff with strict government-imposed controls on food composition, processing, marketing and retailing.

This is where the concept really starts to fall apart and demonstrates the sheer technical ignorance and impracticality of its supporters. If you are going to control something by means of formal regulation, a key component of the regulation has to be an accurate and legally acceptable definition of whatever it is you are trying to regulate.

Can you even imagine trying to regulate food on the basis of the vague, subjective and unquantified Monteiro definition?

It would be a legal and scientific nightmare and no government could even begin to convert it into formal legislative terminology, never mind embark on the immense and tortuous path to bring it into operation and, equally importantly, enforce it.

Equally, could you use the law to control the vast number of product types, processes and raw materials listed in the Monteiro definition? The answer is clearly a resounding no.

The supreme irony is that the cornerstone of the argument from the ‘ultra-processed’ supporters is their position that what they define as ‘ultra-processed’ foods are considered by them to generally possess poor nutritional properties in terms of individual nutrients such as fat and sugars.

Even if this were true (which is highly debatable to put it mildly), why can’t we just stick to assessing nutritional quality by means of quantitative measurement of individual nutrients as we already do? In this context, the ‘ultra-processed’ concept is completely pointless as we have to come back to measuring individual nutrients anyhow!

A more scientific definition

However, let us try to give the people concerned a bit of seemingly much-needed proper scientific help and see if we can develop a definition for ‘ultra-processed’ food that is a bit more specific in nature and might have some chance of standing up in court.

How about ‘…a food that has been produced using multiple processing steps and which bears no resemblance to its original constituents’.

Admittedly not ideal and the attorneys would have field day with it eg how many processing steps is ‘multiple’ and how do you define ‘resemblance’, but at least a bit more manageable than the rambling and unworkable definitions of Monteiro et al.

Alas, we immediately hit problems again. It is well known that two of the products targeted by public health activists and which would clearly be on the hit list of any ‘ultra-processed’ based legislation are potato chips and carbonated soft drinks.

Ironically neither of these involve complex processes, with potato chips involving 5 stages (clean and peel potatoes / slice / fry / season / pack) and soft drinks a mere 4 stages (blend syrup / carbonate water / mix / pack) and are processes which, other than packing and carbonation, are pretty similar to common domestic kitchen processes.

At the other end of the spectrum, a product involving no less than 13 processing steps and for which the finished product bears absolutely no resemblance whatsoever to its starting ingredients, turns out to be a well-known South African breakfast cereal.

It is based on soya beans and maize which have to undergo extensive processing to render them palatable and improve their nutritional quality, and for which the finished product has arguably a better nutritional profile than any other processed or unprocessed food in the world!

Inconsistencies such as these (and there are many more) simply serve to knock another nail in the coffin of the ‘ultra-processed’ concept as a viable classification parameter. 

Looking at the bigger picture

I have already referred to the broad make-up of the ‘ultra-processed’ support brigade. They appear to comprise a group of public health nutritionists and social scientists with strong anti-industry and anti-globalisation credentials whose primary objective seems to be to damage the food industry in any way they can, irrespective of the effects on global food supply, and for whom ideology seemingly overrides the practical requirement to feed over 7 billion increasingly urbanised global consumers.

They also share a common ignorance of food science and technology, supply chain and logistical issues and, perhaps above all, a complete inability to grasp the concept that food needs to taste good.

It may be a little flippant, but there is a certain symbolism in the fact that that the politically-correct ‘people’s food’ catering at the World Public Health Nutrition Congress in Cape Town in 2016, where the ‘ultra-processed’ brigade were in full cry, was frankly revolting, and I noticed envious eyes being cast by the delegates at the burgers and fries that were being happily munched by the students at the next-door venue!

On a slightly more serious level there are intriguing similarities between the ‘ultra-processed’ brigade and those demanding ‘radical economic transformation’ of the South African economy with proponents of both concepts making a lot of noise but completely unable to translate the noise into anything practical.

What conclusions can we start to draw?

One thing is for sure, namely that neither the food industry nor processed food are going to disappear as a result of all this fuss – nor should they as they both provide a vital role in feeding the world and those opposing them have completely failed to provide any viable alternatives to either.

But the most important point is that it should be self-evident that there is processed food of good, average and poor nutritional quality, and it would be grossly unrealistic, not to mention highly unscientific, to base nutritional assessments on the processing status of the food concerned.

We have perfectly adequate measures already in place for determining nutritional quality in as objective a manner as can reasonably be expected, namely the tried-and-tested analytical methods for determining the levels of both macronutrients and micronutrients.

These have the advantage of being supported by established and rigorous scientific methodology and above all are quantitative measures which can be used to set standards and, where appropriate, act as a base for regulatory measures.

However, to base any form of regulation on the ‘ultra-processed’ concept would be impractical to the point of being laughable due to the complete lack of any clear and meaningful definitions.

As things stand, it appears that its brigade of supporters would like to define an ‘ultra-processed’ food as ‘anything that we say it is’, such is the scientific weakness of their position. 

So – any time that some maybe well-meaning, but spectacularly ill-informed, public health nutritionists or consumers start shooting off their mouths to you about the purported evils of ‘ultra-processed’ food, I suggest you gently ask them for a proper scientific definition of their concept and how they intend using it in practice, and also politely suggest they go and learn a bit about food science and technology.

I am sure we have not heard the last of this rather silly idea – indeed Monteiro and company managed to get a session on it into the program at the recent International Congress of Nutrition in Buenos Aires – but my suspicion is that the mainstream nutrition community regard it with considerable scepticism.

It likely will remain with us as an irritating and rather pointless fringe concept but with little mainstream support, let alone practical application and emphatically not as a classification parameter.  

About Nigel Sunley:

NIgel SunleyNigel Sunley is a highly experienced food scientist, with extensive commercial experience, and is one of SA’s foremost labelling and regulatory experts. See more here

E-mail: [email protected]

Additional reading: The food industry and obesity – find the middle ground

No Comments

Sorry, the comment form is closed at this time.