Processed food has changed our world, providing nutritious, tasty food to growing populations. Questions are being raised regarding the “healthfulness” of processed foods, with more and more studies searching for an answer. But first, can anyone tell you what "processed" means?
I suppose we can say that, like pornography, I can identify a processed food when I see it. And while that is reasonable for a casual conversation, when scientific studies are used in crafting public health regulation, we need to be sure we all agree on what “processed” means. A new study shows that for scientists, there is no clear definition.
“…the classification criteria used are ambiguous, inconsistent and often give less weight to existing scientific evidence on nutrition and food processing effects.”
We use classification schemes to identify foods that alter our health, for good or bad. The conventional system found on food labels classified food based on its nutrients. But many contemporary scientific studies ignore nutritional aspects and instead categorize food by its “level of processing” – what is added, lost, how it is transformed. The process-based classifications assume a correlation between processing and nutrition that is not proven. Some argue that foods should be classified based on both nutrients and processing. There is no scientific consensus.
The most widely used food classification is the NOVA system , increasingly used as the basis for dietary guidelines. It has four categories
- unprocessed and minimally processed foods
- processed culinary ingredients – butter or pasta are in this group
- processed foods
- ultra-processed foods
In this qualitative study, researchers screened 470 publications for food processing definitions, identifying 146 definitions and four underlying thematic approaches. All of the categorizations reflected how a product was transformed, often beginning by defining “unprocessed” using terms like “natural” or “raw” – but what is the definition of those words?
The extent of change from its natural state
- A minimally processed food is a single whole food - right off the vine, say blueberries. Some food classification systems require minimally processed fruits and vegetables “to retain an active metabolism of plant tissue.” By that definition, freezing or drying blueberries makes them a processed food.
- Processed foods combine whole foods with ingredients. A tomato picked, steamed, and canned is processed food.
- Ultra-processed food is a combination of ingredients, typically containing no “whole” foods and frequently described as not “real food.” All those plant-based burger meats, mixed with tomato paste, seasonings, and pasta, are not lasagna. It is ultra-processed and therefore not real food.
But these systems are not aligned. In one classification, pasteurized milk is minimally processed; in another, it is highly processed. A transformation that we would consider natural, like fermentation, produces only highly processed foods. Does yogurt seem highly processed?
The nature of change - changes to a food’s inherent properties, or addition of ingredients
This categorization reflects an unsettled scientific debate. Should we consider food and diet in a reductionist way – as nutrients, or more holistically, as a synergism between all the bioactive compounds found in a “whole” food.  Many “minimally processed” foods in one classification undergo physical processes that disrupt the “food matrix,” making them “processed” by another classification.
The difficulty is that these classifications do not reflect, in a systematic way, changes in a food’s inherent properties. For example, thermal treatment of tomatoes will reduce Vitamin C but increases the bioavailability of lycopene – does that make it more or less processed? Crushing a seed destroys its matrix rendering it no longer “whole,” but its nutrients more available. Where does this change belong on the processing scale, and does that change reflect the food’s nutritional value?
“Merely adding ingredients is deemed a form of processing.”
We are all familiar with the addition of sugar, salt, or fat to processed foods. But classification schemes often consider the intent of the additive in determining the degree of processing.
- Enrichment, replacing nutrients lost in processing, is an example of minimal processing.
- But the same additive, when used to fortify, meaning to add a nutrient irrespective of whether it is usually found in that food, makes the product highly processed.
Both enhance a product's nutritional value. How does the categorization of the processing enhance our knowledge concerning our health?
Place of processing
Home-made foods or artisanal foods are generally given a pass, considered as less processed. But while we are not using extruders, and other industrial equipment and ingredients, does that automatically mean home-cooked is unprocessed? If I make gelatin spheres that burst in your mouth using a molecular gastronomy kit at home, why is it less processed than if created in a restaurant kitchen?
Purpose of Processing
Some schema considers the reasons or intent of processing - an arena fraught with opinion and bias, especially true for additives. Additives for food safety or to prevent food spoilage create minimally processed foods. Additives designed to influence our purchasing behavior are “cosmetic,” making the product ultra-processed.
The Underlying Science
The researchers believed that while “there are some links between the evidence and the design of the classification systems, they often lack a solid underpinning.” These food classification systems are designed to facilitate a scientific basis for public health initiatives and regulations. While those delicious and addictive, ultra-processed foods are consistently associated with obesity and cardiovascular disease, there is no firm basis for whether this is due to processing or nutritional content. In fixating on processes, do we ignore “consideration of a balanced diet and portion sizes?”
“From the perspective of food science and technology, processing and nutritional value do not have a linear relationship and these concepts need to be dissociated.”
More importantly, there is not a direct linear relationship between processing and food’s nutritional value. For example, one definition of “highly processed” includes cereal and milk, “although their consumption tends to be inversely associated with several chronic diseases.”
And no apparent correspondence between categories and toxicologic risk. Acrylamide is formed in the “Maillard reaction” when asparagine, an amino acid, is heated in the presence of sugars found in starch-rich foods, creating the tasty crust of grilled meats. Acrylamide content is monitored and regulated by the food industry, not so much when Dad’s grilling. Asparaginase, an enzyme, significantly reduces the formation of acrylamide, a health benefit. Asparaginase is used in industrial frying and is an additive that demonstrates a lack of correspondence between additive and toxicologic risk.
They also note these categorizations seem invested with views and opinions that might bias what studies report. Consider these examples,
- “Most of the classification systems are introduced by expressing concerns about the food transition to industrially-made products and a concomitant increase in chronic disease.”
- One expert warned, “ultra-processed foods contain ‘authorised, but controversial’ additives, referring to sodium nitrite and titanium dioxide “for which carcinogenicity has been suggested in animal or cellular models.”
All of these phrases and words are “value-laden.” Much like the terms natural and whole, which have no clearly defined meaning for foods. Or “junk food” or “empty calories” - pejorative words, suggesting shoddy, poor quality.
Food classification in science is another of those human social constructs. As such, they are susceptible to human bias. As the researchers' question, is focussing on processing “the optimal approach for food communication?”
 Center for Epidemiological Studies in Health and Nutrition, School of Public Health, University of Sao Paulo, Brazil
 “It has been noted that the nutrients tracked and cataloged by food databases represent a small fraction of the >26,000 distinct, definable biochemicals in food.”
Source: Processed food classification: Conceptualisation and challenges Trends in Food Science & Technology DOI:10.1016/j.tifs.2021.02.059