Making Sense of Ultra-Processed Research Clickbait

Clickbait Art at Bondi JunctionNutrition research in medical journals follows trends that define what the cognoscenti can regard healthy – or not. For decades, the bad stuff was fat. Then we switched to the sugar is toxic meme and that was the preoccupation through the 2010s. Now there can be no doubt. Research on ultra-processed foods is providing a steady stream of clickbait with an unwavering message: “ultra-processed foods are killing us.”

The latest installment came from BMJ this week. It is an umbrella review of correlations between ultra-processed food exposure and health outcomes. The finding could hardly be less surprising:

“Greater exposure to ultra-processed food was associated with a higher risk of adverse health outcomes, especially cardiometabolic, common mental disorder, and mortality outcomes.”

Headlines flowed freely from there: “Ultra-processed foods linked to 32 health problems.” And so on into the hundreds of stories.

The Collision of Political,
Social, and Nutrition Science

The definition of ultra-processed, however, is not a very good foundation by itself for deciding which foods are good for you. It is not a tool for describing the nutrient value of any particular food.  Dietitian and nutrition consultant Linn Steward explains the origins and utility of the NOVA system for identifying ultra-processed foods:

“NOVA was a social political reaction to the global food industry‘s arrival in Brazil. It is not a nutrient profiling system and was never intended for that purpose. NOVA captures the value of freshly prepared meals, minimally processed ingredients, social cohesion, and cultural authenticity.”

We should not have to choose between using the NOVA system and nutrition profiling to understand our food. They serve different roles. “Food is more than the sum of its parts,” says Steward.

Finding the Meaning of Ultra-Processed Research

So how do we make sense of the mind-numbing repetition of the same observational research finding over and over again? Diets comprised mainly of ultra-processed food correlate with a higher risk of chronic, diet-related disease. We would suggest two thoughts are worth keeping in mind.

One, ultra-processing is not a sufficient criteria for constructing a list of good and bad foods. Because foods that qualify as ultra-processed are very diverse. Some can be the foundation for a very nourishing meal. Others might be best for only very occasional consumption. Glazed doughnuts do not comprise the breakfast of champions.

Two, the NOVA system does have some utility for identifying problems with the emerging global food supply. We need much more research to fully understand all of the factors – palatability, marketing, distribution, social systems, and more – that are contributing to the relationship between ultra-processed foods and chronic disease prevalence.

This is quite a large task indeed. But not one we can afford to neglect.

Click here and here for a sampling of recent ultra-processed research clickbait. For the recent BMJ paper, click here. Finally, for an excellent discussion with Kevin Hall about making sense of the issues with ultra-processed foods, click here.

Clickbait Art at Bondi Junction, photograph by Floodstreet, licensed under CC BY-SA 4.0

Subscribe by email to follow the accumulating evidence and observations that shape our view of health, obesity, and policy.


March 2, 2024

2 Responses to “Making Sense of Ultra-Processed Research Clickbait”

  1. March 03, 2024 at 9:23 am, David Brown said:

    Close to minute 26 Dr. Hall mentioned that a large fraction of ultra processed foods are high in saturated fats. He admits, though, that researchers do not yet have a mechanistic explanation for why ultra-processed foods cause people to overeat. Perhaps that is because they are not taking the endocannabinoid system into account. Excerpt: “With dietary fats being the only source of FA required for synthesis of endocannabinoids, it is possible that what is being consumed is capable of modulating circulating endocannabinoid levels, thus influencing GPCR signalling in an acute time frame and affecting appetite and subsequent food intake.

  2. March 04, 2024 at 3:30 am, Mary-Jo said:

    ITA with Linn. It’s difficult to use *processing* as a criteria to evaluate how healthy, how ‘good’ foods are for us and to make associations with consumption of processed foods with obesity and other diseases. If by ‘ultra-processed’ foods, researchers and public health experts REALLY mean foods of low nutritional value vs. high content of their calories, sugar, fat, sodium — what we used to call ‘junk food’, why not just use that as a classification or even just the nutritional value profile of foods? There used to be a classification called the INQ -index of nutritional quality- that assigned discrete values to foods based on nutrient-calorie ratio. Imo, it made it very handy for objectively (as possible) investigating quality of dietary intake – getting measurable scores, useful in research and clinical assessments. It’s helpful to patients to see their initial scores and track scores as they made more healthful changes. It’s a bit of work on the part of the dietitian, but now I’m sure software could would make it easier! It’s gone out of favor now, but might be interesting to resurrect it!