Rayonist Sausages and Mackerel

Defining the Slippery Ultra-Processed Boogeyman

New research keeps popping up to remind us that ultra-processed food is a slippery boogeyman. Headlines about their addictive and toxic properties confront us everywhere. They’re “the worst” for your heart, say headlines quoting cardiologists. But the problem with this boogeyman classification is that it’s slippery.

A new paper this week reveals that even experts in nutrition have a tough time identifying ultra-processed foods. In fact, the study found poor consistency in the ability of food experts to sort foods into four different classes of the NOVA system for defining ultra-processed foods. So the authors of this paper, led by Véronique Braesco, concluded:

“Although assignments were more consistent for some foods than others, overall consistency among evaluators was low, even when ingredient information was available. These results suggest current NOVA criteria do not allow for robust and functional food assignments.”

Eye of the Beholder

In their study, Braesco et al asked French food and nutrition specialists to assign foods to the four NOVA groups according to the level of processing of each food. These specialists had two different lists of food to sort. One was a list of 120 marketed products with a complete list of ingredients for each. The other was a list of 111 generic food items with no specific list of ingredients provided.

It made no difference. Because only three of the marketed foods and one of the generic foods received consistent ratings by the experts. Most often, the evaluators placed these foods into two, three, or even all four different groups. Ask a different expert how highly processed a food is and you’ll likely get a different answer.

Inconsistent Associations

Another recent paper tells us that the heterogeneity of ultra-processed food classification leads to inconsistent associations of these foods with type 2 diabetes risk. Ming-Jie Duan and colleagues identified four different diet patterns – all with habitual consumption of ultra-processed foods. Two of those patterns correlated with a higher risk of diabetes. But one of them had no link to diabetes risk. Then the fourth – a pattern high in sweet snacks and pastries – had an association with less risk for diabetes. The authors concluded:

“The heterogeneity of UPF as a general food category is also reflected by the discrepancy in associations of four distinct UPF consumption patterns and incident type 2 diabetes.”

Lacking Precision

Ultra-processed foods have taken on the reputation of a boogeyman despite the fact that their definition is slippery. Even experts can’t seem to sort foods according to the NOVA classification system with good consistency. The classifications leave a lot of wiggle room with different foods seeming to present different degrees of risk – despite falling into the same ultra-processed definition.

So it seems that trying to reform food systems to resist ultra-processed foods might be a bit premature. As a first step, we need better definitions.

Click here for the Braesco study and here for the Duan study.

Rayonist Sausages and Mackerel, painting by Mikhail Larionov / WikiArt

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


March 23, 2022

One Response to “Defining the Slippery Ultra-Processed Boogeyman”

  1. March 23, 2022 at 2:55 pm, John DiTraglia said:

    Like fast food and junk food it’s just not a thing.