“Ultra-processed foods are operationally distinguishable from processed foods by the presence of food substances of no culinary use (varieties of sugars such as fructose, high-fructose corn syrup…modified starches…modified oils…protein sources such as hydrolysed proteins, soya protein isolate, gluten, casein, whey protein and 'mechanically separated meat') or of additives with cosmetic functions (flavours, colours, emulsifiers, thickeners…) in their list of ingredients.”
The history of NOVA is that as the obesity epidemic was spreading, a Brazilian nutrition scientist, Carlos Monteiro, did a large scale project to identify likely causes right as it reached Brazil and obesity in Brazil started to rise. In short, he found statistical evidence indicating, basically, that as as people stopped cooking they became obese. In the book, he observes that connection was so strong that if the one single thing you knew about a person was that they had bought sugar, it predicted that they would not be obese because sugar, while unhealthy, is usually used to cook with. Thus the emphasis on “traditional culinary uses”.
Cooking has a lot of virtues, but evidence that the ingredients themselves are the issue came when an NIH researcher, Kevin Hall, did an experiment where he had people live in an NIH facility and half ate “processed” food (in the NOVA system) and half ate “ultra-processed” food with equivalent nutritional content, and the UPF group gained weight very quickly relative to the “processed” group. The Hall excitement is when Monteiro’s theory really started getting traction.