From scarcity to ultra-processed foods: a conceptual analysis of how their properties interact with characteristics of food intake regulation mechanism.

Since the mid-twentieth century, there has been a rapid rise in the production and consumption of ultra-processed foods (UPFs) so that they now contribute up to 60% of dietary energy intake in several Western countries. Existing accounts emphasize economic, political, food-system, and socio-cultural drivers of this shift. In this conceptual article, we discuss whether from a historical perspective, UPFs are a unique category of foods or rather an intensified extension of a long-standing trajectory of humans using various methods, such as mechanical processing, fermentation and cooking, to improve the digestibility, energy availability and palatability of their foods. Hereto we identify the typical properties used to explain the appeal of UPFs, distinguishing (i) direct intake drivers: engineered palatability, food matrices that enable high eating rates, high energy density and high sensory variety and (ii) market-mediated product attributes: convenient and portable formats, low effective cost, branded and marketed, ubiquitous availability, and long shelf life. We show that for none of these, UPFs form a clearly distinct grouping, but that what sets them apart is that on average they score higher on these properties, and they combine more of them at the same time. We then propose a conceptual framework to assess whether UPFs are unique in the way that they appeal to our evolved food intake regulation system. Hereto we map the properties onto (quantifiable) characteristics of the food intake regulation system in a testable framework. We conclude that a property-based assessment of UPFs suggests them to be an extension, rather than a categorical break from historical traditions, and that understanding the contemporary prevalence of UPFs requires explicit consideration of their interaction with evolved food intake regulation mechanisms, alongside established structural explanations, and we outline implications for future empirical tests and multi-level interventions.