The degree of processing in food and beverage products is becoming an increasingly pressing matter for various stakeholders such as nutritionists, consumers, and policymakers. However, criticisms have been leveled at the most widely used system for assessing food processing, the NOVA classification system. Specifically, the NOVA 4 ultra-processed food category has been deemed too heterogeneous and qualitative in nature, with inconsistencies and ambiguities that limit research into the impact of processed food.
In response to these criticisms, researchers out of Massachusetts have developed a machine learning algorithm called FoodProX that relies on nutrients as input to accurately predict the degree of processing for any food in a reproducible, portable, and scalable fashion. The machine learning algorithm allows the researchers to define a continuous index (FPro) that captures the degree of processing of any food, ultimately helping to quantify the overall diet quality of individuals.
Using the algorithm, the researchers have calculated individual food processing scores (iFPro) for more than 20,000 individuals with dietary records in a representative US national sample from 1999-2006. The findings of the study reveal that individuals with a high food process score are positively associated with a risk of metabolic syndrome, diabetes, and a family history of heart attack or angina.
Despite the fact that ultra-processed breakfast cereals and refined flours are often fortified with nutrients like vitamin B12 and vitamin C, individuals who consume more extensively processed foods exhibited lower levels of vitamins in their bloodstream, according to the researchers. They believe that FoodProX can help address the limitations of the NOVA classification system and monitor the degree of an individual’s reliance on less or more processed food.
In summary, the researchers have developed a novel machine learning algorithm to accurately predict the degree of processing in any food product. The algorithm relies on nutrients as input and can be used to capture the overall diet quality of individuals and reveal the statistical correlations between the degree of processing and multiple disease phenotypes. Ultimately, this technology can help to extend and build upon the current NOVA classification system and provide valuable insights into the impact of processed food on human health.