Study links ultra-processed food consumption to 57,000 annual deaths in Brazil

A team of researchers from São Paulo's universities and Chile's Universidad de Santiago published findings this year that reframed how Brazil thinks about food and death. Working across USP, Fiocruz, and Unifesp, they calculated something no one had quantified before: how many Brazilians die prematurely from eating ultra-processed foods. The answer was 57,000 people a year.

To understand what that number means, consider what else kills Brazilians. In 2019, the year the researchers used for their analysis, homicides claimed 45,500 lives. Lung cancer killed 28,600. Breast cancer killed 18,000. The deaths attributable to ultra-processed food consumption—candy, instant noodles, frozen pizza, packaged snacks, soda, processed meats—exceeded the murder rate and surpassed the combined toll of the country's two deadliest cancers. The study appeared in the American Journal of Preventive Medicine, and it was the first of its kind to attach a mortality number to this category of food.

Ultra-processed foods are industrial formulations made from food fragments and laboratory-synthesized additives: colorants, preservatives, flavorings. The research linking them to weight gain, diabetes, cardiovascular disease, and cancer had been accumulating for years. But previous studies had focused on specific nutrients—sodium, saturated fat, sugar. Eduardo Nilson, a researcher at USP's nutrition and health research center and one of the study's authors, explained why the team took a different approach. The problem isn't just the nutrients. Ultra-processed foods affect the body through multiple pathways: they alter how nutrients are absorbed, trigger inflammatory responses, change the composition of gut bacteria, and introduce chemical contaminants through manufacturing and packaging. Because the mechanisms are so varied, the researchers modeled deaths from all causes in the 30-to-69 age group, not just specific diseases. This approach captured the full weight of the risk factor.

The calculation began with data from Brazil's household budget survey, which showed that Brazilians derive 19.7 percent of their daily calories from ultra-processed foods—roughly 400 calories for someone eating 2,000 a day, equivalent to a single package of instant noodles. The researchers combined this consumption data with 2019 mortality statistics and relative risk estimates drawn from a comprehensive review of scientific literature. Of the 541,100 premature deaths recorded that year in Brazil, 57,000—or 10.5 percent—were associated with ultra-processed food consumption. Sixty percent of these deaths occurred among men. Sixty-eight percent occurred in people between 50 and 69 years old.

The model also projected what could be prevented. If Brazilians reduced their ultra-processed food intake by 10 percent, 5,900 deaths could be avoided annually. A 20 percent reduction would save 12,000 lives. A 50 percent reduction would prevent 29,300 deaths. Nilson noted something striking about these scenarios: Brazil had already lived them. A 20 percent reduction would mean returning to consumption levels from a decade earlier. A 50 percent reduction would match the 1990s. The country had proven it was possible to eat differently. Instead, consumption kept climbing.

The reason was economic. Ultra-processed foods were becoming cheaper while fresh produce grew more expensive—a cruel inversion that hit the poorest Brazilians hardest. For the first time in 2022, ultra-processed foods cost less than fresh food on average. Nilson and his colleagues also modeled what would happen if Brazil's consumption patterns shifted toward those of wealthier nations. If Brazil matched Mexico's ultra-processed food share of 29.8 percent of calories, deaths would nearly double to 113,000 annually. If consumption reached United States levels—57 percent of calories—Brazil could face 194,000 deaths a year from this single dietary factor.

The researchers called for urgent public policy intervention. They pointed to measures that had worked elsewhere: subsidies for healthy foods, institutional purchasing programs, support for family farms, restrictions on advertising ultra-processed foods to children, taxation of these products, and clearer warning labels. Brazil had begun implementing a new labeling system in October, but Nilson and others argued it was weaker than evidence suggested it should be. The system used magnifying glasses to flag excess salt, sugar, and saturated fat, but experts had advocated for the alert-figure model used successfully in Chile, which research showed was more effective at changing behavior. The criteria for flagging products were also less stringent than international health organizations recommended, meaning fewer items would receive warnings. Implementation would be gradual too—only new products would carry the labels immediately.

Nilson saw the research as a foundation for future work. The model could be refined to account for time lags between dietary changes and mortality shifts. It could estimate the impact of specific policies, calculate the burden on Brazil's public health system, and quantify economic losses from premature death and illness. He drew a parallel to tobacco. Decades of research establishing clear risk factors and attributable deaths had built the case for regulation. Ultra-processed foods, he suggested, deserved the same rigorous attention. The data now existed. What remained was the will to act on it.