Are ultra-processed products bad for your longevity?

Meat and fish products in processed form, such as sausages, deli meats and ready-made burgers, show the strongest association with premature death. People who eat a lot of these products have a 6 to 43 percent higher risk of dying, depending on the cause of death. Soft drinks (with or without sugar), dairy desserts and ultra-processed breakfast foods are linked to a slightly elevated mortality risk of 7 to 9 percent.

For premature death overall, the risk estimates found vary widely: from just 4 percent higher in a large study with more than 114,000 participants, to 21 percent higher in a summary of multiple studies. Per 50 grams more per day, that works out to roughly 2 percent additional risk. That sounds modest, but ultra-processed products are rarely eaten in small quantities. The association with cardiovascular mortality appears larger (50 percent higher risk), but the individual studies are of very low quality.

The best-supported association is with type 2 diabetes. For every step up in the amount of ultra-processed food consumed, the risk rises by around 12 percent. This is also the only area in which the evidence is rated as 'moderate', rather than 'low' or 'very low'. For colorectal cancer, a smaller but comparably reliable association of 4 percent per 10 percent more ultra-processed food in the diet has been found.

There are two crucial nuances. First: in a large American study with nearly 35 years of follow-up, no association with death from cancer or cardiovascular disease could be found when those causes of death were examined separately. Second: in people with good overall diet quality, the mortality risk largely disappeared. This suggests that the overall pattern of eating matters at least as much as how much you eat from a bag or a box.

The core scientific debate has not yet been settled: is the processing itself the problem, or are ultra-processed products simply a good marker for an unhealthy eating pattern in general? All available associations have been found in observational research, in which cause and effect cannot be proven conclusively. Notably, of 45 health outcomes studied, 71 percent showed an association with ultra-processed food, but for most of those associations the quality of the evidence is low to very low.