Ultra-processed food showed no overall link to colon cancer, but its parts pulled in opposite directions

If ultra-processed food raises colorectal cancer risk, a study of 461,682 people followed for a median of 18.8 years should be a good place to see it. Overall, it didn't show up. Among US adults aged 50 to 71 in the NIH-AARP study, the fifth who ate the most ultra-processed food had about the same rate of colorectal cancer as the fifth who ate the least, across 10,075 cases.

The label covers a lot of ground, from breakfast cereal to hot dogs. When the researchers split it apart, the pieces went different ways. Reconstituted meat and fish products were linked to about 10% higher risk. Breakfast cereals were linked to 16% lower risk, and bread to 9% lower. Fiber and calcium that came from ultra-processed sources also tracked with lower risk: 28% lower for every extra 10 grams of fiber a day and 23% lower for every 300 milligrams of calcium.

Overall diet quality mattered too. Compared with people who ate well and ate little ultra-processed food, people with poor diets had 22% higher risk in the top ultra-processed group and 28% higher risk in the second-lowest group. Poor diet quality carried extra risk even when processed intake was low.

There are real gaps. Diet was self-reported once, in 1995 and 1996, by mostly white, older Americans, and today's food supply is different. Labeling whole foods as ultra-processed can misclassify products, and additives couldn't be studied. With so many secondary comparisons, some of these splits could be chance. The results point toward what is in the food rather than how it was made, but an observational study like this can raise that idea without settling it.