Ultra-processed foods and cancer risk explained: Study of 461,000 adults finds unexpected results

Not all ultra-processed foods showed the same association with colorectal cancer in a study of more than 461000 US adults. While researchers found no significant link between overall consumption and cancer incidence.

The study published in PLOS Medicine used data from NIH-AARP Diet and Health Study. Researchers followed 461682 adults aged 50 to 71 who enrolled in 1995 and 1996. 10075 participants developed colorectal adenocarcinoma, over a median of 18.8 years.

Ultra-processed foods made up a median of 43% of participants’ daily energy intake. Despite that people who consumed the most did not have a statistically significant difference in colorectal cancer incidence compared with those who consumed the least, as per the study.

Overall Finding Was Different From the Results for Individual Foods

So when researchers compared highest and lowest groups for total ultra-processed food intake, they found the hazard ratio was 0.97. And the 95% confidence interval ranged from 0.91 to 1.03. It indicates no statistically significant association.

The result remained broadly unchanged after the researchers accounted for factors such as diet quality and body mass index. They also found no association when they examined cancer by its location in the colon/ rectum.

But the analysis looked beyond total intake. When researchers separated foods into individual categories, the results varied.

Breakfast Cereals and Bread Were Associated With Lower Incidence

The study found that the people in highest intake group for ultra-processed bread had a 9% lower incidence of colorectal cancer than those in the lowest group. Breakfast cereals were associated with a 16% lower incidence, while frozen desserts and milk drinks were associated with an 18% lower incidence.

Reconstituted meat and fish products showed the opposite association. Participants with the highest intake had a 10% higher incidence than those with the lowest intake.

These figures do not establish that particular foods protect against cancer or cause it. As the study was observational it identified patterns in the data rather than proving that eating or avoiding a food directly changes cancer risk.

Meat, Whole Grains and Calcium Show Different Patterns

The researchers also examined nutrients and food components based on whether they came from ultra-processed/ non ultra-processed sources.

Higher intake of whole grains and fiber from ultra-processed sources was associated with lower colorectal cancer incidence. Dairy and calcium were also associated with lower incidence, regardless of whether they came from ultra-processed or non-ultra-processed foods.

Meat intake, meanwhile, was associated with higher incidence from both sources.

The findings point to differences within the broad ultra-processed food category. However, the study did not establish whether nutrients from foods with different levels of processing have equivalent effects on cancer risk.

Replacing Some Ultra-Processed Foods Was Linked to a Small Difference

Researchers also estimated what might be associated with replacing 10% of food grams from ultra-processed sources with the same proportion from minimally processed foods. The substitution was associated with a 2% lower incidence of colorectal cancer.

The Study Has Limitations

Participants reported what they ate at the beginning of the study. And researchers did not repeatedly measure their diets throughout the follow up period. The information was self-reported and questionnaire was designed to measure dietary intake rather than food processing. Some foods may therefore have been classified incorrectly.

The study also reflects a specific population, older US adults recruited in the mid-1990s, most of whom were white. The results may not apply to younger or more diverse groups, or fully reflect today's food supply. The researchers cautioned that findings from analyses of individual food groups should be interpreted carefully because multiple comparisons can produce chance associations.