Key Takeaways
- A study of 461,682 US adults found no statistically significant association between total ultra-processed food consumption and colorectal cancer incidence.
- During a median follow-up of 18.8 years, researchers documented 10,075 colorectal cancer cases.
- Higher consumption of certain ultra-processed meat-containing products was associated with greater colorectal cancer incidence.
- Some ultra-processed food categories, including bread, breakfast cereals, and dairy-based products, were associated with lower incidence.
- The findings indicate that different ultra-processed foods may have different relationships with colorectal cancer, but they do not establish that individual foods cause or prevent the disease.
Introduction
A large study of 461,682 adults found that associations with colorectal cancer varied across ultra-processed food categories, with no significant association for total intake.
Ultra-processed foods are often discussed as a potential contributor to colorectal cancer. However, a large US study suggests that grouping these foods together may conceal important differences in their nutritional composition and associations with cancer incidence.
Published October 15, 2026, in PLOS Medicine, the study analyzed dietary information from 461,682 older adults and found no statistically significant association between total ultra-processed food intake and colorectal cancer incidence.
The findings differed when researchers examined specific food categories. Some meat-containing products were associated with higher incidence, while certain categories containing whole grains or dairy were associated with lower incidence.
The results raise a question for dietary research: How much can food processing classifications tell us about cancer risk without also considering the foods’ nutritional characteristics?
What the Study Examined
Researchers from the US National Cancer Institute and collaborating institutions analyzed data from the NIH-AARP Diet and Health Study.
The analysis included 461,682 adults aged 50 to 71 who were free of cancer at enrollment in 1995–1996.
Participants completed food-frequency questionnaires describing their usual diets. Researchers classified foods according to the NOVA system, which groups foods by their degree and purpose of industrial processing.
For the primary analysis, ultra-processed food consumption was measured in grams per 1,000 calories consumed. Participants were divided into five intake groups, separately for men and women.
The team also examined dietary components previously associated with colorectal cancer, including meat, whole grains, fiber, dairy, and calcium, distinguishing between ultra-processed and non-ultra-processed sources.
Participants were followed through 2018 using cancer registry information.
Over a median follow-up of 18.8 years, researchers documented 10,075 colorectal adenocarcinoma cases.
Statistical models accounted for several demographic, lifestyle, and health-related factors, including age, sex, smoking, physical activity, alcohol consumption, education, race and ethnicity, family history of cancer, and self-reported health status.
The researchers aimed to determine whether colorectal cancer incidence differed according to total ultra-processed food intake and whether specific food groups and dietary components showed different associations (Abar et al., 2026).
What Researchers Found
Overall Ultra-Processed Food Intake Was Not Significantly Associated With Colorectal Cancer Incidence
Ultra-processed foods accounted for a median of 43% of participants’ daily energy intake.
When researchers compared participants in the highest and lowest fifths of ultra-processed food consumption, they found no statistically significant association with colorectal cancer incidence.
The adjusted hazard ratio was 0.97, with a 95% confidence interval of 0.91 to 1.03.
A hazard ratio of 1.00 represents an equal estimated hazard between the groups being compared. The confidence interval in this analysis included 1.00, meaning the findings were compatible with a modest decrease or increase in relative hazard.
The researchers also found no statistically significant overall association when examining cancers of the proximal colon, distal colon, and rectum separately.
These results do not prove that ultra-processed foods are harmless or that every food in this category has the same relationship with cancer.
Instead, they indicate that total ultra-processed food intake, considered as one broad category, was not significantly associated with colorectal cancer incidence in this particular cohort.
Meat-Containing Products Showed a Different Pattern
When the researchers examined individual ultra-processed food categories, higher consumption of reconstituted meat and fish products was associated with greater colorectal cancer incidence.
Participants in the highest consumption group had a 10% higher adjusted hazard than those in the lowest group.
The adjusted hazard ratio was 1.10, with a 95% confidence interval of 1.04 to 1.18.
In a separate analysis of dietary components, each additional daily serving of meat from ultra-processed sources was associated with an 8% higher adjusted hazard of colorectal cancer.
Meat from non-ultra-processed sources also showed a positive association, although the estimated increase was smaller.
Importantly, this dietary-component analysis examined meat intake according to its processing classification. It should not be interpreted as an estimate for processed meat alone.
The findings are broadly consistent with established evidence linking processed meat consumption to colorectal cancer. However, because the study was observational, it cannot establish that the foods examined caused the additional cancer cases.
Some Ultra-Processed Foods Were Associated With Lower Cancer Incidence
Several ultra-processed food categories showed associations in the opposite direction.
Compared with participants in the lowest consumption groups, those in the highest groups had lower adjusted colorectal cancer hazards for:
| Ultra-processed food category | Adjusted hazard ratio | Relative difference |
|---|---|---|
| Bread | 0.91 | 9% lower |
| Breakfast cereals | 0.84 | 16% lower |
| Frozen desserts and milk drinks | 0.82 | 18% lower |
These results should not be interpreted as evidence that individual products within these categories prevent colorectal cancer.
For example, the category labeled “frozen desserts and milk drinks” included foods with potentially different nutritional profiles. The observed association does not establish that eating ice cream or other sugary desserts reduces cancer risk.
Similarly, the results for bread and breakfast cereals do not mean that every product in those categories is equally beneficial.
The associations may reflect differences in whole-grain content, fiber, dairy, calcium, overall dietary patterns, or other characteristics that the study could not fully separate.
The researchers also examined dietary components from ultra-processed and non-ultra-processed sources.
Higher intake of calcium and dairy from both sources was associated with lower colorectal cancer incidence.
Whole grains and fiber from ultra-processed sources were also inversely associated with incidence in the mutually adjusted models.
These food-category and nutrient analyses were secondary investigations involving multiple comparisons. Their findings require confirmation in other populations.
The results indicate that nutritional composition may matter when examining the relationship between food processing and colorectal cancer, but they cannot establish whether these foods directly influenced cancer development.
How the Findings Compare With Earlier Research
A 2022 Study Found Higher Colorectal Cancer Risk Among Men
The new findings differ from an influential study published in The BMJ.
Researchers analyzed three prospective US cohorts involving 206,248 men and women.
During 24 to 28 years of follow-up, they documented 3,216 colorectal cancer cases.
Among men, those consuming the most ultra-processed foods had a 29% higher adjusted hazard of colorectal cancer than those consuming the least.
The association was particularly pronounced for distal colon cancer, where the estimated hazard was 72% higher.
However, the researchers did not find a significant overall association between total ultra-processed food consumption and colorectal cancer among women (Wang et al., 2022).
The 2022 study also identified differences between food categories.
Certain ready-to-eat meat products and sugar-sweetened beverages were associated with greater colorectal cancer incidence among men. Yogurt and dairy-based desserts were associated with lower incidence among women.
These findings overlap with the newer research in an important way: both studies indicate that different ultra-processed food groups may have different relationships with colorectal cancer.
The studies are not directly interchangeable. They involved different populations, dietary assessments, exposure measures, and statistical analyses.
Why the Studies May Reach Different Conclusions
The 2026 study assessed ultra-processed food intake using grams per 1,000 calories and included older adults recruited in the mid-1990s.
The 2022 study assessed intake in servings per day and incorporated repeated dietary questionnaires collected over time.
Differences in how researchers measure food intake can influence which products contribute most heavily to the overall ultra-processed food category.
The nutritional composition of those products may also matter.
A packaged whole-grain bread, processed meat product, and sugar-sweetened beverage can all fall within the same broad processing classification while differing substantially in fiber, protein, saturated fat, sodium, and other components.
Consequently, a single measure of total ultra-processed food consumption may combine dietary exposures associated with higher and lower colorectal cancer incidence.
Neither study establishes that industrial processing itself is responsible for the observed associations.
What the Findings Mean for Colorectal Cancer Prevention
The findings do not suggest that people should disregard established dietary recommendations or increase their consumption of ultra-processed foods.
Instead, they show why the nutritional composition of individual foods matters when interpreting research on food processing and colorectal cancer.
The researchers found that some ultra-processed food categories were associated with higher colorectal cancer incidence, while others were associated with lower incidence.
For consumers, the most practical approach is to focus on overall dietary quality rather than assuming that every food classified as ultra-processed has the same health implications.
Established cancer-prevention guidance supports limiting processed meat and choosing foods that provide whole grains and dietary fiber.
The study also found inverse associations involving dairy and calcium intake, although these findings do not establish that increasing either nutrient will prevent cancer in an individual.
A packaged whole-grain bread and a processed meat product may both fall within the ultra-processed category, but they can differ substantially in their nutritional composition.
These findings should not be used to label individual products as cancer-preventing or cancer-causing.
Diet is also only one aspect of colorectal cancer prevention. Appropriate screening remains essential because it can identify precancerous growths and detect cancer at an earlier stage.
People should follow screening recommendations based on their age, medical history, family history, and individual risk factors.
Limitations
The study has several important limitations.
First, it was observational. Researchers could identify associations between dietary intake and colorectal cancer incidence but could not establish that particular foods caused or prevented cancer.
Second, dietary information was self-reported. Participants may have inaccurately recalled their usual food intake or the quantities they consumed.
Diet was assessed at enrollment, meaning the analysis could not fully account for changes in eating habits during the following decades.
The researchers also classified foods using historical dietary information. Some questionnaire items combined foods with different processing characteristics, creating the possibility of classification errors.
Although the statistical models accounted for several known risk factors, unmeasured or imperfectly measured characteristics could still have influenced the findings.
The researchers identified a statistical issue involving the assumption that relative hazards remained constant over time. They therefore examined results across different follow-up periods and found generally similar estimates. Nevertheless, this issue is relevant when interpreting the long-term results.
The study enrolled adults aged 50 to 71 during the 1990s. Their diets may differ from those of younger adults today, particularly because food manufacturing practices and available products have changed.
The findings should therefore not be assumed to apply equally to younger populations or current food products.
The secondary analyses of specific food groups and nutrients also involved multiple statistical comparisons, increasing the possibility that some associations arose by chance.
Finally, the study examined colorectal cancer incidence rather than the biological mechanisms through which specific foods or dietary components might influence cancer development.
Final Thoughts
A study of more than 461,000 US adults found no significant association between total ultra-processed food consumption and colorectal cancer incidence.
However, the results differed across food categories.
Certain meat-containing products were associated with higher colorectal cancer incidence, while some foods providing whole grains, fiber, dairy, or calcium were associated with lower incidence.
Earlier research has reported an overall positive association between ultra-processed food consumption and colorectal cancer among men, showing that the evidence is not entirely consistent.
The new study adds an important distinction: grouping all ultra-processed foods together may obscure meaningful differences between the foods people consume.
For readers, the practical message is not that ultra-processed foods are universally safe or dangerous. It is that dietary quality, food composition, and established cancer-prevention recommendations remain important when evaluating long-term health.
Funding and Disclosures
The 2026 study was funded by the Intramural Research Program of the National Cancer Institute, part of the US National Institutes of Health.
Several authors were current or former employees of the funding institution. The researchers reported that the funder had no role in study design, data collection and analysis, publication decisions, or manuscript preparation.
The authors declared no competing interests.
References
Abar, L., O’Connell, C. P., Hong, H. G., Herrick, K. A., Kahle, L., Lerman, J. L., Liao, L. M., Zhang, X., Zhang, X., Zhao, L., Zouiouich, S., Sinha, R., Khandpur, N., Martínez Steele, E., & Loftfield, E. (2026). Associations of food processing and dietary risk factors with colorectal cancer incidence in the NIH-AARP Diet and Health Study: A cohort study. PLOS Medicine, 23(10), e1005251. https://doi.org/10.1371/journal.pmed.1005251
Wang, L., Du, M., Wang, K., Khandpur, N., Rossato, S. L., Drouin-Chartier, J.-P., Martínez Steele, E., Giovannucci, E., Song, M., & Zhang, F. F. (2022). Association of ultra-processed food consumption with colorectal cancer risk among men and women: Results from three prospective US cohort studies. BMJ, 378, e068921. https://doi.org/10.1136/bmj-2021-068921
