New research published in JAMA Oncology finds an association between higher intake of ultra-processed foods and an increased likelihood of precancerous colorectal adenomas in women under 50. The analysis followed 29,105 nurses who underwent at least one colonoscopy and were tracked from June 1991 through June 2015, with diet questionnaires every four years. Women averaging three servings of ultra-processed foods per day had about a 3% chance of a precancerous polyp, while those consuming 10 or more servings daily showed roughly a 5% chance. The study did not find a similar link with serrated lesions.
Key takeaways
- The cohort included 29,105 nurses under age 50 who had at least one colonoscopy; follow-up spanned June 1991–June 2015 with diet surveys every four years.
- Women averaging ~3 ultra-processed servings per day had an estimated 3% risk of a precancerous polyp versus ~5% for those at 10+ servings per day.
- The increased association was observed for adenomas (precancerous polyps); no clear association emerged for serrated lesions.
- Only about 5% of adenomas are cancerous at detection, but roughly 75% of colorectal cancers originate from these growths (Cleveland Clinic figure).
- Higher ultra-processed food intake correlated with higher BMI, greater likelihood of past smoking, lower physical activity and reduced intake of fiber and vitamin D among participants.
- Ultra-processed foods represent more than half of the average U.S. calorie intake, according to the CDC, and have been linked in other studies to diabetes, cardiovascular risk, sleep problems and higher mortality.
Background
Interest in the role of diet in colorectal cancer risk has grown as incidence among younger adults has risen in recent decades. Public-health researchers and clinicians have tracked dietary patterns alongside screening data to try to separate lifestyle drivers from genetic and screening-related factors. Ultra-processed foods—broadly defined by public-health researchers as industrial formulations with many ingredients and additives—have become a dominant part of the U.S. food supply, prompting questions about long-term effects.
Past observational studies have connected ultra-processed diets to metabolic outcomes such as obesity and type 2 diabetes; mechanistic work has suggested impacts on gut bacteria and inflammation. Screening-detected lesions, including adenomas and serrated lesions, are important precursors in colorectal carcinogenesis, and understanding lifestyle correlates of each lesion type helps shape prevention guidance. Stakeholders include clinicians, public-health agencies, food policy makers and patients seeking practical dietary advice.
Main event
The new study, released in JAMA Oncology, examined dietary questionnaires completed every four years by participating nurses, matching reported intake to subsequent colonoscopy findings. Investigators categorized products as ultra-processed based on ingredient profiles commonly used in nutritional epidemiology and then estimated daily servings. Analyses adjusted for multiple covariates including age, BMI, smoking history, physical activity and nutrient intakes such as fiber and vitamin D.
Results showed a graded relationship: modest intake levels (around three servings daily) were associated with a baseline roughly 3% probability of finding a precancerous adenoma, while very high intake (10+ servings daily) aligned with an approximate 5% probability. The team did not detect a statistically significant relationship between ultra-processed intake and serrated lesions in this cohort.
Authors noted that participants who consumed more ultra-processed items also tended to have other health-related risk factors—higher body mass index, more frequent past smoking, lower activity and lower dietary fiber—factors that can confound or mediate observed associations. The paper stresses observational design limits and calls for further mechanistic and interventional work to test causality.
Analysis & implications
These findings add to a growing body of observational evidence linking highly processed industrial food products to adverse health outcomes. Because the study is observational, it cannot prove causation, but the dose‑response pattern strengthens the argument that greater exposure may elevate risk. The approximately two-percentage-point difference between low and very high consumers represents a modest absolute increase for an individual woman but could translate into substantial population-level impact given the prevalence of ultra-processed foods.
Public-health implications include potential reinforcement of dietary guidance that favors whole foods and limits packaged, ingredient-heavy products. If mechanisms suggested by experimental work—microbiome alteration, increased gut permeability and chronic inflammation—are validated, targeted interventions (dietary counseling, reformulation, labeling or policy changes) could meaningfully reduce incidence of colorectal precursors among younger adults.
Clinicians should view these results in the context of established screening recommendations and risk factors. For younger adults and clinicians assessing risk, counseling on diet quality, weight management and smoking cessation remains evidence-based practice. For policy makers, the study adds weight to ongoing discussions about defining and regulating ultra-processed foods at the federal level.
Comparison & data
| Measure | Low–moderate UPF intake (~3 servings/day) | High UPF intake (10+ servings/day) |
|---|---|---|
| Estimated risk of precancerous adenoma | ~3% | ~5% |
| Cohort size | 29,105 nurses under 50 with ≥1 colonoscopy (June 1991–June 2015) | |
The table summarizes the principal numeric comparisons reported in the paper. While relative differences appear meaningful, absolute risks in this screening-detected sample remain modest for individuals. Researchers adjusted for multiple confounders, but residual confounding—especially from correlated lifestyle factors—remains a concern in interpreting the magnitude of effect.
Reactions & quotes
Public-health and clinical reactions emphasized caution while acknowledging the study’s contribution to the literature on diet and colorectal risk. Experts highlighted the consistency of these findings with prior links between processed diets and metabolic disease.
You do see an increase in risk with more servings of ultra-processed food products.
Dr. Céline Gounder, CBS News medical contributor / KFF Health News
Dr. Gounder, who spoke to CBS News about the paper, framed the result as one piece in a larger puzzle about how processed diets affect gut health and cancer risk.
I don’t know if there are any good ultra-processed foods.
Dr. Céline Gounder, CBS News medical contributor / KFF Health News
That blunt assessment was offered as practical guidance: experts argue that reducing ultra-processed intake is a straightforward, low-risk intervention while causal mechanisms are further investigated.
A good rule of thumb is: if you can’t pronounce an ingredient on the package, it is probably an ultra-processed food.
Dr. Céline Gounder, CBS News medical contributor / KFF Health News
Practitioners and patient advocates have repeated similar heuristics while awaiting formal federal definitions and potential regulatory action.
Unconfirmed
- Whether the association represents direct causation: the study is observational and cannot definitively prove that ultra-processed foods cause adenomas rather than correlate with other risk factors.
- The precise biological pathway(s): microbiome changes, increased gut permeability and inflammation are plausible but not yet confirmed as direct mediators in humans.
- Whether specific additives, processing methods or food categories (beyond broad packaged foods and sweetened products) drive risk remains unresolved.
Bottom line
The JAMA Oncology analysis of 29,105 nurses suggests that frequent consumption of ultra-processed foods correlates with modestly higher detection of precancerous colorectal adenomas among women under 50. For individuals, the absolute risk differences reported are small, but widespread high intake of processed products means the population-level implications could be meaningful if the association reflects a causal effect.
Clinicians and public-health officials can consider these results when advising patients: promoting whole-food diets, maintaining healthy weight and following screening guidance remain central. Researchers should pursue mechanistic studies and randomized interventions to test whether reducing ultra-processed food intake lowers adenoma occurrence.