Prostate cancer is the most commonly diagnosed cancer in American men. Ultra-processed food makes up more than half the calories in the average American diet. So a study connecting the two got a great deal of attention.
Here is the finding, and here is what its design permits you to conclude.
The study
Published in The American Journal of Medicine in 2026, the analysis used a nationally representative sample of more than 17,000 US men.
After adjusting for age, smoking, race and ethnicity, and poverty status, men eating the most ultra-processed food had roughly 24% to 34% higher prostate cancer risk, depending on how the consumption groups were divided. The most commonly cited figure is about 30%.
The authors state plainly that the study is observational and cannot prove that ultra-processed foods caused the cancers.
What “ultra-processed” actually means
The term comes from the NOVA classification, and it is about industrial formulation rather than nutrition. Group 4 covers foods made largely from substances extracted from foods — protein isolates, modified starches, hydrogenated oils — plus additives such as emulsifiers, artificial colours and flavour enhancers.
The practical list: soft drinks, packaged snacks, most breakfast cereals, processed meats, industrial breads, reconstituted meat products, instant noodles, many “health” bars.
The category is also genuinely criticised for being too broad. Packaged wholegrain bread and a can of soda both land in Group 4 despite behaving very differently in the body. That imprecision is a real weakness in this entire field, not just this study (what the research actually shows).
Why observational is the crux
This is the part worth understanding, because it applies to most nutrition headlines you will ever read.
Confounding. Men who eat the most ultra-processed food differ systematically from those who eat the least — in income, education, physical activity, alcohol, obesity, and access to healthcare. The study adjusted for several of these. Adjustment reduces confounding; it does not remove it, and unmeasured factors remain unmeasured.
Detection bias, which matters specifically here. Prostate cancer is largely found through PSA screening. Men who see doctors more often get screened more often and get diagnosed more often. If dietary patterns correlate with healthcare access — and they do — some of this association could reflect who gets tested rather than who gets cancer. This is a well-known problem in prostate cancer epidemiology.
Dietary recall is imprecise. These studies typically depend on people remembering and reporting what they ate, which is unreliable in known and systematic ways.
Incidence is not mortality. Many prostate cancers are indolent and never cause harm. A study of diagnoses does not tell you about the aggressive disease that actually matters.
What would make it stronger
Plausible mechanisms exist: ultra-processed foods drive obesity and insulin resistance, both linked to cancer risk; some emulsifiers affect the gut barrier and inflammation in animal models; and certain packaging compounds are endocrine-active.
Plausibility supports a hypothesis. It does not confirm one, and nutrition has a long record of mechanistically plausible findings that failed in trials.
The honest position: this is a consistent, biologically plausible association from a good sample, with real limitations that cannot be resolved by this design. It adds to a body of evidence linking ultra-processed food to several cancers and to cardiovascular disease. It does not establish that a soda causes prostate cancer.
Why the advice does not change much anyway
Here is the practical resolution. The recommendation that follows from this study — eat less ultra-processed food — is already supported by stronger evidence for other outcomes.
A tightly controlled inpatient trial found people ate around 500 more calories per day on an ultra-processed diet than on a matched unprocessed one, with the same nutrients available. That is a randomised finding about eating behaviour, not an association.
And the reduction overlaps almost entirely with what already has the best evidence: more fiber, more whole foods, less liquid sugar (what to cut, and the pattern with the strongest trial evidence).
You do not need this study to justify eating less packaged food. That is exactly why the weakness of its design is not a reason to dismiss the advice — nor a reason to overstate the finding.
When to see a doctor
Discuss PSA screening with your doctor starting at age 50, or 45 with a family history or if you are African American, where risk is higher and disease tends to present earlier. Screening is a genuine shared decision, since it carries a real risk of overdiagnosis and treatment of cancers that would never have caused harm.
See a doctor promptly for difficulty starting or stopping urination, a weak stream, blood in urine or semen, new erectile difficulty, or persistent pain in the hips, back or pelvis.
The honest bottom line
A nationally representative study of more than 17,000 men found roughly 30% higher prostate cancer risk among those eating the most ultra-processed food. It is observational, it cannot exclude confounding, and prostate cancer is unusually vulnerable to detection bias because diagnosis depends on who gets screened. Treat it as one more brick in a wall rather than a demonstrated cause — and note that the advice it implies is already better supported by other evidence.
Frequently asked questions
Do ultra-processed foods cause prostate cancer?
This study cannot establish that. It found an association — roughly 30% higher risk among the highest consumers in a sample of more than 17,000 US men — but it is observational. Men who eat more ultra-processed food differ in income, activity, obesity and healthcare access, and prostate cancer diagnosis depends heavily on PSA screening, so some of the association may reflect who gets tested. The authors themselves state it cannot prove causation.
What counts as an ultra-processed food?
Under the NOVA classification, Group 4 covers foods made largely from substances extracted from foods — protein isolates, modified starches, hydrogenated oils — plus additives like emulsifiers, artificial colours and flavour enhancers. In practice: soft drinks, packaged snacks, most breakfast cereals, processed meats, industrial breads and many health bars. The category is criticised for being too broad, since packaged wholegrain bread and soda both fall into it.
When should men start PSA screening?
Generally discussed from age 50, or from 45 for men with a family history of prostate cancer or who are African American, where risk is higher and disease tends to appear earlier. It is genuinely a shared decision with your doctor rather than an automatic test, because PSA screening carries a real risk of overdiagnosis — detecting cancers that would never have caused harm but that lead to treatment with lasting side effects.