A large null finding sharpens the food question

# A large null finding sharpens the food question OCTOBER 16, 2026 People who ate the most ultra-processed food were no more likely to develop colorectal cancer than people who ate the least, according to a large US study published in PLOS Medicine in 2026. The study followed participants from 1995-96 through 2018. Useful evidence includes results that complicate a familiar story. This large cohort found no overall association between ultra-processed food intake and colorectal cancer, within the study's population and methods. This was an observational study in people. It cannot prove cause and effect. Researchers tracked 461,682 US adults who were aged 50 to 71 and cancer-free when they joined the NIH-AARP Diet and Health Study in 1995-96. They followed the group for cancer diagnoses through 2018. In that time, 10,075 people were diagnosed with colorectal adenocarcinoma, the most common type of colon and rectal cancer. At the start, participants filled out a food frequency questionnaire. It asks how often a person eats a long list of foods. The researchers sorted each food into the Nova system, which ranks foods by how much industrial processing they undergo. Ultra-processed foods sit at the top of that scale. These foods were a big part of the diet. The median intake was 293 grams per 1,000 calories a day. That was 43% of daily energy. Across the US, the authors note, ultra-processed foods make up more than half of adults' energy intake. The team split participants into five equal groups by how much ultra-processed food they ate, separately for men and women. They compared the top group with the bottom group. They adjusted for age, sex, race and ethnicity, education, physical activity, smoking and alcohol use, among other factors. The result was flat. The hazard ratio, a measure of relative risk over time, was 0.97 for the highest group versus the lowest. The 95% confidence interval ran from 0.91 to 1.03, which includes no difference. The trend across all five groups was also not significant (p trend = 0.55). The more useful findings came from looking inside the category. Whole grain, fiber, dairy and calcium that came from ultra-processed foods were each linked to lower colorectal cancer incidence. Meat that came from ultra-processed foods was linked to higher incidence. The authors suggest this may explain the flat overall result. Some ultra-processed foods carry components tied to lower risk, such as whole grains or dairy products. Others carry components tied to higher risk, such as processed meat. In the total, the opposing links may partly cancel out. The authors say this may also help explain why earlier studies on ultra-processed food and colorectal cancer have disagreed. FOR YOU For personal health, the practical message is about ingredients more than labels. The authors say their findings support existing advice to eat whole grains, fiber, dairy and calcium and to avoid processed meat to help prevent colorectal cancer. A packaged food that is high in fiber or whole grain is not the same as a packaged meat product, even if both are called ultra-processed. Talk with your clinician before making major dietary changes or starting supplements. The study has limits, which the authors state. Because it is observational, other factors the researchers could not fully account for may have influenced the results. Diet was self-reported, so some people's intake may have been recorded inaccurately, whether by chance or in a consistent direction. The findings apply to older US adults, not necessarily to younger people or other countries. Nothing here says ultra-processed food is healthy overall. The study looked at one disease, colorectal cancer. Other health outcomes were not part of this analysis. Watch for other large cohorts to break down ultra-processed food by its components in the same way, and to test whether the pattern holds in younger and more varied groups. The result belongs alongside the positive findings, with its own limitations visible. A better account of food and health will need room for both. Sources: PLOS Medicine (2026), PLOS Medicine (via Medical Xpress) PLOS Medicine (2026),: https://doi.org/10.1371/journal.pmed.1005251 PLOS Medicine (via Medical Xpress): https://medicalxpress.com/news/2026-10-total-ultra-food-intake-linked.html ---