2026 · PLOS Medicine
Associations of food processing and dietary risk factors with colorectal cancer incidence in the NIH-AARP Diet and Health Study: A cohort study
Verdict
MIXED
The study only partly follows the scientific method. There is no theoretical framework to inherit.
1·Study methodology · no theoretical framework
MIXED
Partly follows the scientific method. No theoretical framework attached.
2·Paradigm · Basic Epidemiological Comparison
NOT APPLICABLE
No theoretical framework. The study verdict decides.
How the verdict is decided
This study does not operate inside a theoretical framework. Using hospital records, billing codes, or diagnostic names as labels on visible events is not the same as depending on that framework. The verdict is the study methodology score alone.
1·Study methodology
Did this study test its claim with methods that are independent, falsifiable, and non-circular?
MIXED
This study claims that eating ultra-processed food overall is not linked to colorectal cancer, because processed grains and dairy lower the risk while processed meat raises it. The tests split.
Independently Verifiable
Pass
Checking boxes on a food survey and counting colon cancer diagnoses in state registries are records of foods eaten and tumors found, which require no theoretical model to observe.
Falsifiable
Fail
The authors credit their null finding to opposing effects of grain and meat subgroups and warn of residual confounding and questionnaire error, so any overall link or lack of link is explained away in advance.
Non-Circular
Pass
Sorting questionnaire food items by factory processing and counting state cancer registry diagnoses does not use the cancer outcome to build the food groups.
Why
Nearly half a million older adults filled out a 124-item food questionnaire once in the mid-1990s. The authors linked those checked boxes to government food databases, sorted the items into factory-processed and non-processed buckets, and tracked the people in state cancer registries through 2018 until 10,075 colon and rectal cancers were recorded. Both the survey answers and the tumor diagnoses are visible records, and the food buckets were not built from the cancer counts. The study fails on falsifiability. When total ultra-processed food showed no link to cancer after statistical adjustments, the authors explained the null away by splitting the food bucket into grains and dairy that lowered risk and meats that raised it. With residual confounding and questionnaire error also waiting in the limitations, no number from this regression could lose.
2·Paradigm · Basic Epidemiological Comparison
The three tests do not apply. There is no theoretical framework.
NOT APPLICABLE
Why
This study compares groups defined by food questionnaire answers against tumor diagnoses recorded in state cancer registries. Nutrient and food-group estimates calculated from food composition tables are physical estimates of what people ate, and registry histology codes sit on physical tumors found in patients. Even the diet quality index is just a score of how closely a person's food list matches federal advice, used here as a statistical control rather than a disease construct. Because neither the food groupings nor the cancer counts depend on an unobservable entity or a theoretical model of the body to exist, the study is a basic epidemiological comparison and no paradigm score applies.
From the study
UPF intake was not associated with CRC incidence (hazard ratio 5th versus 1st quintile 0.97; 95% confidence interval [0.91–1.03]; p trend = 0.55). Whole grain, fiber, dairy, and calcium intakes from UPF sources were inversely but meat intake from UPF sources was positively associated with CRC incidence.
The main limitations of the study relate to its observational design and include potential residual confounding and exposure misclassification owing to systematic or random error in self-reported dietary data.
Dietary intake was assessed using a food frequency questionnaire (FFQ) that included 124 food and beverage items, with portion sizes and 21 questions on low-fat, high-fiber foods and food preparation methods.
Incident cancer cases were identified through cancer registry linkage in study states and three states popular for relocation (Arizona, Nevada, and Texas).