"Eat this, avoid that" headlines about food and cancer are not all equal. Some rest on decades of consistent research; others on a single preliminary study. Pick a real claim, then adjust the four things scientists actually weigh — and watch a claim slide from preliminary to strong. No single food causes or cures cancer; this is about reading evidence.
Evidence strength
preliminarylimitedprobablestrong
What the four grades mean: expert panels (such as the World Cancer Research Fund / AICR review and IARC) sort diet–cancer relationships into tiers — roughly strong/convincing, probable, limited-suggestive, and preliminary / no conclusion. They grade the evidence, not any one meal or person.
Pick a diet–cancer claimtap one below, then test its evidence
Grade it yourself
These four questions decide how strong a claim is. They start at the real research picture for the claim you chose — change any of them to see how the grade would shift.
Best study design availablehow the evidence was gathered
Consistency across studiesdo independent studies agree?
Dose–responsemore exposure, more effect?
Amount of evidencehow many studies, how many people
Reading evidence is not medical advice. Cancer risk reflects whole eating patterns over years, plus genetics, age, tobacco and alcohol, infections, activity, environment, and access to care — never one food on its own. "Strong" evidence describes a population average and a direction of risk, not a guarantee for any individual. For decisions about your own health, talk with your clinician.