Seed oils
Also called: vegetable oils, sunflower oil, rapeseed oil, canola, linoleic acid.
Last reviewed 2026-09-16. How this works.
Causing cancer Grade D
No human evidence of benefit, or none that survived testing.
The claim that seed oils cause cancer has been looked for in large studies and was not found.
People looked. The association the claim depends on is not there.
The claim
This claim runs the other way from most in the chapter. It is that seed oils cause cancer, or make an existing one worse, and it has spread very widely online.
What it does
The argument is usually made from the fatty acid composition of these oils and from inflammation pathways, and often from the observation that intake rose over the same century that cancer incidence rose. A shared trend is not a mechanism, and this is the correlation problem the book spends its first part on.
The evidence
This entry is not an absence of evidence. It is an accusation that was checked. Pooling prospective cohort studies, the relative risks for high compared with low linoleic acid intake were 1.05 for breast cancer, 0.92 for colon cancer and 0.83 for prostate cancer, every interval crossing one. The paper states plainly that none of the combined estimates indicated a significantly increased risk.1
A later pooled analysis of vegetable oil intake and breast cancer reported an odds ratio of 0.88 with no significant dose-response trend.2
The distinction this entry exists to make is the one the book draws in chapter 12. Absent evidence is usually an economic fact rather than a verdict: nobody funds a trial of a food nobody can patent. Seed oils are the mirror image, an accusation that was investigated and failed. Treating those two as the same thing is the error this book exists to correct, and it is easy to make in either direction.1
Three cautions belong on the page rather than buried in a note. The pooled cohort analysis is from 1998, and consumption patterns have changed a great deal since. Its own authors record that confounding cannot be excluded, because people who use more of these oils differ in other ways too. And the later meta-analysis was reached through a secondary source rather than read.12
One more caution, and it points at the sources that agree with this page rather than the ones that do not. Some of the recent review literature in this area carries titles that announce their conclusions. The book's own test for reading a source does not get suspended for a source arguing your side. That observation is the author's characterisation rather than a finding of any paper, and it is recorded as such.2
The studies this rests on
-
Pooled prospective cohort estimates for linoleic acid intake and cancer1
A meta-analysis, summarising observational studies, measured a clinical outcome.
no significantly increased risk for breast, colon or prostate cancer with high compared with low intake.
relative risk 1.05, interval 0.83 to 1.34. breast cancer; colon 0.92, interval 0.70 to 1.22; prostate 0.83, interval 0.56 to 1.24
What is absent
No randomised trial has assigned people to eat more or less seed oil and counted cancers, and none is likely to. What exists is observational, it is dated, and it does not support the claim.
Scope: seed oil intake and cancer incidence.
Established by published review, Zock and Katan 1998, American Journal of Clinical Nutrition, checked 2026-09-16. Result: no combined estimate from within-population studies indicated a significantly increased risk of cancer with high compared with low linoleic acid intake.1
What to ask
Is there anything I have read online about food and cancer that you would want to correct?
The dish-to-person gap
Seed oils is eaten as food rather than taken as a measured compound, and the register holds no human pharmacokinetic measurement for it.
When to call your team — no warning signs are recorded
If you have changed how you eat because of something you have read, or are thinking about it, the people treating you can only weigh that if they know about it. That is true when nothing has been written down here, and it is most true then.
The register records no warning signs for this entry. That is not a statement that nothing can go wrong.
Interactions
The register holds no measured interaction between this food and a cancer treatment class. An empty list means nobody has measured it, not that nothing would happen. Concentrated extracts are a different question from the food.
Nobody having looked is not the same as nothing being there. That is what this line says and all it says.
The whole interaction list, and how to turn it into a pharmacy review request.
Find a trial instead
This is a claim of harm from an ordinary food rather than a treatment to enrol in.
References
Zock PL, Katan MB. Linoleic acid intake and cancer risk: a review and meta-analysis. American Journal of Clinical Nutrition 1998;68(1):142-153.
doi:10.1093/ajcn/68.1.142 · PubMed 9665108
How this was reached: transcribed from chapter note, via docs/CHAPTER-NOTES.json, chapter 10 note 10.
A meta-analysis of vegetable oil intake and breast cancer risk. Asian Pacific Journal of Cancer Prevention 2015;16(12):5125-5135.
doi:10.7314/apjcp.2015.16.12.5125 · PubMed 26163654
How this was reached: transcribed from chapter note, via docs/CHAPTER-NOTES.json, chapter 10 note 11.