The omega 6 to omega 3 ratio is one of the most confidently repeated ideas in popular nutrition and one where the evidence is considerably weaker than the confidence.
What is the claim?
That ancestral diets contained roughly equal amounts of omega 6 and omega 3, while modern diets contain fifteen or twenty times more omega 6.
Since omega 6 derived signalling molecules are generally more inflammatory, the argument runs that this shift promotes chronic inflammatory disease.
The two families also compete for the same enzymes, so high omega 6 intake reduces conversion of ALA to EPA and DHA.
What is the problem with it?
The theory is plausible and the evidence connecting the ratio to disease outcomes is weak.
Observational studies of linoleic acid, the main dietary omega 6, have generally found neutral or favourable associations with cardiovascular disease rather than harm.
Trials replacing saturated fat with vegetable oils rich in omega 6 have generally shown neutral or beneficial effects on cardiovascular risk.
Does omega 6 cause inflammation?
The expected effect has been difficult to demonstrate. Studies increasing linoleic acid intake have generally not found increases in inflammatory markers.
Conversion of linoleic acid to arachidonic acid, the inflammatory precursor, is limited and tightly regulated, which may explain why the theory does not translate.
What does matter?
Absolute omega 3 intake appears more important than the ratio. Someone eating oily fish twice weekly has adequate EPA and DHA regardless of their omega 6 intake.
Improving the ratio by cutting omega 6 without increasing omega 3 does not raise EPA and DHA levels meaningfully.
Does the ratio matter for conversion?
- High omega 6 does reduce conversion of ALA to EPA and DHA
- This matters most for people relying entirely on plant sources
- It matters little for anyone eating fish or taking EPA and DHA directly
- Conversion is poor regardless, so the effect is on an already small number
What about seed oils?
Seed oils have become a target of considerable online criticism, largely on ratio grounds.
The evidence connecting them to harm is weak, and much of the concern extrapolates from mechanism to outcome without the outcome data supporting it.
Highly processed foods containing these oils have other characteristics that may explain observed associations.
What is the practical advice?
Focus on increasing omega 3 rather than obsessing about reducing omega 6. Eat oily fish, or supplement if you do not.
That achieves what the ratio argument is aiming at, without requiring a restructuring of your entire diet on contested grounds.
What is the summary?
An appealing theory that outran its evidence. Raising omega 3 is the useful action; the ratio follows from it without being the target.
This article is general information and not medical advice.
Where did the ancestral figure come from?
Estimates of ancestral diets are necessarily indirect, based on modelling and on studies of contemporary hunter gatherer populations.
They are reasonable estimates rather than measurements, and the confidence with which specific ratios are quoted exceeds what the evidence supports.
Does the ratio appear in guidance?
UK and European guidance does not set a target ratio, focusing instead on absolute intakes of omega 3 and on total fat quality.
That reflects the weakness of the outcome evidence for the ratio specifically.
What is the summary?
An appealing theory that outran its evidence. Raising omega 3 is the useful action, and the ratio improves as a consequence rather than being the target.
What should you actually change?
Eat oily fish twice weekly, or take a supplement if you do not. That single action addresses the underlying concern the ratio argument is pointing at.
Restructuring your entire fat intake on contested grounds is a great deal of effort for an uncertain return.
Where the ratio genuinely matters is for people relying entirely on plant ALA, since omega 6 competes for the enzymes that convert it.
For everyone else, the question is largely academic and the practical action is the same either way.