Credible nutrition scientists have opposing views on whether omega 6 in vegetable oils is good for heart health. But who is right?
Part 1 of this series examined claims that omega 6 in vegetable oils causes ill health by increasing the risk for cancer, macular degeneration, Parkinson’s disease and inflammatory conditions, such as rheumatoid arthritis, ulcerative colitis and asthma. These claims, frequently expressed on the internet and in social media, were found to be baseless.
Part 2 will consider whether omega 6 increases or decreases the risk for coronary heart disease. Unlike the wild social media claims about omega 6, this issue is argued in the scientific literature with credible scientists on both sides. Whose argument is stronger?
The conventional wisdom
Leading heart health agencies such as the Heart Foundation, the American Heart Association and the World Health Organisation assert that polyunsaturated fats (mainly omega 6) are heart healthy and that they may be the ideal replacement for saturated and trans fats in the diet. The case appears to be strong:
• Polyunsaturated fat (mainly omega 6) has a beneficial effect on blood lipids. When saturated fat is replaced by polyunsaturated fat the level of serum LDL-cholesterol and the total cholesterol/HDL cholesterol ratio both fall, which is consistent with lower risk for coronary heart disease.
• Epidemiological studies indicate that polyunsaturated fat (mainly omega 6) is heart friendly. A pooled analysis of 11 prospective cohort studies conducted in Europe and the United States shows that replacing saturated fat with polyunsaturated fat lowers coronary risk more than other macronutrients.
• In randomised controlled trials coronary heart disease risk falls when saturated fat is replaced by polyunsaturated fat (mainly omega 6) – see Harvard’s meta-analysis.
So the conventional wisdom is based on three consistent lines of evidence with a rare meta-analysis of randomised controlled dietary trials sitting on top. Impressive. How could you argue against it?
Image: source
The alternative view
Although in the minority, credible scientists do argue against it. In relation to mechanism, it is argued that omega 6 antagonises omega 3, offsetting the heart health benefits of the latter. Omega 6 is also said to pro-inflammatory (though this is a dubious claim – see Part 1 of this series). These effects are presumed to overwhelm the beneficial effects of omega 6 on blood lipids.
In the absence of supporting epidemiological evidence, the argument that omega 6 has adverse effects on heart health really rests on a recent meta-analysis of randomised controlled trials that comes to different conclusions to the Harvard meta-analysis mentioned above.
Duelling meta-analyses
How can two meta-analyses of essentially the same data lead to differing conclusions?
Although both meta-analyses considered polyunsaturated fat-for-saturated fat exchanges, the recent meta-analysis sought to separate those trials that used pure omega 6 as a substitute for saturated fat from those that used mixed polyunsaturated fats i.e. omega 6 plus omega 3. But the trouble was there were few studies with pure omega 6 interventions, just the large Minnesota Coronary Survey (men and women) and the tiny Rose Corn Oil Trial. Undeterred, the researchers went looking for more pure omega 6 trials and identified the 40-year old Sydney Diet-Heart Study. The data for this study had to be completely re-analysed in order to be considered in the meta-analysis, an amazing effort in itself.
When the results of the mixed polyunsaturated fat studies were meta-analysed they showed a fall in heart disease risk, essentially the same result as that observed in the Harvard meta-analysis. But the effect of the pure omega 6 studies was a non-significant rise in heart disease risk which led the researchers to conclude that there was ‘a possibility of harm’ associated with omega 6.
A sceptical nutritionist’s view
The conclusion that omega 6 is possibly harmful essentially relies on two studies, both of which are problematic. The Sydney Diet-Heart Study did not employ a pure omega 6-for-saturated fat exchange as claimed. This old trial replaced saturated fat with Miracle polyunsaturated margarine, which in the early 1970s contained about 15% trans fatty acids. So, the intervention was the replacement of saturated fat with omega 6 and trans fats. Given the high risk associated with trans fats is it any wonder that heart disease risk did not fall? But why attribute the lack of benefit to the omega 6 and not to the trans fats?
The Minnesota study was a large well designed trial that was supposed to be conducted in institutionalised subjects over five years. However, a policy of de-institutionalisation brought the study unstuck. Rather than subjects receiving five years of dietary intervention, the average time on the trial was just over a year – too short to see any benefit from a lipid lowering trial of this size. So the observed lack of benefit in this trial is actually the expected result.
Randomised controlled dietary trials are notoriously difficult to conduct and all of the trials included in these meta-analyses have shortcomings, but basing a contrarian position on the non-significant findings of these two trials is drawing a rather long bow.
Following the publication of the recent meta-analysis several experts wrote to the relevant journal encouraging the authors to ‘consider all the evidence’ in coming to conclusions about omega 6. Good advice for us all.
Conclusions
At this stage the evidence that omega 6 in vegetable oils confers a heart health benefit when it replaces saturated fat in the diet is much stronger than the counter argument. However, if you are still sitting on the fence, it is worth remembering that no heart health agency actually recommends that pure omega 6 replace saturated fat. The advice is to replace saturated fat with polyunsaturated fats – both omega 6 and omega 3. Both duelling meta-analyses support this position so maybe the two sides of this argument are not as far apart as it seems.
Image: source
Bill,
Recently, the consensus of Internet nutritionists is that omega-6 is unhealthy in the amounts in which it is currently consumed. The claim that it’s inflammatory pops up, which you covered in part 1, but there’s also the claim that PUFAs are particularly unstable, and are readily oxidised in the body. Apparently this can lead to atherosclerosis or cancer, but most of the time the argument simply ends with the word “oxidised”. I’ve seen this claim appear an awful lot lately. In particular, the Internet consensus is that vegetable oils are undesirable, and should be replaced with butter or coconut oil.
Hi Shane. A couple of points here. The omega 6 content of the Australian diet has been steadily falling since about 1985. The last time I looked it was about 4-5% of daily calories, which is right at the lower end of recommended intakes, so we are not drowning in omega 6 as is frequently suggested.
There have been randomised controlled trials in humans lasting years in which omega 6 intakes of 14-15% were consumed – three times current intakes. The only consequence was a fall in coronary events. There is also consistent evidence from epidemiology, mechanistic studies and long-term feeding studies in primates which all indicates that omega 6 or polyunsaturated fats more generally are protective against heart disease. This is the evidence that the WHO, American Heart Association, NHMRC and the Heart Foundation take into consideration when they produce their dietary recommendations. A ‘consensus of internet nutritionists’ rates as the lowest form of scientific evidence available. Expert organisations wouldn’t even look at it.
In relation to cancer, the amount and type of fat that people eat has no effect on cancer risk. Regards, Bill
Of course. I didn’t mean to give the impression that I actually support these claims, but I do think it’s important to be aware of what the blogs are saying. And it looks like omega-6 is the next nutrient to be obsessed over and vilified.
It is a complete fabrication to say the Sydney Diet Heart study was confounded by trans-fats. Dr C Ramsden looked into this and concluded that the margarine they were given was most likely low in trans fats, as it was chosen to be high in PUFA.
It was more likely that the control group were consuming more trans fats. Do you know something about this study that Dr Ramsden was unable to find?
see http://www.bmj.com/content/346/bmj.e8707
Nevertheless, the question of how trans fats affected this clinical study is an unknown – however it was a far bigger problem in other clinical studies which your preferred meta-analysis relied up. Your Havard meta-analysis assumes all these studies can be summed up as replacements of saturated fat with polyunsaturated fats, yet they were almost all multi-intervention studies, with the Oslo diet study notably advising the intervention group to avoid stick-margarines, and being provided with Cod Liver Oil (high in Vitamin D and O3 DHA/EPA).
The Sydney Diet Heart Study is the stand out study of the crowd, with the most simple of substitutions applied, it should be the study you trust the most!
Using your interpretation of the SDHS, you imply that benefits associated with replacing SFA with Linoleic Acid were more than negated by increases in trans-fat consumption (unproven), and negated to near statistically significant amounts. Yet, you do not consider the implications that this would have for the Oslo diet heart study or others where we know trans-fat consumption was significantly decreased in the intervention group. To uphold the veracity of the Havard meta-analysis is to apply a shocking double standard here.
The SDHS is not an outlier either, it’s findings were similar to the Rose et al corn oil study, although this study was not as well designed.
Your other evidence is of course population studies, which we know are very confounded. For every population study that purports to find a correlation, there is another that contradicts it. In some ways the macro population figures can be a better guide, as they are not reliant on food questionnaires. If replacing saturated fat with Linoleic acid was beneficial, you would not expect French paradoxes or the Isreali paradoxes (http://www.ncbi.nlm.nih.gov/pubmed/8960090), yet the existence of these paradoxes show that our assumptions are very likely wrong!
Finally, relying on surrogate end-points to suggest Omega 6 is beneficial is putting your faith in the theory and nothing else. The theory can only be trusted as far as its ability to make predictions go, and this theory failed to meet expectations in the only study that really tested it.
Hello Gordon. The margarine used in the Sydney Heart Study was Miracle, produced by Unilever. I consulted to this company for 14 years and I can assure you that the Miracle margarine used in the Sydney Heart Study contained about 15% of its fatty acids as trans fats, similar to all margarines available at the time. The first trans-free margarine produced by Unilever (and the first in Australia) was Becel, but that came over a decade after the Sydney study.
So the intervention in this study was not the replacement of saturated fat with omega 6 as Ramsden asserts; it was the replacement of saturated fat with omega 6 + trans fatty acids. Ramsden argues that the adverse effects observed in this study were due to omega 6 whereas most other observers say that they were due to the trans fatty acids.
I put this directly to Ramsden when he presented this work at a conference in Sydney but he had no answers. Regards, Bill
Hi Bill,
Even though margarine trans fat contents are well below what they were, do you not think the small amounts they do still contain (e.g. Olivani, with the Heart Tick, contains 0.06g/10g serving) would add up over a life time of consuming that product? That, along with the taste, means I don’t touch the stuff ( I also ensure I consume butter sparingly)
Kind regards,
Emma
Hello Emma. That’s the typical level of trans fat in modern margarines. It’s very low – butter has about 5-6 times as much trans fat.
It’s almost impossible to totally avoid eating trans fats. To do so you would have to avoid beef, lamb, cheese and all whole milk dairy products as well as butter and margarine.
Better to look at limiting the total intake of trans fats. The World Health Organisation recommends that trans fat intake be less than 1% of calorie intake. Current intake in Australia averages about 0.5-0.6% of calories intake, which is quite low. These days most of our trans fat intake comes from dairy fat and meat fat. If you limit saturated fat intake you will be limiting the intake of these animal fats anyway and your total trans fat intake will be low. Regards, Bill