There’s something odd about the recent study on omega-3 enriched margarine…

There’s been a lot of reporting of a study that came out on Sunday in the New England Journal of Medicine [1], and it’s going to be the focus of today’s blog. Here, in brief, is the design of the study: almost 5000 men and women who had had a heart attack previously were randomised to include in their diet one of the following four foods:

  1. margarine
  2. margarine enriched with EPA and DHA (omega-3 fats found in fish)
  3. margarine enriched with ALA (alpha-linolenic acid – a form of omega-3 fat found in plant foods such as flaxseed
  4. margarine enriched with EPA, DHA and ALA

During a follow-up period of 40 months, consumption of EPA/DHA and/or ALA was NOT associated with a reduced risk of ‘cardiovascular’ events such as heart attacks and strokes or death from these things. There appeared to be some benefits for diabetics from omega-3-enriched margarine, but taken as a whole, the results were pretty disappointing.

This study has generally been reported as evidence which shows that ‘perhaps omega-3 fats are not as healthy as we’ve been led to believe’. That’s certainly an interesting take, seeing as it was omega-3 fats that were being tested here. But omega-3 fats added to margarine. The issue here? Well, margarine in a plastic, chemicalised, highly processed food, that evidence links with adverse effects on health. See more about this here.

Now, imagine for a moment that omega-3 fats are supremely healthy. Would it make sense to eat them in conjunction with a non-food likely to have toxic effects on health? Would you drop your vitamin C in a glass of arsenic before chugging it down?

The interesting thing about how this study is being reported is that journalists have immediately leapt on the idea that omega-3 is left wanting. But what about the margarine? I can’t find any mainstream publication which is questioning the value that margarine may have. The authors of this study do not consider this possibility either.

It’s a shame that this study used non-enriched margarine as a placebo. The authors of this study really missed a trick here because a better placebo would have been, in my view, no margarine at all. Even though this would not have a been a true placebo, it would nonetheless have been interesting to see how those who ate margarine fared compared to those who did not.

Another potential flaw in this study related to the amount of additional omega-3 consumed by the participants derived from the margarine. This turned out to be 376 mg of EPA/DHA. And even looking at conventional recommendations, that’s unlikely to be enough to have real benefit (recommendations for those with cardiovascular disease are usually to have 800-1000 mg per day).

Another odd thing about the study is this sentence in the abstract results section:

“In the prespecified subgroup of women, ALA, as compared with placebo and EPA–DHA alone, was associated with a reduction in the rate of major cardiovascular events that approached significance (hazard ratio, 0.73; 95% CI, 0.51 to 1.03; P=0.07)”.

The confidence limits here (0.51 to 1.03) span 1, and what this means is that there was NO statistically significant reduced risk associated with ALA consumption in women. So, why even suggest it? This looks like authors desperate to find something positive in a study which was, overall, quite negative.

In such circumstances, it can sometimes be useful to find out if there was any involvement in the study from ‘industry’. And guess what  – there was! The trial was supported by, amongst other bodies, Unilever (a major manufacturer of margarine, including omega-3 enriched ones).

Could this fact, one might ask, have anything to do with the fact that we’ve not heard a word of dissent about margarine, and the authors appeared to be clutching at straws in terms of their interpretation of the data? And could it have anything to do with the fact that the study design did not include a group who ate no margarine at all?

References:

1. Kromhout D, et al. n–3 Fatty Acids and Cardiovascular Events after Myocardial Infarction – Alpha Omega Trial Group NEJM August 29, 2010

14 Responses to There’s something odd about the recent study on omega-3 enriched margarine…

  1. Jennifer 31 August 2010 at 5:06 pm #

    I agree. The effects of margarine may override any benefit from the Omega 3s, particularly considering that margarine is usually put on grains. Every one should remember that this study applies to those who have already had a heart attack, not to those looking for preventative wellness. And Omega 3s also help couples conceive and promote mental health. Here are a couple of articles that detail why it’s important to get your Omega 3s:
    http://biovedawellness.com/2010/08/nutritional-protection-omega-3-essential-fats/
    http://biovedawellness.com/2010/07/fish-oil-can-it-help-you-conceive-a-baby/

  2. Lucien Grange 31 August 2010 at 5:58 pm #

    I find it particularly frustating that the composition of
    the margarine is nowhere indicated. Apart from the margarine
    being a poor non-food, most brands are very high in omega 6
    which having common bio-chemical pathways with omega 3
    could easily explain how the weak doses of omega 3s were
    ”crowded out” by large consumption of omega 6.

    Any idea about marge composition ?

  3. John Briffa 31 August 2010 at 6:11 pm #

    Lucien

    If you read the paper, it says in reference to the margarine that the specific composition “could not be identified”. Yes “Unilever provided an unrestricted grant for the distribution of trial margarines to the patients.” From this, it appears we are to accept that Unilever provided the margarine but no-one was allowed to know exactly what was in it. Though I may be missing something…

  4. Lucien Grange 31 August 2010 at 10:29 pm #

    Having now checked the supplementary tables which are
    attached to the study – which I might profitably have done
    earlier !- I see that the statement that no one could know
    what the margerine contained applies to the randomized patients.
    The Linoleic acid omega 6 varies between 37.7% and nearly
    40% of the product. Not good !

    If you don’t mind, dear Doctor, I shall have some of that
    butter with you.

  5. dennis 1 September 2010 at 12:50 am #

    Another example of reductionism. Isolated Omega 3 fats are a food extract not a wholefood. Why would we expect one isolated component of a wholefood to be elevated to such a ‘magic bullet’ panacea status ? Of course we should eat natural foods as close as possible to the food our ancestors ate – sea fish, fresh greens and even seeds from which omega 3′s can be extracted – but there’s little or no merit in consuming extracted omega 3 fats on their own and it’s pointless or dangerous to consume margarine which is an entirely synthetic ‘food’. Also avoid refined, processed, preserved and intensively farmed / unnaturally fed foods whose lipid profile is unnatural. The press always distorts the conclusion that food ‘containing x is good’ into ‘x is good’
    where x = omega 3, vitamin D or whatever is currently in vogue.

  6. Jamie 1 September 2010 at 1:02 am #

    I read the alerts for this and saw two things – the low 400mg dose of EPA/DHA and the fact that these were added to margarine. At this point, this study was dead in the water in terms of anything meaningful coming from it. And given the source of the funding, it comes as no surprise.

    I also thought to myself, ‘John will be all over this like white on rice’. Correct! Nice post.

  7. Reijo Laatikainen 1 September 2010 at 8:54 am #

    Margarines are currently almost devoid of trans-fat (you can see trans-fat content of margarines in supplementary tables as Lucien states). Butter includes more trans-fat. So your claims on toxicity are not substantiated by trans-fat. What is the toxic element of margarines other than trans-fat?

    We still have two RCT studies showing benefit of omega-3s in similar setting (Jelis study and GISSI Prevenzione). On the other hand, a study from Germany, not published yet as paper, found no benefit either after MI (http://bit.ly/aWaQ4C). No study has shown negative effect.

  8. Jamie 2 September 2010 at 8:19 pm #

    Margarines are generally less than 1-2% trans fat… a lot lower than they used to be, but not devoid of trans fat. How much does it need to contain in order to cause damage when consumed over many years.

    The main trans fat in butter is conjugated linoleic acid. It is well accepted in nutritional circles that CLA does not carry the same risk as other forms of trans fat (though it is potentially not without risk).

  9. dennis 2 September 2010 at 8:55 pm #

    The point about margarine is that its components have no nutritional benefit – not that it is acutely toxic (except perhaps for the any synthetic transfats) – however if any significant calories are obtained from margarine it can displace calories from foods which have nutritional benefits – not a good idea. So there is no point in consuming margarine and much to lose from consuming significant amounts. If one really believes that isolated fish oils are beneficial (and I accept there may be a little bit of evidence for this) then they can be taken. But the right thing to do is to correct the diet by eating natural foods, avoiding foods with unnaturally high levels of omega6

  10. Reijo Laatikainen 3 September 2010 at 12:59 pm #

    I was aslo disappointed on how low the used doses were (376 mg of EPA+DHA + dietary intake of 120 mg at baseline). Authors did not speculate in their discussion if total intake of EPA+DHA c. 500 mg was insufficient. A bit strange. In JELIS and GISSI Prevenzione studies, supplementation of EPA and DHA was at least double (1 800 mg EPA and 850 mg EPA+DHA respectively) and positive outcomes were reported.

    @Jamie. I agree what you say about CLA. But I’m not sure if CLA really forms so much of total trans-fat. To my understanding CLA content of butter is something like 15 % of trans fat, but I haven’t seen any comprehensive data though.

    This Canadian study ( http://bit.ly/cXLtHv ), that I was able to find via PubMed, states that in butter, cream and other dairy products total trans fat if c. 5.5 % of total fat whereas CLA content is about 0.5-0.9 % of total fat. It sounds like CLA content is approximately 10-20 % of total trans fat. What do you think?

  11. elaine 3 September 2010 at 1:22 pm #

    but surely if they were testing whether it is worth buying the omega 3 enriched marg over regular marg, this is a perfectly good experiment (though journalists will always c*ck up the reporting!)
    PS Isn’t margarine disgusting?

  12. Xenia 3 September 2010 at 5:08 pm #

    @ Reijo Laatikainen:

    Please. CLA = God made, trans fats in the magarine = man made (or better make that ‘factory made’). When was this ever the same? This is just like trying to compare natural tomato and a GMO monstrosity created from tomato.

    But it is not just about the trans fats. It is simply about processed, artificial products masquerading as food. So you need no “other” toxic ingredient. There might be one, but what’s important is the fact that this is simply not food! Humans might have lost their instincts about that (and basically their sound mind …) but whenever in doubt, ask the animals. Ants won’t go anywhere NEAR any margarine, take my word for that. It is more than clear that this is a piece of plastics to them. Not food! But go and leave on your doorstep a piece of butter and it will disappear in no time.

  13. dennis 3 September 2010 at 8:27 pm #

    Absolutely right Xenia. The biggest disaster for the western diet was in 1973 when the US FDA withdrew Congress’s 1938 ‘imitation food’ law and all sorts of synthetic imitation foodlike substances could pass as food. If margarine was better than butter, cows would have evolved to produce margarine !

  14. Florence 3 September 2010 at 10:20 pm #

    I am pleased that a couple of people have mentioned Omega 6, because I believe that here lies a major problem. People are consuming way too much Omega 6 these days and the balance between that and Omega 3 has shifted considerably. Dr Briffa, I think you mentioned in a previous blog some time ago that the ratio of Omega 6 to Omega 3 should be 1 to 3:1 when it is now 20 to 30:1. This can’t be good, so why are people not being advised to reduce their consumption of Omega 6, rather than up their Omega 3?

Leave a Reply to Reijo Laatikainen Click here to cancel reply.