PDF | E-mail

Fuelling fishy furphies

Fuelling fishy furfies

A recent Adelaide Advertiser article opened with the sentence:

Taking omega-3 supplements can protect us against heart disease, inflammatory diseases like arthritis and are important for the development of the brain and visual system in infants.

As an opening sentence, this has the kind of definiteness that normally comes from a journalist writing about some issue they don't know anything about. Like when the sports writer has to cover for the medical writer or vice versa. Scientists, on the other hand, are frustratingly famous for rejecting black and white in favour of grey with their finely nuanced statements. But this article wasn't from the pen of any time-starved journalist out of their element and meeting a tight deadline, it was from an Adelaide University academic ... Dr Beverly Muhlhausler.

Is there evidence that omega-3 oils are the magic bullet she claims? Here's a few recent Omega-3 studies and a 2006 British Medical Journal paper:

1. Omega-3 no match for Alzheimer's This describes research where Alzhiemer's patients were given a long chain omega-3 oil in an attempt to slow progression of the disease ... it didn't.
2. Fish oil pregnancy doubt This describes a major study where pregnant women were given either omega-3s or a placebo to see it would reduce post natal depression in the women or improve subsequent cognitive and language development in her child. It did nothing.
3. Low-Dose Omega-3 Fatty Acids Don't Protect Heart Patients Having had a heart attack, patients took omega-3 (they say low-dose, but I'd say modest) to prevent further trouble. It did nothing.
4. Hot off the press ...being published on 1st of December, a study of high dose omega-3s to prevent abnormal heart rhythm. As is becoming common enough to be laughable, pissy little mickey mouse trials by people enthusiastic about the treatment looked promising, but when the research was done properly with randomised, prospective, double blind placebo controlled trials ... nothing happens. Actually, there's nothing wrong with mickey mouse trials, they guide decisions on larger trials, but they shouldn't guide media hype.
5. Lastly, a not quite so recent, but far more important study. A 2006 meta-analysis of all the high quality studies into heart disease and omega-3 oils: Risks and benefits of omega 3 fats for mortality, cardiovascular disease, and cancer: systematic review. The conclusion? Zip. Nothing. No effect.

I'll provide a little historical context for the study of omega-3 oils later, but as a prelude let me note that studies on fish and health are intrinsically difficult. Wealthy people are healthier people for all kinds of reasons and they also happen, on average to eat more fish ... because it's expensive. The association of fish and health may not be causal just as smokers are more likely to carry matches but it's not the matches which cause their lung cancer.

 

But could Dr Muhlhausler or another omega-3 advocate find research with different findings than the studies I've just referred to? Yes, she could. I contacted Dr Muhlhausler and asked her for evidence supporting her view and I'll look at it later. But regardless of the number and quality of studies she or I can produce, I think it is undeniable that any definitive statements asserting that omega-3 oils protect against heart disease are simply false and misleading. Whatever else it is, the evidence isn't even close to conclusive. What does conclusive evidence look like?

Compare the omega-3 evidence with a known causal link

Contrast the unproven link between omega-3 and heart disease with the convincingly proven link between red and processed meat and bowel cancer. This link is now considered to be causal by the World Cancer Research Fund (WCRF). In 2007 a researcher being paid by Meat and Livestock Australia presented results from 32 studies in a special issue of Nutrition and Dietetics on red meat in the Australian diet. Every single study he produced ... all 32 of them ... found that the people eating the most red meat had a higher rate of bowel cancer than those eating the least red meat. The only type of inconsistency he could, and did, seize on in an imaginative and desperate attempt to protect the product of the organisation paying him was that not all of the studies found a statistically significant increase. Of course not.

To prove this or that food causes or prevents disease is extremely difficult. The WCRF, whose weighty 2007 report by 150 authors concluded that red and processed meat causes bowel cancer, requires multiple types of evidence. First a body count. Consisent evidence that people consuming or avoiding the food get sicker quicker. This is provided in spades by the 32 studies. Secondly they require plausible biological mechanisms. There is no shortage of those in the case of red meat and there is clear clinical evidence which supports at least one of these mechanisms ... you can see the DNA damage in colon cells after a red meat meal.

In the case of omega-3 oils. There's plenty of theory, but the body count is missing ... and not through want of trying. There has been a steady stream of studies. But the evidence of protection against heart disease from omega-3 oils is far weaker than evidence that red meat causes bowel cancer.

Emotional blackmail

Telling people they should take omega-3 oils for their heart is one thing. Telling pregnant women to take them for the health of their unborn child is rather different. The clear implication of Muhlhausler's assertion that pregnant women need omega-3 supplements to ensure proper development of their babies brain is even worse than just plain false. The attempt to lay a guilt trip on pregnant women is, in my view, quite despicable. How could a pregnant women resist such pressure and "take risks" with her babies future?

Consider in contrast Australia's highest health authority ... the National Health and Medical Research Council (NHMRC). If Muhlhausler was correct then the NHMRC would insist on long chain omega-3 oils being added to infant formula. They don't. The optional addition of these products to infant formula is a recent marketing initiative. Whatever the disadvantages associated with using infant formula, it has produced millions of perfectly good brains and functional eyes all over the planet for decades with zero long chain omega-3 oils.

We can actually go one step further. In Brains biceps and Baloney I go through a little of the data on the increases in intelligence across the world over many decades with no increase in long chain omega-3 oils ... even among children in Kenya with almost no animal products in their diet at all.

Fish, chicken, fermentation or transgenic plants?

On a positive note, Muhlhausler's article shows evidence of a refreshing and growing acknowledgment by at least some nutritional researchers and dietitians that environmental issues are of some consequence. She understands that fishing at current levels isn't sustainable and that if there really was a need for more long chain omega-3s, then aquaculture won't meet it. Contrast this with the CSIRO in 2005 launching the most environmentally destructive diet on the planet in apparent total ignorance of climate change, deforestation, or the crippling of the Murray Darling by the dairy industry ... to name just a few issues. The 2009 launch of a new CSIRO Total Wellbeing Recipe book shows that the CSIRO remains not so much ignorant as environmentally malicious, despite having some of the best environmental scientists on the planet who have (politely and diplomatically) rubbished the diet.

But long chain omega-3s from chicken? Apart from raising serious risks of a global pandemic, the chicken industry causes widespread and intense suffering in chickens. Chickens for meat production have been bred to grow furiously and their skeletal system frequently fails to keep up with their musculature. Crippled chickens is the result.

Crippling chickens for arthritis relief

A 2008 UK study on leg disorders will be reasonably close to the mark for any factory farmed broiler production system all over the world. This is a highly globalised and homogeneous industry. In a tour de force, the researchers scored the walking ability of 51,000 birds in 206 flocks. The average percentage of birds in a flock who could walk normally at 40 days was just 2.2 percent with 25 percent scoring 3 or more on a scale of 0 to 5 (5 is totally unable to walk). It's hard to think of a more vicious and sadistic animal production industry ... except perhaps the fishing industry. Are the chickens in pain? Or perhaps they just can't walk properly. This is easy to test and the research has been done. Give chickens access to feed with and without carprofen (an analgesic and antiinflamatory drug) and they will choose the carprofen.

There is a particular irony in using an industry built on suffering and join pain producing a compound that people will take in desperation for their arthritis. This website has recently been updated with the latest information on the brutality of the industry Muhlhausler is advocating for.

So while Muhlhausler shows some environmental awareness, she is either ignorant or unconcerned with pain and suffering as an issue in the production of long chain omega-3 oils.

The international pharmaceutical industry isn't known for its animal compassion, but it cares about the mighty dollar, so it is very busy working on transgenic crops to produce LCO3s as well as bacterial fermentation systems. These perhaps solve the ethical dilemmas for people wanting to try LCO3 supplements. Getting ripped off? There's no solution for that one.

But what about the Eskimos?

Let's step back a little and go through some of the history of the fuelling of the omega-3 band wagon.

A 1975 paper contained an unreferenced claim that "coronary atherosclerosis is almost unknown amount these people [Greenland Eskimos] when living in their own cultural environment". The paper has received 520 citations in the scientific literature, according to Google Scholar. That's a huge number of citations. But it is over a long time period. One of the citing papers was a 1991 paper in the prestigious medical journal The Lancet. That paper in turn has had 446 citations.

In 2003 a study appeared from some scientists at the Greenland institute of Public Health into the coronary health of Inuits ... these are the largest subgroup of peoples called Eskimo. At last, a quarter century after the horse had bolted and the stable was paved over with a car-park, someone had finally decided to actually check the data! This is precisely the way science isn't supposed to work ... but sometimes does. A researcher with a good reputation makes a claim which soon becomes gospel. Gospel is for religions, and rumour is for daytime soapies, science should be different. No matter, better late than never. The lead author of the study was Preben Bjerregarde who has over 60 peer reviewed papers ... he's no mug. Checking this claim is easier said than done. Inuit can be nationals of Canada, Greenland or Alaska and national databases don't always make selection based on ethnicity easy. That difficulty is compounded by the fact that even fewer Inuit lived traditional lifestyles in at the dawn of the millenium than 25 years earlier. So Bjerregarde and his crew combed the scientific literature and their own extensive medical records for data.

Cutting to the chase ... here's a quote from the conclusion:

Mortality from stroke is similar or probably higher among the Inuit than among other western populations. The evidence for a low mortality from Ischemic Heart Disease is fragile and rests on unreliable mortality statistics.

 

The claim that coronary disease was unknown is rubbish, it may not even be low. How many citations has Bjerregarde received since 2003? Just 39. It seems very few people want to know.

Muhlhausler's evidence

When I asked Dr Muhlhausler what she based her quite strong statements in The Advertiser on, I also sent her a link to the meta-analysis in the British Medical Journal. Her response was to say that "...more recent meta analyses have supported a protective effect of long chain omega-3s on cardiovascular disease".

She attached two studies in support of her claim ... a review in the Journal of the American College of Cardiology and a review paper from the Lancet..

The second paper doesn't support Muhlhausler's level of certainty at all, stating explicitly:

"Information about the role of [long chain omega-3s] intake or supplementation in primary prevention of coronary artery disease is scarce, but the available evidence suggests that those with hyperlipidaemia and diabetes might benefit most."

The paper is principally concerned with the use of long chain omega-3s (LCO3) as drugs for people with particular problems. This is a very different issue. Primary prevention is about healthy people taking LCO3 to stay healthy. Secondary prevention is about people taking LCO3 after a heart attack to prevent another one.

The first paper, on the other hand does indeed wax lyrical about the primary and secondary protective value of long chain omega-3s. But it's a review rather than a meta-analysis. What's the difference? A review focuses on what are considered to be the important studies in an area without pretending to be complete in any sense. A meta-analysis tries to consider all studies. It then applies rigorous standards about which studies to include or omit from the final analysis which combines the results of the studies quantitatively. In my view a good meta-analysis carries more weight than all but the best reviews and this review, with lead author Carl Lavie, doesn't inspire confidence.

For example. Lavie praises an early secondary prevention trial as follows:

In a randomized trial (DART) (16) ... performed 2 decades ago in 2,033 men with recent MI, -3 PUFA, either in the form of oily fish or fish oil capsules, reduced 2-year all-cause mortality by 29% with the benefit almost entirely attributable to a reduction in CHD mortality.

This is the worst kind of cherry picking. One of the reasons the DART trial is famous is that the apparent mortality reduction after 2 years reversed into a 30 percent mortality increase after three more years. After 10 years, its pretty much neck and neck despite the consumption of a whole lot of extra fish.

As for primary prevention, Lavie doesn't have much at all. The paper he cites as the major primary prevention put people with high cholesterol on both statins and LC03 ... and achieved a modest improvement compared to statins alone. Is this primary prevention?

In summary, Muhlhausler either doesn't have any evidence to match the careful meta-analysis I cited at the beginning of the piece, or she chose to send me second rate or irrelevant studies.

Compared with what?

It's important to compare whatever reductions or increases in heart disease that long chain omega-3 oils may cause with other ways of reducing heart disease. There are plenty of other games in town. The DART study discussed above is a large study by omega-3 study standards, but it involved just a couple of thousand people. The EPIC-Oxford study is a much bigger UK study of lifestyle and disease with over 64,000 people. It's participants on average suffered less than half the UK average death rates from heart disease ... and had similarly low rates of all major diseases. What did they do? The participants are called "Health Conscious". Those that eat meat eat about a quarter of the UK average and about a third were vegetarians. More than half ate no fish at all. About a third had pre-existing health conditions at the start of the study ... many people become vegetarian as their normal diet starts to make them sick by middle age. Despite this, the overall mortality rates were still half the UK average. The death rate from heart disease of the vegetarians was lower than the sub-group who ate fish.

Concluding remarks

Dr Muhlhausler's Advertiser article has potentially mislead tens of thousands of South Australians. The certainty with which she made her claims was unjustified. Unfortunately, the remedies for such deceptions are few and ineffective. Blog pieces like this don't have the reach of mainstream media and few people can spend the time sifting through the academic journals to make an informed judgement. I would urge Dr Muhlhausler to write to the Advertiser with a retraction for publication.