By Jerome Burne
If you aren’t confused about whether taking omega-3 will keep your brain sharp or is simply a waste of money, you haven’t been paying attention. Last week the media carried reports of a new study that found that taking a gram of fish oil a day didn’t protect elderly people from failing memory and poorer thinking skills.
So is this another bust for supplements and a chance to dismiss omega-3 supplements as ‘codswallop’ and a ‘waste of millions of pounds’?
Absolutely not. Read on before tipping those golden capsules in the bin because far from being another of those supposedly “scientific” studies showing that they are useless, it is actually a very good example of why randomised controlled trials (RCTs) of supplements can be very unreliable and misleading.
In fact anyone with a good knowledge of nutrition and supplements could have told the researchers that they had set up the trial in such a way that a cynic would say it was designed to ‘prove’ no benefit.
Glaring errors included:
- Testing the possible benefit of omega-3s to slow down cognitive decline in people who did not have problems with thinking and memory. Supplements are not magic brain boosters.
- Neglecting to test how much omega-3 the trial subjects had in their body already. It’s pointless taking a supplement if you already have a healthy level.
- Using people who were already taking statins. The researchers should have known there is plenty of proof that statins make omega-3 less effective.
The media were respectful because the trial ticked all the gold standard, evidence based medicine boxes. It was placebo controlled, run from Harvard, had 4000 subjects, who were treated for five years. When a trial like this says something is rubbish – it is rubbish.
That’s curious because we are all familiar with the idea of ‘bad pharma’ – the readiness of drug companies to spin the results from their RCTs so drugs look safer and more effective than they really are. So couldn’t the opposite be going on – making treatments look far less effective than they are? This sort of trial – all too common – also needs a name. How about ‘black trials’?
More details on why this omega-3 study is distinctly fishy
Flaw number 1:
All the people in the trial were at low risk for memory problems. They were educated, healthy and had a good diet. Testing omega-3 on them for brain protection was as sensible as testing a drug for period pains on men.
What they were all suffering from, in fact, was a disease of the retina called Age-related Macular Degeneration (AMD), and the trial was originally set up to test if various other minerals and vitamins could help that condition, not brain function. The omega-3 test was bolted on later.
Just how little their memory and thinking skills declined is set out in the trial summary. Over the five years the thinking ability of both those getting a placebo and those being supplemented dropped by a tiny amount – 0.19 a year. Most people’s memory gets a bit worse as they get older. A drop of 0.19 is in the normal range.
In drug trials, finding no difference between those who’d taken the drug and those who got the placebo means the drug doesn’t work. But if you are trying to keep someone healthy with a supplement, finding no decline in the placebo group means there isn’t a problem that needs fixing.
Flaw number 2:
The researchers failed to record the test subjects’ omega-3 levels before the trial began. Not checking existing levels in a drug trial is OK because no one already has statins or anti-depressant chemicals in their bloodstream. But if you are testing a natural substance, then patients in both the treatment and the placebo group will already have more or less of it.
So if you want to run a fair trial, you need to know who has a little and who has lots. Supplementing someone who is deficient is very likely to improve things; giving extra to patients with healthy levels won’t do any good at all. All of the patients in the treatment group in this trial got the same amount – a gram a day.
So it is already looking pretty useless – giving a supplement to protect mental functioning of people whose wits were already pretty sharp and giving a supplement whether patients need it or not. Unfortunately lots of supplement trials play both tricks -guaranteed cut the chances of showing benefit.
Flaw number 3:
Not paying attention to drugs being given alongside the supplement. Omega-3 becomes far less effective when given together with a statin. So it is very relevant that 44% of the patients in this trial were also getting statins.
Studies on the benefit of omega-3s in heart disease have found that when statins are given at the same time the benefit disappears.
Statins also make it easier for the body to use omega-6 fatty acids which crowd out omega-3 and, unlike omega-3, increase the insulin resistance that is a feature of diabetes. Diabetes is now a well-known side effect of statin use. And diabetics are known to be at greater risk for all the cognitive problems that come with Alzheimer’s.
Many drugs make supplements work less effectively – the diabetes drug metformin affects the amount of B12 you absorb from food – which are rarely considered when prescribing. Yet warnings the other way round – about the possible damaging impact of supplements on drugs – are of course common.
The research on the omega-3/statin link also found a damaging effect of statins in the brain. They can be toxic to the units that make energy (mitochondria) found in every cell of the body. While omega-3s protect brain cells, the higher the statin dose, the more damage there is.
Flaw number 4:
Doing a trial on the wrong patient group – people with an eye disease with no cognitive problems – makes it more likely something unexpected could skew the results. In this case the wild card is an amino acid called homocysteine, high levels of which have been linked with cognitive problems and Alzheimer’s for decades
Research published earlier this year reported that high levels of homocysteine are common among people with AMD. The importance of this is that high homocysteine is something else that can interfere with the brain-protective effect of omega-3 (see below).
Flaw number 5:
Ignoring all the evidence supporting omega-3 supplements on the brain and assuming that the result of one very unreliable trial is enough to claim the supplement has no benefit. ‘Our brains are packed with omega- 3s,’ says Dr Carrie Ruxton, dietitian for the Health Supplement Information Service. ‘Children certainly need them for proper brain development and they have been shown to help protect against diabetes and stroke.’
It’s true that some earlier trials have also found no benefit, such as an RCT by the London School of Hygiene and Tropical Medicine but it was just as flawed as the current one. The scientists gave a gram a day to healthy older people with no cognitive problems without testing their existing levels and, surprise, reported no improvement.
So what needs to be done?
Something similar to the AllTrials campaign, which is aimed at forcing drug companies to be more transparent about their results, is needed to save patients being fooled and quite possibly damaged by these misleading and unscientific trials condemning supplements. Supplements do of course need to be tested but let’s do it in a relevant way.
It just so happens there is a good example of a far more useful randomised trial of omega-3 and the brain also published last week that avoided the familiar flaws. After six months on Effalex Active – a multi-nutrient supplement containing just over a gram of omega-3s – memory and speed of information processing had improved.
It was small, just 27 post-menopausal women and there are plans for a larger version. ‘We measured the omega-3 levels at the beginning and end,’ says lead researcher Dr Simon Dyall, senior lecturer specialising in nutrition and neurology at the University of Bournemouth. ‘Those with more at the end did better.’
A perfect example of why there is a need for a rethink in the way we test non-drug treatments is what has happened to the work of David Smith, Professor Emeritus of Pharmacology at Oxford, who showed several years ago that taking high doses of B vitamins to lower the level of a damaging amino acid homocysteine, dramatically slowed the rate of brain shrinkage which is linked to cognitive problems. He’s been a victim of the practice of using these irrelevant black trials to dismiss his work. Full story here.
Why being hooked on RCT’s isn’t helping
But as well as showing how to do this research without falling into the common pitfalls, both Smith and Dyall’s work suggests a new direction. Cognitive decline doesn’t happen because just one thing goes wrong, it’s the result of a combination of unhealthy lifestyle factors and testing them one-at- a-time is of limited use.
So a few trials are starting that treat people with combinations. Following on from Smith’s study, Dyall included raised doses of B vitamins in the supplement his older women were getting.
Recent research by Smith has found that B vitamins need omega-3 to work properly. Without a high level of omega-3 patients didn’t gain the cognitive benefit from taking the high doses of B vitamins that was bringing down their homocysteine. So having high levels of homocysteine in the blood, as patients in the AMD trial did, will be reducing the effect of omega-3 More evidence that omega-3 is involved in protecting against cognitive decline.
But the more treatments you pack in, the less suitable RCTs, designed to test drugs one at a time, become. Researchers are already beginning to discuss the design flaws and pitfalls in testing supplements with RCTs – see here and here.
So a few researchers are talking about what are called ‘multidomain interventions’. For instance, high B vitamins, plus omega-3, plus exercise plus dietary change. This seems likely to benefit patients much sooner than waiting for the pharmaceutical magic bullet.
In the meantime I plan to watch out for the flawed misleading trials, hang on to my golden capsules along with high dose vitamin B and keep a critical eye, naturally, on future multidomain studies. I’ll let you know what I find.
Jerome Burne
Latest posts by Jerome Burne (see all)
WINNER: 
Spot on Jerome. Multidomain, multifactorial trials are definitely a great way to go. So is imperative (Comparative -see later comment) Effectiveness Research, where, for example, one multifactorial intervention can be be compared against a drug intervention and a zero intervention control. Even more complicated, but nevertheless also more likely to deliver more pronounced effects, are individualised multifactorial interventions that take into account epigenetic profiles and SNPs of individuals. The days of looking for silver bullets, whether they’re conventional drugs or ‘green’ medicines, should be numbered….
Should be: Comparative Effectiveness Research (CER)!
I hoped you’d look at this when I saw the report. I’ll keep taking the omega-3. My doctor told me that taking a vitamin B supplement wouldn’t help because the body couldn’t absorb it in tablet form? Bad advise, I wonder?
Absorption of vitamins a tricky and specialized subject, wouldn’t trust a regular doctor to be up to speed on it. They do regularly give injections of it, no idea how that works for lowering homocysteine as my understanding is that we don’t store B vitamins. More informed comments welcome.
High dose oral B vitamins (Folic acid, B12 and B6) readily lower plasma homocysteine. There is no need for injections. This treatment has been used in many trials, including the VITACOG trial which showed that in people with Mild Cognitive Impairment who had high homocysteine the B vitamins slowed shrinkage of the Alzheimer-related parts of the brain by 90% and slowed cognitive decline (Douaud et al PNAS 2013, 110:9523-9528. Preventing Alzheimer’s disease-related gray matteratrophy by B-vitamin treatment).