Omega-3 Fatty Acids, Intelligence, Mental Health and Food-Marketing Mind Games
Docosahexaenoic acid (DHA) and arachidonic acid (ARA) are both present in large quantities in the grey matter of the human brain, and both accumulate at a rapid rate during the central nervous system growth spurt that takes place shortly before and after birth. Given this, it would be a surprise if these fatty acids did not benefit the brain in some respect. However, there are several reasons to be wary of the way in which this inferred relationship is popularly portrayed, and several clarifications that need to be made.

DHA is one of many omega-3 fatty acids. Alpha-linolenic acid (ALA) is referred to as the "parent" of the omega-3 family because it is the precursor to eicosapentaenoic acid (EPA), which is the precursor to docosapentaenoic acid (DPA), which is the precursor to DHA. Linoleic acid (LA) is referred to as the "parent" of the omega-6 family because it is the precursor to gamma-linolenic acid (GLA), which is the precursor to ARA. As a result, it is sometimes assumed that consuming generous quantities of ALA (e.g. from flax seeds or walnuts) and LA (e.g. from walnuts and most seeds and vegetable oils) will have a beneficial effect on the brain.

Often the emphasis is placed entirely on "omega-3", with all types and sources of omega-3 being deemed equally worthy, and all persons of all ages being portrayed as beneficiaries. Since the inherently "good" and "healthy" status of omega-3s is seen to be self-evident, it is further assumed that their goodness is dose-responsive and that brains will get bigger and moods better (both without limit in space or time) as more seafood is eaten and as more flax oil is poured over more grass-fed meat and more green vegetables. So single-minded is the focus on the putative benefits of omega-3s that the potential positive or negative impact of any other food, nutrient, lifestyle or socio-economic factor is completely disregarded.

Producers and marketers of food items including organic yogurts and chocolate bars to which flax seeds have been added are adamant that eating their wares is the pathway to mental prowess and emotional contentment hitherto not imagined. Their rationale is that their products contain ALA, which can be converted into EPA and DHA, which "have been shown" to improve mood, intelligence and whatever organ, function or ailment you can think of.

In reality, the conversion rate of ALA into DHA in humans is barely higher than zero. Supplementing with generous quantities of ALA from flax did result in significantly greater concentrations of EPA and DPA (but not DHA) in the plasma of lactating women, but not in greater concentrations of any of them in the breast milk (where DHA was most needed). There was a "trend" for EPA (only present in minute quantities in the brain) to increase in breast milk, but it looks like there was also a trend for the trendiness of this trend to become less trendy each week. (The conversion of LA into GLA and ARA is similarly low.) Consuming the longer-chained omega-3s (DHA and EPA) directly from fish, fish oil or the yolks of some varieties of egg is considered optimal. Even then, positive outcomes related to brain function, mental health and other aspects of health have often been mild and/or have resulted from small rather than large doses of DHA and/or EPA (when they have resulted at all). This conflicts with popular belief, but it makes sense considering that (a) ARA is even more prevalent in the grey matter than DHA but is often overlooked, that (b) the total-body concentrations of both DHA and ARA are small compared to those of other fatty acids (despite their particular prevalence in one part of the body), and that (c) the potential beneficial impact of DHA and/or ARA on brain development must be much smaller in the vast majority of people (all those older than a few months) whose central nervous systems are not in the midst of a massive growth spurt.

Let's take a closer look at some relevant research, starting with that relating to the brain development and function of those most likely to benefit (i.e. the recently born and the soon-to-be-born). By four years and seven years of age, there was no link between DHA and ARA levels in phospholipids from umbilical venous plasma or red blood cells (both indicative of DHA and ARA status during the most crucial period) and cognitive performance. By then, any benefit from infancy fatty acid status had been outweighed by lifestyle (smoking during pregnancy) and socio-economic (parental education and/or intelligence) factors.

In another comparison, this one involving maternal supplementation with cod liver oil (rich in DHA and EPA) or corn oil (quite high in omega-6 but with no ARA), maternal DHA intake appeared to favourably affect offspring IQ at the four-year mark, but any beneficial effect had been outweighed by other factors at the seven-year mark. Maternal plasma phospholipid levels of ALA and DHA during pregnancy were linked with better "sequential processing" among the offspring, but that was merely one aspect of the intelligence test, and overall performances didn't differ.

After a different trial, this one involving the addition of DHA or DHA-plus-ARA to infant formulas, the trial authors generously concluded that DHA and ARA supported visual and cognitive development at 39 months of age. The plainer way of putting it would be to say that DHA-plus-ARA didn't make a damn bit of difference - and DHA alone appeared to be producing lower vocabulary scores at the 14-month mark!

Formula-fed infants often fare less well than breast-fed ones, and their visual acuity (at four years of age) was worse in this trial, but those whose formulae were supplemented with DHA-plus-ARA were able to match the breast-fed ones. However, those whose formulae were supplemented only with DHA did worse than the breast-fed ones (as did the formula-fed control group) on verbal IQ tests.

The intelligence of offspring was estimated earlier (at nine months rather than four years) in another trial, and those whose mothers had eaten "DHA-containing functional food" up to delivery did better at certain problem-solving tasks but not at "recognition memory" ones. "DHA-containing functional food" was the rather grand way of describing cereal bars to which a high-DHA, low-EPA fish oil (or corn oil for the placebo group) had been added.

The patterns emerging from the above research are (a) a weak trend for omega-3 in the form of DHA to be more efficacious than omega-6 in the form of LA, and (b) a not-quite-so-weak trend for a mixture of DHA and ARA to be more efficacious (or less deleterious) than DHA alone. Hardly a glowing endorsement!

Despite the potential modest benefits of modest amounts of ARA on brain development, ARA has a bad reputation for being "pro-inflammatory" via its role as a substrate used by "cyclo-oxygenase" (COX), which in turn can produce "prostaglandin H2" (PGH2), which in turn can stimulate the production of vasoconstrictive, platelet aggregative "eicosanoids" such as "thromboxane A2" (TXA2). On the other hand, COX can fuse with "prostacyclin synthase" (PGIS) to form vasodilative, platelet anti-aggregative "prostacyclin" (PGI2), and a protein has been constructed to produce prostacyclin from ARA continually. Natural means of increasing prostacyclin synthesis appear to be vitamin E and vigorous muscular contractions.

EPA is thought to be "anti-inflammatory" owing to its ability to oust ARA from phospholipids, and research on the effect of omega-3s on mood and mental health in adults has therefore often focused more on EPA than on DHA. Let's have a scan through that research.

One review focused on a frequently found association between the presence of major depression and a higher ratio of ARA to EPA (or total omega-6 to total omega-3) in plasma phospholipids, erythrocytes, cholesteryl esters or adipose tissue. Although the association seems quite consistent, the possibility was raised that it is not causative and is merely an artefact of an underlying process that does cause depression. Erythrocyte samples from depressed and non-depressed subjects were both depleted of total omega-3 after being experimentally subjected to oxidative attack via hydrogen peroxide.

The same review also looked at depression treatment studies in which EPA and DHA (or either one isolated in ethyl form) were given in various doses alongside standard anti-depressants. The addition of smaller rather than larger doses of EPA and/or DHA (e.g. 1g per day rather than 2g or 4g) often did appear to offer some benefit on more serious forms of depression, and EPA appeared more efficacious than DHA when they were given separately, but (as noted in the next review) DHA has rarely been tested in isolation and on one occasion its potential efficacy (on subjects who were mostly only mildly depressed) was cancelled out by a chance turn for the better in some members of the placebo group.

Another review found little or no overall impact of EPA and/or DHA on schizophrenia, and no conclusive evidence of any significant effect on ADHD ("attention-deficit hyperactivity disorder") and similar conditions, but observed a similar positive effect of EPA and/or DHA on major depressive disorder and bipolar disorder. The effect of DHA alone was not better than placebo when 2g per day were given, but the subjects did not have depleted DHA levels to begin with, and the other DHA-only trial was the one in which many members of the placebo group happened to get better as well. EPA alone seemed to be effective only when smaller rather than larger doses were given, but larger amounts (as high as 6.2g) were effective when they were given alongside DHA (as high as 3.4g) in fish oil (which may also contain ARA and fat-soluble vitamins such as A/retinol and D3/cholecalciferol). Interestingly, the odd instances in which EPA and DHA appeared effective against ADHD or related conditions were those in which they were given alongside ARA and/or GLA and vitamin E.

A thorough meta-analysis of 12 randomized controlled trials investigating the effect of various doses of EPA and/or DHA (with or without other treatments) on various types/definitions of depression or mood disorder in various amounts of subjects of various ages over various time periods painted a less optimistic picture of omega-3 supplementation. Six candidate trials were excluded from the analysis due to unsuitable data, but only two of those had reported a beneficial effect from EPA and/or DHA. The analysts used various statistical models to try to account for the heterogeneity of the data. "Random-effects models" place greater weight on smaller studies and can prevent a small number of large trials from dominating the results, but "fixed-effects models" don't give such weight to smaller studies and can prevent the results from being dominated by "publication bias" (the non-publication of research that hasn't found a desired effect, a bias which is more likely to occur after smaller studies on which less time and money have been expended). "Funnel plots" are used to test for publication bias, with asymmetrical plots indicating the potential for such bias, and funnel plot asymmetry was marked even when one large trial (which found no positive effect from fish or fish oil) was excluded. However, excluding that one large trial did result in a small overall positive effect from EPA and/or DHA, and there was also a positive effect from EPA and/or DHA (including that one large non-positive trial) on major rather than minor depression (consistent with the previous reviews). The problem was that there was evidence of potential publication bias for the positive effect on major depression, but not for the non-positive effect on minor depression.

Despite the above uncertainty, the research seems fairly consistent on a few points: (a) any efficacy of EPA and/or DHA is stronger on major rather than minor depression; (b) smaller rather than larger doses are more often efficacious; but (c) a wider range of conditions may be positively affected, and/or larger doses may be effective, when EPA and DHA are given together and/or alongside ARA and/or GLA in products that may also contain large quantities of fat-soluble vitamins and/or other essential nutrients. The last point emphasises the folly of over-excitedly developing a narrow focus on any one factor (essential nutrient or not), and fits in with the research suggesting that a combination of DHA and ARA (rather than DHA alone) is more beneficial for infants.

It would be surprising if EPA in any dose has a mood-elevating effect not shared by DHA, not least because DHA can actually be "retro-converted" into EPA. The subjects in this experiment were fed daily 6g doses of ethyl-EPA or ethyl-DHA. EPA was converted into DPA but not detectably into DHA, and also lowered ARA, whereas DHA was partly converted into EPA and didn't lower ARA!

Another interesting point, as outlined in the second review cited above, is that the beneficial effect of small EPA doses may not be due to ousting ARA from phospholipid membranes. On the contrary, it appears that small doses of EPA increase the activity of enzymes that also enhance the uptake of ARA into phospholipids, but that large doses of it do the opposite and thereby could facilitate the use of unincorporated ARA for the production of pro-inflammatory eicosanoids!

"Omega-3" generally does not offer any evident benefit to the brain, but one omega-3 fatty acid (DHA) and one omega-6 one (ARA) are essential components of the brain, and it is therefore a no-brainer that some DHA and some ARA are better than none. However, they do not possess a magical dose-responsive ability to boost brain size and power, and the amounts required are probably fairly modest for infants and even more modest for grown adults. Adults suffering from stronger forms of depression often have depleted omega-3 levels, and EPA and/or DHA appear to have a small benefit in these subjects, but again the required doses are modest, especially when they are given in isolation. The most important conclusion is that too great a focus should never be placed on a single factor. DHA in some amount is important for infants, but ARA is equally or more important and the impact of both can easily be overridden by lifestyle and socio-economic factors. EPA (along with DHA, ARA, GLA and the fat-soluble vitamins given with them) holds modest promise for those afflicted by certain forms of depression and mood disorder, but large, unnaturally isolated quantities of EPA could do more harm than good. Despite the encouragement of the media, it is helpful if the relationship between different nutrients is not conceived as a struggle between good and evil!

Written: April 2009