After a month, the participants were exposed to an oxidising agent. Again, as expected, there was significantly less fatty-acid oxidation with the highest doses of EPA and DHA compared with the sugar placebo pills.

But, in contrast to the earlier benefits, after six months, the oxidative damage in red blood cells was significantly greater with the highest doses of EPA and DHA. The authors concluded that long-term megadosing with these omega-3s could actually lead to atherosclerosis (Am J Clin Nutr, 1996; 64: 297-304).

High levels of the omega-3 alpha-linolenic acid (LNA) have also been implicated in an increased risk of heart attack. LNA is less effective than fish oil at reducing triglyceride fats (Am J Clin Nutr, 1997; 66: 591-8). In one study, fatty tissue taken from heart-attack victims was found to be high not only in trans fatty acids, but also in linoleic acid (LA, an omega-6) and LNA (Eur J Clin Nutr, 2000; 54: 618-25). Other data show an increase in atherosclerosis over the long term with omega-3s (Arterioscler Thromb Vasc Biol, 1996; 16: 1197-202).

Similarly, when the Nurses Health Study of 88,795 healthy women compared breast cancer rates in those who derived more than 30 per cent of their calories from fat with those who derived less than 20 per cent of their calories from fat, it found that women on the low-fat diet had the higher rates of breast cancer (JAMA, 1999; 281: 914). Closer examination of the findings revealed that all types of fats, except omega-3s, were protective. In fact, omega-3s derived from fish oil actually increased the risk of breast cancer.

Animal, vegetable, or what? Such findings only add fuel to the debate over not just optimal amounts, but optimal sources of essential fats. For years, fish and fish oils have been the preferred source of omega-3s. But most recently, with the increasing pollution of marine life, the pendulum has swung in favour of vegetable sources such as flaxseed (linseed) and hempseed oils.

But accumulating evidence suggests that replacing fish oils (which contain easily assimilated EPA and DHA omega-3s) with flaxseed oil, which has relatively high amounts of LNA (but is an earlier form or precursor to EPA and DHA) may not be as sound as we thought.

Although LNA can be converted in the body into EPA and DHA, in some people, particularly the elderly, this process may be inefficient (Am J Clin Nutr, 1991; 54: 438-63). Other studies have found that, even in healthy individuals, the amount of LNA converted into EPA in the body may be very low (Prostagl Leukotr Essential Fatty Acids, 2000; 63: 287-92).

It is generally believed that around 14 per cent of LNA is converted to EPA. But in one (albeit small) study, only 0.2 per cent of the LNA in a flaxseed oil supplement was available for conversion to the final omega-3 fatty acid - compared with the 23 per cent of EPA already present in fish oil (J Lipid Res, 2001; 42: 157-65).

Poorly metabolised oils may increase free-radical damage. The inability of some individuals to metabolise LNA may be one reason why its supplementation has been linked to higher rates of prostate cancer (J Natl Cancer Inst, 1994; 86: 281-6; Int J Cancer, 1997; 71: 545-51). LNA has been touted as a potent inhibitor of other types of cancer cells, but the evidence is inconclusive (Am J Clin Nutr, 1997; 66 [Suppl]: 998S-1003S).

Proponents of flaxseed oil claim that the improvement in skin condition often noted with supplementation is a sign that it’s working to improve the whole body. But this could simply be a result of the high phytoestrogen content of the oil rather a reflection of LNA action - and if this is the case, are these phytoestrogens doing harm or good?