Elsevier

Progress in Lipid Research

Volume 42, Issue 6, November 2003, Pages 544-568
Progress in Lipid Research

Review
Problems with essential fatty acids: time for a new paradigm?

https://doi.org/10.1016/S0163-7827(03)00038-9Get rights and content

Abstract

The term ‘essential fatty acid’ is ambiguous and inappropriately inclusive or exclusive of many polyunsaturated fatty acids. When applied most rigidly to linoleate and α-linolenate, this term excludes the now well accepted but conditional dietary need for two long chain polyunsaturates (arachidonate and docosahexaenoate) during infancy. In addition, because of the concomitant absence of dietary α-linolenate, essential fatty acid deficiency is a seriously flawed model that has probably led to significantly overestimating linoleate requirements. Linoleate and α-linolenate are more rapidly β-oxidized and less easily replaced in tissue lipids than the common ‘non-essential’ fatty acids (palmitate, stearate, oleate). Carbon from linoleate and α-linolenate is recycled into palmitate and cholesterol in amounts frequently exceeding that used to make long chain polyunsaturates. These observations represent several problems with the concept of ‘essential fatty acid’, a term that connotes a more protected and important fatty acid than those which can be made endogenously. The metabolism of essential and non-essential fatty acids is clearly much more interconnected than previously understood. Replacing the term ‘essential fatty acid’ by existing but less biased terminology, i.e. polyunsaturates, ω3 or ω6 polyunsaturates, or naming the individual fatty acid(s) in question, would improve clarity and would potentially promote broader exploration of the functional and health attributes of polyunsaturated fatty acids.

Introduction

Things should be a simple as possible but no simpler (Einstein)

An ‘essential’ nutrient is one that is needed for normal development and function of mammalian cells throughout the life cycle. In its active or precursor form, there is a minimum amount of such a nutrient that must regularly be provided in the diet. This dietary requirement generally varies with species, gender, age and presence of physiological and pathological challenges (pregnancy, lactation, infancy, aging, infection, disease, etc.). Inherent in the concept of dietary essentiality are selectivity and conditionality; not all nutrients are essential in the diet nor are those that are essential needed in the same amount throughout the life cycle. Amongst dietary fats, the ‘essential fatty acids’ (EFA) have been identified as ‘essential’ nutrients. However, there has been a tendency to oversimplify not only the symptoms of EFA deficiency, but also which ones are needed in the diet and in what amount. Thus, in the case of EFA, the cardinal rules of selectivity and conditionality are being frequently strained or ignored. Hence, the quotation above aptly sums up my view of the state of the art regarding EFA.

This paper outlines several examples of why things are less simple with EFA than they seem or than they are portrayed most of the time. For instance, the term EFA is widely used to refer to about eleven 18–22 carbon ω6 and ω3 polyunsaturated fatty acids (PUFA; Table 1). However, it can legitimately be applied to well over thirty ω6 and ω3 PUFA that vary in chain length from 14 to at least 40 carbons in chain length, and contain from 2 to at least 8 double bonds (Table 1). Clearly the term EFA can be applied to a large number of fatty acids that are unlikely to be nutritionally ‘essential’ in the true sense. The outcome of this loose application of the term EFA is that no one, even amongst the specialists, really has the same definition of an EFA. Many specialists focus this term more or less depending on the circumstances. At the same time, the emerging biological importance of other long chain fatty acids, the so called ‘non-EFA’, is largely ignored. These and other problems with EFA are described here. Some of these issues were described in earlier reviews [1], [2]. Some originated many years ago and persist, while others are more recent. All contribute to a weak and unsatisfactory foundation from which to teach the nutritional importance of PUFA and from which to explore their biological functions.

Section snippets

A brief history of polyunsaturated fatty acids

There are several excellent reviews of the early research into nutritional and metabolic aspects of PUFA [3], [4], [5]. The late 1920s and early 1930s saw the first reports that dietary fats contained two vitamin-like substances soon identified as linoleate (18:2ω6) and α-linolenate (18:3ω3). These two ‘parent’ PUFA were jointly known for a while as vitamin F, a term abandoned in favour of EFA. The first 25 years or so of research on the nutritional importance of these two PUFA can be

Problem 1: too many or too few fatty acids qualify as essential fatty acids

People working on PUFA metabolism can be categorized as minimalists, inclusionists or pragmatists. The minimalists identify linoleate and α-linolenate as the only two EFA, because (i) they are the principal PUFA in the diet, (ii) they correct all known symptoms of PUFA deficiency (in rats), and (iii) they can be converted to all subsequent PUFA in their respective families. If so, there is a dilemma because there is clearly much support, including approval from most national government

Problem 2: non-essential fatty acids

Dealing with non-EFA is the hangover of having created the EFA. There are four aspects to this problem. First, PUFA that have no known function or requirement at any time in the life cycle should technically be considered non-EFA. The 22 carbon PUFA (ω6 docosapentaenoate [22:5ω6] and ω3 docosapentaenoate [22:5ω3]) serve as examples. They are EFA only by the inclusionist's definition because they are ultimately derived from the parent EFA, linoleate and α-linolenate, respectively. They have no

Problem 3: essential fatty acid deficiency is not the same as deficiency of ω6 polyunsaturates

A major oversight from the seminal work that established dietary PUFA requirements during the 1950s, and one which unfortunately remains with us to the present, is that no serious attempt was made to study the nutritional requirement for ω6 PUFA using diets deficient only in ω6 PUFA, i.e. using diets containing adequate amounts of ω3 PUFA. Surprisingly, amounts of dietary linoleate needed to achieve or restore normal growth and function have always been determined in animals that were given fat

Problem 4: α-linolenate reduces the requirement for ω6 polyunsaturates

An abundant literature demonstrates that raised intake of various ω3 PUFA reduces tissue content of ω6 PUFA, especially long chain ω6 PUFA. This literature is usually interpreted to imply that an increased risk of ω6 PUFA deficiency occurs when ω3 PUFA intake is elevated. I disagree that symptoms of ω6 PUFA deficiency have ever been reported under intakes of ω3 PUFA that could reasonably be expected to occur in nature. Mohrhauer and Holman [20] demonstrated that up to 1.8% of dietary energy of

Problem 5: preferential β-oxidation of eighteen carbon polyunsaturates

The rules governing dietary essentiality of a nutrient don't specify what that nutrient has to do, only that some symptoms must occur when the nutrient is absent from the diet, which can be prevented or corrected when the nutrient is present. In principle and from long experience, we expect essential nutrients to contribute to the structure or function of cells in some anabolic sense. All nutrients that are organic molecules are likely to undergo some degree of β-oxidation (hereafter oxidation)

Problem 6: greater tissue loss of polyunsaturates than saturated or monounsaturated fatty acids during fasting-refeeding and weight cycling

Whole body fatty acid balance studies bring to light the interesting point that various forms of undernutrition can markedly stimulate oxidation of the two parent PUFA. Undernutrition increases fatty acid oxidation, which is how ATP is regenerated when there is an inadequate supply of glucose. It is surprising, though, how actively linoleate and α-linolenate are utilized in this process. In fact, during fasting/refeeding or weight cycling, they are sufficiently oxidized that tissue PUFA

Problem 7: carbon recycling

For nearly 30 years now, tracer studies with 13C or 14C-labelled linoleate or α-linolenate have demonstrated that carbon from these two PUFA is actively incorporated into cholesterol and non-EFA in amounts that can easily exceed the amount incorporated into the respective long chain PUFA (reviewed in Ref. 31). The amount of carbon recycled into de novo lipogenesis varies with age, the tissue studied and the experimental conditions. Very different animal models show that, under normal rearing

Problem 8: synthesis of linoleate and α-linolenate from their sixteen carbon homologues

Even the minimalists adhere to designating linoleate and α-linolenate as EFA. The rationale is simple. Linoleate and α-linolenate cannot be synthesized de novo in mammals and they, or their long chain PUFA derivatives, have functions in the body not fulfilled by other compounds and for which symptoms occur when they are deficient in the diet. That is reasonable except that PUFA deficient rats can synthesize linoleate and α-linolenate by elongation from their sixteen carbon homologues,

Problem 9: more is not better

Few if any well-controlled studies have ever been done looking at linoleate requirements in humans. None have been done recently. The best estimates arising from experimental measurement (as opposed to epidemiological assessment of average linoleate intake) suggest that about 1% of energy intake as linoleate is sufficient to meet ω6 PUFA requirements in healthy adults [3], [4], [5]. Higher requirements are assumed to exist during pregnancy, lactation and early development but are still thought

The proposed new paradigm

In the 1960–1980s, the term ‘P/S ratio’ (PUFA to saturates) helped raise awareness of the differential effects that saturated, monounsaturated and polyunsaturated fatty acids in the diet have on several health outcomes. For instance, particular emphasis was placed on the relation between elevated dietary P/S ratio and reduction of serum cholesterol. Most of the time, the ‘P’ in the P/S ratio is linoleate. The P/S ratio is less common now in the technical literature on PUFA because, in light of

Acknowledgments

Financial support of research done in the author's lab and described here was received from NSERC, CIHR, Dairy Farmers of Canada, Unilever, Milupa, Martek, Flax Council of Canada, Epilepsy Canada, Stanley Thomas Johnson Foundation and Bloorview Children's Hospital Foundation. Mary Ann Ryan, Richard Bazinet, Kathy Musa-Veloso, Ursula McCloy, Chantale Menard, Jody Cole, Cynthia Dell, Jilin Yang, Zhen-yu Chen, Matt Anderson, Sergei Likhodii, Tom Brenna, and Michael Crawford did much of the work or

References (83)

  • R.T. Holman

    Prog Chem Fats Other Lipids

    (1971)
  • R.T. Holman

    J Nutr

    (1960)
  • H. Sprecher

    Biochim Biophys Acta

    (1968)
  • J.M. Bourre et al.

    J Nutr

    (1997)
  • H. Mohrhauer et al.

    J Lipid Res.

    (1963)
  • S.C. Cunnane et al.

    J Lipid Res.

    (1997)
  • S.M. Greenberg et al.

    J Nutr

    (1950)
  • R.P. Bazinet et al.

    J Lipid Res.

    (2003)
  • S.A. Roberts et al.

    Am. J. Clin. Nutr.

    (2001)
  • S.C. Cunnane et al.

    J Nutr

    (1997)
  • Z.Y. Chen et al.

    Metabolism

    (1993)
  • S.C. Cunnane et al.

    Am. J. Clin. Nutr.

    (2001)
  • S.C. Cunnane et al.

    Prostaglan Leukotr Essent Fatty Acids

    (1999)
  • S.C. Cunnane et al.

    J Lipid Res.

    (1998)
  • R.C. Sheaff-Greiner et al.

    J Lipid Res.

    (1996)
  • V. Wijendran et al.

    J Lipid Res.

    (2002)
  • J. Farquharson et al.

    Lancet

    (1992)
  • M. Makrides et al.

    Am. J. Clin. Nutr.

    (1994)
  • E. Vos et al.

    Am. J. Clin. Nutr.

    (2003)
  • L.W. Kinsell et al.

    Lancet

    (1957)
  • A. Poulos

    Prog Lipid Res.

    (1989)
  • C.K. Van Pelt et al.

    J Lipid Res.

    (1999)
  • P.J. Jones et al.

    Am. J. Clin. Nutr.

    (1985)
  • S.C. Vasdev et al.

    J Molec Cellular Cardiol

    (1977)
  • P. Bjorntorp

    J Biol Chem.

    (1968)
  • W.J. Bettger et al.

    J Nutr

    (1979)
  • G. Bruckner et al.

    J Nutr

    (1982)
  • S.D. Phinney et al.

    Metabolism

    (1993)
  • O.S. Privett et al.

    Am. J. Clin. Nutr.

    (1977)
  • J. Rafael et al.

    J Nutr

    (1984)
  • E. Aaes-Jorgensen et al.

    J Nutr

    (1958)
  • G.O. Burr et al.

    J Biol Chem.

    (1929)
  • R.T. Holman

    J Nutr

    (1960)
  • F.W. Quackenbush et al.

    J Nutr

    (1942)
  • A.R. Whorton et al.

    J Nutr

    (1977)
  • B.L. Walker

    J Nutr

    (1968)
  • B.L. Walker

    J Nutr

    (1967)
  • V.A. Ziboh et al.

    J Lipid Res.

    (1972)
  • R.O. Peluffo et al.

    Biochim Biophys Acta

    (1976)
  • S.C. Cunnane

    Can J Physiol Pharmacol.

    (1996)
  • S.C. Cunnane

    Br J Nutr

    (2000)
  • Cited by (0)

    View full text