Diagnosing Essential Fatty Acid Deficiency
To the Editor:
Dr Gould and colleagues reported the effects of an intensive cholesterol-lowering program using mostly low-fat diets.1 To monitor essential fatty acid (EFA) status, they measured the ratio 20:3ω9/20:4ω6 (triene/tetraene or T/T ratio). The authors stated that a ratio >0.4 in serum phospholipids indicates EFA deficiency (p 1532). In 1987, Siguel et al2 3 showed that measures of T/T currently in use were in error by a factor ≥10. The use of old technology, which had inadequate peak separation and erroneous peak integration, led to huge errors in measuring 20:3ω9. When 20:3ω9 was properly measured, T/T ratios >0.02 (in whole plasma) indicated EFA deficiency. For healthy people, T/T ratios in red blood cells (RBCs), RBC phospholipids, and serum phospholipids, which reflect the subject’s plasma levels, give reference values that are >10 times below the values reported by Dr Gould (Siguel, unpublished data, 1997). The T/T reference values used by Gould et al appear to be too high and would account for the statement that “no patient showed evidence of EFA deficiency.” If we seek to detect abnormal cholesterol using a method that can only measure cholesterol >1000 mg/dL, most patients would be found to have normal cholesterol, and saturated fat and obesity would be found not to increase cholesterol.
Patients are unlikely to develop the severe EFA deficiencies that would have dramatic symptoms, such as substantial hair loss, overt clinical dermatitis, and T/T ratios >0.4; this would require severe depletion of EFAs over a period of many years. However, biochemical evidence of EFA deficiency can be detected in patients after several months on a low-fat diet when appropriate technology is used. Insufficient levels of EFAs, a condition I characterized as EFA insufficiency (EFAI),4 and imbalance of the ω3/ω6 ratios is associated with abnormal lipid levels,5 hypertension, coronary artery disease, and a higher-than-average probability of premature death as a result of heart disease.6 Patients who routinely follow low-fat diets deprived of EFAs are likely to develop EFAI and ought to be warned about the long-term consequences of EFA deficiency. Researchers can now monitor EFA status quite accurately and measure the biochemical onset of EFA deficiency over the duration of a study.
These issues have far greater implications for public health than a mere technical discussion on reference levels. The existence of EFA abnormalities affects both medical and food policy in the United States. Failure to diagnose EFA abnormalities may lead to expensive treatments for preventable chronic diseases followed by unnecessary premature death. Billions of dollars are now spent on drug treatments and surgical procedures for conditions like high blood pressure, abnormal cholesterol ratios, and coronary artery disease, which respond quite well to correction of polyunsaturated fatty acid (PUFA) abnormalities.4
Corporations are now developing a wide range of fat substitutes and fat replacements. Replacing fat with fat substitutes (or substances that inhibit fat absorption) obviously reduces the amount of fat in the diet, which may help some patients, but also significantly reduces the amount of EFAs in the body, which can have disastrous consequences.
One obvious concern about the safety of these new products has been the likelihood that these food substitutes cause substantive EFA deficiency in the population. Companies engaged in the development and marketing of food substitutes would like to prove that their fat substitute (or related product) is healthful and does not lead to EFA deficiency. It is quite simple to prove that EFA deficiency does not exist by measuring EFA deficiency with a technology that can only detect severe EFA deficiency. By publishing articles, journals must be careful that they do not endorse, unintentionally, use of a technology that cannot detect EFA deficiency except in very rare and exceptional cases. Corporations could use this published information to state that subjects using their new products did not experience EFA depletion and that they used the same methods of measurement reported in the peer-reviewed journal. This is not a merely hypothetical issue; I have been told of at least one company that plans to show that its product does not cause EFA deficiency because the T/T ratios of subjects remained <0.4, citing articles that rely on outdated technology to support their contention.
A similar concern arises with regard to drug treatment for heart disease, particularly when combined with a diet high in monounsaturated fatty acids (MUFAs), as proposed by many public health organizations and popular books such as The Zone.7 For example, the Heart Owner’s Handbook states that MUFAs are better than PUFAs (p 33).8 However, MUFAs obviously cannot replace the PUFAs missing in low-fat foods. These matters are critical now that many new food products are either under government review or already approved for consumer use. Failure to use a sensitive blood test would obscure EFA deficiency caused by the extensive use of low-fat foods or foods high in MUFAs. Failure to diagnose and treat PUFA abnormalities can cause unnecessary premature death.9
Editors who publish articles involving the analysis of fatty acids should insist that authors provide the following information and reviewers be qualified to understand it: column used (ID, phase, length, etc), gas-liquid chromatograph (GLC) (brand model), injection method (split, splitless, etc), amount of injection, how fatty acid methyl esters were prepared, GLC method (eg, temperature run, duration, gases used, flows), how samples were integrated, software used (model, version, etc), quality control for peak integration, how the area of each peak was evaluated for integration errors, and how peaks were identified. Articles should include (for review, if not for publication) at least 2 representative chromatograms for subjects in each group showing peaks and integration: 1 that shows all peaks and 1 enlarged enough to see the baseline and noise levels. This is needed to evaluate the accuracy of the integration and the presence of artifacts or extra peaks that would affect the results. These data are simple to produce and readily available to the authors.
(Dr Siguel has a patent on a blood test to diagnose fatty acid abnormalities.)
- Copyright © 1998 by American Heart Association
Gould KL, Martucci JP, Goldberg DI, Hess MJ, Edens RP, Latifi R, Dudrick SJ. Short-term cholesterol lowering decreases size and severity of perfusion abnormalities by positron emission tomography after dipyridamole in patients with coronary artery disease: a potential noninvasive marker of healing coronary endothelium. Circulation.. 1994;89:1530–1538.
Siguel EN, Chee KM, Gong J, Schaefer EJ. Criteria for essential fatty acid deficiency in plasma as assessed by capillary column gas-liquid chromatography. Clin Chem.. 1987;33:1869–1873.
Siguel E. A new relationship between polyunsaturated fatty acids and total/HDL cholesterol. Lipids.. 1996;31:S51–S56.
Siscovick DS, Raghunathan TE, King I, Weinmann S, Wicklund KG, Albright J, Bovbjerg V, Arbogast P, Smith H, Kushi LH, Cobb LA, Copass MK, Psaty BM, Lemaitre R, Retzlaff B, Childs M, Knopp RH. Dietary intake and cell membrane levels of long-chain n-3 polyunsaturated fatty acids and the risk of primary cardiac arrest. JAMA.. 1995;274:1363–1367.
Sears B. The Zone. New York, NY: Harper Collings Publishers, Inc; 1996.
Texas Heart Institute. The Heart Owner’s Handbook. New York, NY: John Wiley & Sons; 1996.
In his letter about our study published in Circulation 3 years ago,R1 Dr Siguel raises essentially two concerns: (1) how to assay blood for essential fatty acid deficiency and (2) whether low-fat diets cause essential fatty acid deficiency. Both are addressed separately below.
There are two essential fatty acids, linoleic acid and alpha-linolenic acid, required for synthesis for arachinoids, eicosanoids, prostaglandins, thromboxanes, and leukotrienes that are important for cell membrane function.R2 R3 R4 With essential fatty acid deficiencies, an “abnormal” intermediate, eicosatrienoic acid (20:3ω9), accumulates and the “normal” intermediate metabolite, arachidonic or an eicosatetranoic acid (20:4ω6), decreases such that the triene to tetraene ratio (20:3ω9/20:4ω6) or the T/T ratio increases.R2 R3 R4
Current texts of nutrition indicate that a T/T ratio of ≥0.4 is associated with clinical linoleic acid deficiency,R3 R4 a criterion used in our 1993 study. Table 3 (page 666) of Dr Siguel’s 1987 study in the journal Metabolism shows an average T/T ratio of 0.45 in six patients with clinical essential fatty acid deficiency as the result of malabsorption syndromes or prolonged fat-free parenteral alimentation.R5 A normal reference group had a T/T ratio of 0.008, and a subset of subjects from the Framingham Study had values of 0.014. A patient with essential fatty acid deficiency corrected to a T/T ratio of 0.02, in the normal range, after intravenous infusion of adequate essential fatty acids.
In his second 1987 study,R6 Dr Siguel reports a T/T ratio in a normal reference group as 0.013±0.006. In 10 patients with clinical malabsorption syndromes, the T/T ratio was 0.30±0.61, with a value of P=.06 for the difference, not statistically significant at the .05 level. The normal T/T ratio in older literature is reported as 0.1±0.07.R6 On the basis of these data, Dr Siguel claims in his letter that “measures of T/T currently in use were in error by a factor ≥10.” With his average value of the T/T ratio of 0.45R5 and 0.3R6 in patients with essential fatty acid deficiency by new technologyR5 compared with 0.4 used traditionally,R3 R4 his basis for this claim is questionable, particularly since his own value of 0.3 was not significantly different from his reference group at the P=.05 level of probability.R6 His normal reference values for the T/T ratio of 0.013 compared with older literature of 0.1R6 do not redefine the criteria for clinical fatty deficiency, which are 0.3 to 0.45 by his own numbers, the same as traditionally used. Therefore, the criteria for our 1993 article in Circulation are the same as those traditionally usedR3 R4 and consistent with Dr Siguel’s own data.R5 R6
In a publication of 1994,R7 Dr Siguel proposed a new intermediate stage of essential fatty acid insufficiency (as opposed to deficiency) with a number of other end point measures proposed as being more sensitive than the T/T ratio as indicators of essential fatty acid insufficiency. On the basis of fatty acid analyses in 47 patients with documented coronary artery disease, he reasoned that excessive dietary levels of saturated fatty acids saturate or disturb the transport of essential fatty acids, thereby producing a possible abnormal T/T ratio or other proposed biochemical markers of possible essential fatty acid deficiency despite apparently relatively normal diets in these patients. In this study,R7 the T/T ratio in the normal reference population was 0.013±0.001 and in the coronary artery disease patients was 0.016±0.001, a statistically significant difference of unclear biological importance, particularly in view of a T/T ratio of 0.3±0.61 in patients with clinical essential fatty acid deficiency that was not significantly different from normal at the P=.05 level in his previous study.R6 By his T/T ratio criteria in the 1994 study,R7 5 of 47 patients were claimed to have deficiency of essential fatty acids without reference to the type of diet the patients were on. Dr Siguel then extrapolated this point of view to the more extended hypothesis that a low-fat diet causing essential fatty acid insufficiency may increase the risk of coronary artery disease.R7
In a 1996 studyR8 and in his letter, this point of view is still further extrapolated to the more extended hypothesis that low-fat diets cause essential fatty acid deficiency and heart disease and that “billions of dollars are now spent on drug treatments and surgical procedures for conditions like high blood pressure, abnormal cholesterol ratios, and coronary artery disease, which respond quite well to correction of PUFA abnormalities.” His letter goes on to indict the low-fat food industry and “public health organizations” and concludes that “failure to diagnose and treat PUFA abnormalities can cause unnecessary premature death,” where “correcting PUFA abnormalities” apparently does not mean low-fat food or cholesterol-lowering drugs. The data documenting these extended hypotheses are not reported, to my knowledge. In fact, high intake of linoleic acid of >12% of calories reportedly decreases HDL cholesterol that is associated with higher risk of atherosclerosis.R3 R9 R10
Reading the 1987R5 R6 and 1994R7 articles of Dr Siguel suggests that he knows a great deal about measuring fatty acids and about their possible interactions, with thoughtful, provocative hypotheses about their potential biological importance. However, his own data do not support the claim of an error by a factor of 10 in the criteria of the T/T ratio indicating clinical essential fatty acid deficiency. His data actually support the current traditional criteria. The reader and Dr Siguel need to clearly separate documented criteria and conditions for clinical essential fatty acid deficiency supported by Dr Siguel’s own dataR5 R6 from a hypothesized, intermediate stage of essential fatty acid insufficiency in free-living people on undefined diets, on the basis of changes in fatty acid ratios of unknown significance, without defined clinical abnormalities, due to a hypothesized adverse effect on lipid profiles, by an undefined level of dietary polyunsaturated fatty acids.R7 R8 These latter provocative hypotheses are just that—unconfirmed hypotheses. Furthermore, these hypotheses involve the effects of substantial amounts of dietary polyunsaturated fatty acids, particularly linoleic acid and alpha-linolenic acid, on lipid metabolism, not clinical deficiency of these essential fatty acids in the diet of well individuals without malabsorption syndromes or parenteral alimentation.
As to the second issue, do low fat-diets cause essential fatty acid deficiency? The Recommended Dietary Allowances (RDA) published by the National Research Council do not make recommendations on minimum daily requirements of essential fatty acids. The reason is a lack of data and the difficulty of identifying essential fatty acid deficiency in a free-living population other than patients with malabsorption syndromes or total parenteral alimentation. The United Kingdom Reference Nutrient Intakes suggest minimum consumption of 1% of calories from linoleic acid and 0.2% from alpha-linolenic acid.R3 R4 For a 1500-calorie diet commonly needed to achieve lean body habitus, ≈1.67 g of linoleic acid and 0.33 g of alpha-linolenic acid would be minimum requirements on the basis of these criteria. For an 1800-calorie diet, 2 g of linoleic acid and 0.4 g of alpha-linolenic acid would be the minimal requirements by these criteria.
Soybean oil, walnut oil, corn oil, grapeseed, sunflower, and cottonseed oil contain >50% linoleic acid. Soybean oil, walnut oil, canola, and rapeseed oil contain 7% to 11%, and linseed oil contains 53% alpha-linoleic acid.R11 R12 R13 R14 R15 Only soybean and walnut oils have significant balanced amounts of both linoleic acid and alpha-linolenic acid in proportion to their essential requirements. Alternatively, the compounds synthesized from alpha-linolenic acid, docosahexanoic (DHA) and eicosapentenoic (EPA), can be acquired from seafood, specifically omega-3 fish oil. In view of the ubiquity of the two essential fatty acids in Western diets, essential fatty acid deficiency is considered rare in free-living adults if it occurs at all,R3 R4 being seen only and uncommonly in severe malnutrition, untreated malabsorption syndromes or prolonged incomplete fat-free intravenous alimentation.
However, I have seen 14 lean patients on self-imposed, well-documented diets of <5% of calories as fat for 1 to 3 years (less than my guidelines of 10% of calories as fat) who had clinical symptoms consistent with essential fatty acid deficiency. These symptoms included mild but definite temporary recent memory loss, difficulty concentrating, episodic somnolence during the day, visual scotoma, decreased visual acuity, and/or sexual dysfunction. One, several, or all of these symptoms first appeared on very low-fat diets and reverted to normal on increasing sources rich in essential fatty acids. In addition to these symptoms, 4 of these 14 subjects had their first episode of atrial fibrillation or tachyarrhythmia without alcohol exposure or identifiable cause and without recurrence after increasing essential fatty acid intake. Nine of these subjects had their first episode of overt vasovagal syncope preceded by symptoms of similar but less severe vasovagal reactions, all of which disappeared on increasing sources of essential fatty acids without other identifiable changes in food, weight, or medications. Two individuals developed flaking, soft nails that became normal after increased intake of essential fatty acid sources. None had the typical fatigue, elevated triglycerides, and low HDL commonly caused by excess carbohydrate and inadequate protein previously discussed in “Letters to the Editor” of JAMA in response to Dr Siguel.R16 Therefore, clinically relevant essential fatty acid deficiency may occur in otherwise well-nourished, active, well individuals on very low-fat diets of <5% of calories as fat. Any treatment powerful enough to heal is also powerful enough to harm if misused or if its side effects are not understood. Very-low-fat food also has similar beneficial or bad potential and must be implemented properly with no less than 10% of calories as fat.
For individuals adherent to very-low-fat foods, I recommend specific sources of essential fatty acids. Soybean oil contains 51% linoleic acid and 6.8% alpha-linolenic acid.R11 R12 R13 R14 R15 One teaspoon of soybean oil containing ≈4.7 g of oil provides 2.4 g of linoleic acid and 0.38 g of alpha-linolenic acid, the minimum requirements to prevent clinical essential fatty acid deficiency on the basis of the criteria of the United Kingdom Reference Nutrient IntakesR3 R4 and consistent with discussion of the recommended dietary allowances of the US National Research Council.R2 One form of soybean oil is soya lecithin,R15 containing 1.2 g of oil per capsule, so that 4 capsules per day would provide this minimum requirement.
One teaspoon of walnut oil provides 2.4 g of linoleic acid and 0.25 g of alpha-linolenic acid. One to two teaspoons of walnut oil ingested per day provide the minimal requirements for essential fatty acids. It is the only oil besides soybean oil with balanced equivalent amounts of both linoleic and alpha linolenic acids in approximate proportion to their essential requirements.
One teaspoon of corn oil, grapeseed, sunflower, or cottonseed oil provides the same amount or more of linoleic acid, since all contain >50% linoleic acid. However, these oils lack the alpha-linolenic acid. Alpha-linolenic acid can be obtained from a teaspoon of linseed oil, rapeseed oil, or canola oil, containing ≈2.4, 0.5, and 0.5 g of alpha-linolenic acid, respectively. Alternatively, instead of alpha-linolenic acid, fish consumption or fish oil could be used as sources of EPA and DHA instead of oils containing alpha-linolenic acid from which the body synthesizes these compounds.
In the study by Siscovick et al,R17 quoted by Dr Siguel, one fish meal per week consisting of 5.5 g of N-3 fatty acids per month was associated with a 50% decrease in risk of primary cardiac arrest in patients without known heart disease after adjustment for confounding other risk factors. This level of intake is the equivalent of only 0.2 g of fish oil per day, not the “high intakes” of PUFA that Dr Siguel recommends in his 1996 studyR8 and implies in his letter. Although the Siscovick studyR17 has limitations, it suggests that very small amounts of fatty acids may be beneficial.
To keep this discussion with Dr Siguel in perspective, it is important to emphasize that low-fat or very-low-fat food, with adequate essential fatty acids, protein, and other nutrients and cholesterol-lowering drugs have been documented to stabilize or partially reverse coronary and cerebrovascular atherosclerosis with a profound decrease in cardiac and cerebrovascular events. Inadequate lowering of dietary fat or serum cholesterol in the management of patients with coronary artery disease because of extended, unvalidated hypotheses about potential essential fatty acid deficiency could be detrimental to optimal patient management.
Gould KL, Martucci JP, Goldberg DI, Hess MJ, Edens RP, Latifi R, Dudrick SJ. Short-term cholesterol lowering decreases size and severity of perfusion abnormalities by positron emission tomography after dipyridamole in patients with coronary artery disease: a potential noninvasive marker of healing coronary endothelium. Circulation. 1994;89:1530–1538.
Subcommittee on the 10th Edition of the RDAs, Food and Nutrition Board Commission on Life Sciences National Research Council. Recommended Dietary Allowances. 10th ed. Washington, DC: National Academy Press; 1989.
Sanders T. Essential fatty acids. In: Macrae R, Robinson RK, Sadler MJ, eds. Encyclopaedia of Food Science, Food Technology, and Nutrition. Academic Press; 1993:1651–1654.
Bender DA, Bender AE. Nutrition. A Reference Handbook. Oxford University Press; 1997:131–133.
Siguel EN, Maclure M. Relative activity of unsaturated fatty acid metabolic pathways in human. Metabolism. 1987;36:664–669.
Siguel EN, Chee KM, Gong J, Schaefer EJ. Criteria for essential fatty acid deficiency in plasma as assessed by capillary column gas-liquid chromatography. Clin Chem. 1987;33:1869–1873.
Siguel EN, Lerman RH. Altered fatty acid metabolism in patients with angiographically documented coronary artery disease. Metabolism. 1994;43:982–993.
Siguel E. A new relationship between total/high density lipoprotein cholesterol and polyunsaturated fatty acids. Lipids. 1996;31:S-51–S-56.
Mensink RP, Katan MB. Effect of dietary fatty acids on serum lipids and lipoproteins. Arterioscler Thromb. 1992;12:911–919.
Dupont J. Vegetable oils. In: Macrae R, Robinson RK, Sadler MJ, eds. Encyclopaedia of Food Science, Food Technology and Nutrition. Academic Press; 1993:4711–4713.
Snyder HE. Soya beans. In: Macrae R, Robinson RK, Sadler MJ, eds. Encyclopaedia of Food Science, Food Technology and Nutrition. Academic Press; 1993:4215–4218.
Hammond EG, Johnson LA, Murphy PA. Soya beans: properties and analysis. In: Macrae R, Robinson RK, Sadler MJ, eds. Encyclopaedia of Food Science, Food Technology and Nutrition. Academic Press; 1993:4223–4225.
Prasad RBN. Walnuts and pecans. In: Macrae R, Robinson RK, Sadler MJ, eds. Encyclopaedia of Food Science, Food Technology and Nutrition. Academic Press; 1993:4828–4834.
Szuhaj BF. Phospholipids. In: Macrae R, Robinson RK, Sadler MJ, eds. Encyclopaedia of Food Science, Food Technology and Nutrition. Academic Press; 1993:3553–3558.
Gould KL. Letter to the editor. JAMA. 1996;275:1402–1403.
Siscovick DS, Raghunathan TE, King I, Weinmann S, Wicklund KG, Albright J, Bovbjerg V, Arbogast P, Smith H, Kushi LH, Cobb LA, Copass MK, Psaty BM, Lemaitre R, Retzlaff B, Childs M, Knopp RH. Dietary intake and cell membrane levels of long-chain n-3 polyunsaturated fatty acids and the risk of primary cardiac arrest. JAMA. 1995;274:1363–1367.