based formulas do not (unless they have been supplemented).
Let's now review the research linking levels of EFAs in body tissues with EFA consumption via diet and/or supplementation.
Animal studies on impact of EFAs in diet: chickens. Anderson et al. [1990] studied the effects of n-3 fatty acid (FA) supplementation in newly hatched chickens with an n-3 FA deficiency. Levels of DHA were measured in the brains and retina of deficient chicks fed 4 diets containing different levels of n-6 FA. The chicks fed a diet that included DHA and EPA recovered--that is, the levels of DHA/EPA returned to normal levels after 3 weeks. The researchers concluded that linolenic acid as the sole source of n-3 fatty acids is inadequate, and supplementation with DHA is advisable. Anderson and Connor [1994] provides additional insight into this research.
EFA studies on rats. Woods et al. [1996] tested the hypothesis that the "right" ratio of LA/LNA (LA = linoleic acid, an n-6 fatty acid; LNA = alpha-linolenic acid, n-3) in infant formula could support the same levels of DHA production as results from breast-feeding (in rats). A series of milk formulations with LA/LNA ratios of 10:1, 1:1, and 1:12 were tested on infant rats, and the resulting levels of DHA in the rat's brains were measured. Woods et al. [1996, pp. 691, 693] conclude that:
This suggests that there is no level of LNA that is capable of supporting the [levels of] neural DHA found in the dam-reared [breast-fed by mother rats] rat pups...
If it is accepted that human essential fatty acid metabolism is not more efficient than that of the rat, it follows that infant formulas like those in our study would not support proper human nervous system accretion of LC-PUFAs [long-chain polyunsaturated fatty acids].
Woods et al. [1996] do suggest that LNA might assist DHA synthesis in the brain, but not the retina, of the rats.
EFA studies on monkeys. Lin et al. [1994] studied the effect of various n-3 fatty-acid levels in the diet on the retinas of rhesus monkeys. Monkeys on deficient diets, when fed a diet that included DHA, recovered to near-normal levels, except for certain lipids. In the n-3 FA-deficient monkeys, docosapentaenoic acid (DPA, 22:5n-6, a long-chain derivative in the n-6 FA family) partially replaced DHA in the monkeys' retinas. However, Lin et al. [1994, abstract] describe the replacement as "functionally incomplete," suggesting that the resultant DPA levels were too low, and/or did not provide equivalent functional support for the monkeys' visual systems (in comparison to the monkeys whose diet included DHA).
EFA levels in "living foods" adherents (raw vegans). Agren et al. [1995] studied the EFA levels of vegans following the living foods diet, a vegan diet that emphasizes sprouted and fermented raw or "living" foods. The study began with 12 people who claimed to be following the living foods diet, but two were excluded because they were found not to be actually following the diet (it is a difficult and idealistic diet that few can follow successfully, in the long-term), and another subject was excluded who displayed fatty-acid levels that were 4-5 standard deviations from the means observed (again suggesting probable non-compliance with the claimed diet).
Comparison of the EFA levels in erythrocytes, platelets, and serum lipids in living foods raw vegans vs. controls (standard Western diet) showed that the EPA levels in the raw vegans were only 29-36%, and DHA levels 49-52%, of the analogous levels in the control group. Additionally, the n-6/n-3 ratio in the raw vegans (5.55) was about double the ratio observed in the controls (2.70). (Note: Agren et al. [1995, Table 2, p. 367] report n-3/n-6 ratios; the n-6/n-3 ratios in the preceding sentence were obtained by inverting the ratios reported in Agren et al. [1995] for erythrocyte totals.)
EFA levels in conventional vegans. Krajcovicova-Kudlackova et al. [1997] analyzed EFA levels in blood plasma, comparing several groups: vegans, vegetarians (whether lacto- or lacto-ovo-vegetarian was not specified), semi-vegetarians (those who ate fish) and a control group following the standard Western diet. The vegan group had significantly lower levels of both EPA (62.5%) and DHA (67%) compared to the control group, while the levels of EPA and DHA were not significantly different between vegetarians and the control group (computed from Krajcovicova-Kudlackova et al. [1997, Table 2, p. 367]). N-6/n-3 ratios were found to be 19.48 for vegans, 14.71 for vegetarians, and 13.07 for the control group.
Note that the n-6/n-3 ratios for vegans and vegetarians are different (statistically significant, P = 0.001) from the control (standard Western diet) group. Although the difference between the vegetarian group vs. control group is statistically significant, in real-world terms the amount of the difference might not be very meaningful, as 14.71 and 13.07 are "close." However, the corresponding difference between the control group and vegans, whose ratio is 19.48, is considerably different enough that the result is likely to be meaningful in real-world terms as well.
Impact of DHA supplements on EFA levels in vegetarians. Conquer and Holub [1996] studied the effect of DHA supplementation on the EFA levels in phospholipids, in vegetarians. (Phospholipids are compounds of lipids with alcohols, which also have a phosphate residue [Ansell 1973]. Lecithin is an example of a phospholipid [Strickland 1973]. Phospholipids are present in blood platelets, serum, and elsewhere in the human body.)
Two volunteer groups of vegetarians consumed either capsules providing 1.62 gms/day of DHA, or (control group) a similar quantity of corn oil capsules (placebo, no DHA content). The DHA in the capsules was from algae rather than animal sources. It was found that consumption of DHA capsules had a dramatic effect on DHA levels in phospholipids (increase of 225% in blood platelet, 246% in serum, phospholipids) and EPA levels (176% in blood platelet, 117% in serum, phospholipids). Additionally, the consumption of DHA capsules significantly lowered the n-6/n-3 ratio after 3 weeks (from 8.9 to 3.3); ratios were 3.1-3.3 for those taking DHA, 9.5-9.9 for those on the corn oil placebo. The consumption of DHA also improved a number of biomarkers associated with heart disease (cholesterol, triglycerides; see the study for details).
EFA levels in human milk. Sanders and Reddy [1992] compared the levels of EFAs in breast milk from women who followed vegan, vegetarian, and standard Western diets (the latter serving as control group). Levels of DHA were found to be lowest in the milk from vegan mothers (37.8% of control group level), highest in the standard diet, with (lacto-)vegetarians in the middle (81% of control group level). The study authors note [Sanders and Reddy 1992, p. S71]:
The proportion of DHA in erythrocyte total lipids of infants breast-fed by vegans was 1.9% compared with 3.7% in infants fed a [cow's] milk formula containing butterfat as the sole source of fat and 6.2% in infants breast-fed by omnivores [standard Western diet] at 14 weeks postpartum.
That DHA levels in infants were lower in those who were breast-fed by vegan mothers than in those fed the cow's milk formula is somewhat surprising in light of the data that the vegan human milk contained more DHA as a percentage of energy. This may be caused by (relatively) high levels of linoleic acid (n-6) in the vegan breast milk suppressing synthesis of DHA. Sanders and Reddy [1992, Table IV, p. S74] report n-6/n-3 ratios of: 16.4 for vegans, 13.0 for vegetarians, 13.5 for control. (Note: Sanders and Reddy report that the ratio is higher for vegans, but the statistical significance thereof is not clear in their paper.) For a related paper, see Reddy et al. [1994].
EFA levels in tissues of infants. Farquharson et al. [1992] analyzed the brain tissues of infants who died of crib death, also referred to as "sudden infant death syndrome" (SIDS). They found significantly higher levels of DHA and DPA in the cerebral cortex tissues of breast-fed infants vs. those fed formulas (that were not supplemented with DHA; breast milk contains DHA). They note that the higher linoleic acid content of the formulas might inhibit synthesis of DHA.
Carlson et al. [1986] compared the EFA levels in blood cell phospholipids of preterm infants, for those receiving human milk vs. those fed with formula. Feeding with preterm human milk increased phospholipid DHA levels, while feeding with formula caused a decline.
Salem et al. [1996] provides proof that infants are able to synthesize, in vivo, essential fatty acids from precursors. Also noted, however, is that the level of DHA synthesis is inadequate to meet needs in early life.
Connor et al. [1996] studied the effect on DHA levels of supplementing the diet of pregnant women with fish oil capsules (high in DHA and EPA) in the latter stages of pregnancy. They found that blood DHA levels were higher in the blood of infants whose mothers consumed the fish oils. In the supplemented infants, DHA levels were 35.2% higher in red blood cells, and 45.5% higher in plasma, vs. the non-supplemented controls.
EFAs and infant development. Gibson et al. [1996] discusses a number of studies that show improved visual function in breast-fed infants vs. those receiving formulas. The authors relate this to the presence of DHA in breast milk, and the lack (or lower levels) thereof in formulas. Also see Hoffman et al. [1993] and Uauy et al. [1996] in this regard.
Low conversion rates of LNA to DHA. Nettleton [1995] reports that alpha-linolenic acid (LNA) can be converted to EPA and DHA, but the efficiency of the conversion is low [Dyerberg, Bang, and Aagard 1980; Sanders and Younger 1981; as cited in Nettleton 1995]. Emken et al. [1994], as cited in Conquer and Holub [1996], report a conversion rate of LNA to DHA of ~5% in adults. Salem et al. [1996] suggest (Table 1, p. 50; see also p. 51) a minimum conversion rate of ~1% from LNA to DHA in infants (0.9 mg of DHA produced from 100 mg precursor).
Flaxseed oil a suboptimal source for n-3 acids. Flaxseed oil is used by some vegans as a source of n-3 oils. However, the study of Kelley et al. [1993] fed volunteers a diet in which 6.3% of calories were from flaxseed oil. (Given the cost of flaxseed oil, this is a large amount and would be expensive.) They observed a statistically significant increase in EPA levels in peripheral blood mononuclear cells (PBMNC) in those receiving flaxseed oil. EPA levels in serum were not affected by flaxseed oil. Regarding DHA levels, no increases (in PBMNC or serum) were seen from flaxseed oil supplementation.
Gerster [1998] is a review article that looks into the issue of whether adults can adequately convert ALA to EPA and DHA. The tables in Gerster [1998, pp. 165-166, 168] list a number of studies that showed a pattern similar to the one described above. That is, flaxseed oil supplementation may increase EPA, but not DHA.
- Studies that found no changes in DHA levels from flaxseed oil supplementation: Dyerberg et al. [1980], Singer et al. [1986], Kelley et al. [1993], Cunnane et al. [1995], Mantzioris et al. [1994], Layne et al. [1996].
- Studies that reported DHA decreased with flaxseed oil supplementation: Sanders and Roshanai [1983], Allman et al. [1995].
- Studies that reported DHA increased with flaxseed oil supplementation: Kestin et al. [1990], and Sanders and Younger [1981]. Note that the latter study found an increase in DHA for vegan subjects only, in plasma glycerides, but not in platelets (they found no differences in DHA in the non-vegan control group).
The weight of the studies above suggest that flaxseed oil does not provide efficient support for DHA synthesis.
Preformed DHA may be necessary for optimum levels. Gerster also makes the point that fish oils are superior sources for EPA and DHA, and concludes [1998, pp. 159-160]:
These findings indicate that future attention will have to focus on the adequate provison of DHA which can reliably be achieved only with the supply of the preformed long-chain metabolite.
For another perspective, Nettleton reports [1995, pp. 33-34]:
The fact that LNA is less effective than EPA and DHA in enriching tissues with n-3 FA [fatty acids] means that foods rich in LNA, vegetables and seed oils will be less satisfactory sources of n-3 FA for human health than seafoods or other animal foods enriched with EPA and DHA. That is not to say, though, that oils such as canola and soy are not useful sources of n-3 FA...
Finally, consuming oils with LNA helps reduce the amount of linoleic acid we consume. In rats, the ratio of n-3 to n-6 fatty acids consumed is more effective than the absolute amount of n-3 FA in inhibiting eicosanoid synthesis from arachidonic acid (Boudreau et al. 1991). It may be the same for people also...
Tissues that possess the enzymes needed to convert one form of n-3 FA to another will get by with available precursors, but others may require particular n-3 FA preformed. Dietary sources providing both EPA and DHA would appear to have advantages over those supplying only LNA.
Effects of balance of fatty acids. Langley [1995] reports that typical vegan diets have a high intake of linoleic acid, which can suppress the production of DHA, i.e., the typical vegan diet has a poor balance of the EFAs. The effect of the balance of fatty acids is also discussed in a number of the studies cited above.
The remarks above concerning fatty acid balance are cogent, as a number of companies strongly promote (and some vegan "diet gurus" recommend) flaxseed and/or hempseed oils as supplements to vegans to "make sure" they are getting the right balance of fatty acids in their diet. One must wonder how "natural" vegan diets are, if exotic, expensive oils are "required" to provide the right balance of linoleic and alpha-linolenic fatty acids in a vegan diet. If vegan diets are truly "natural," one would intuitively expect it to be relatively easy, as a vegan, to get the optimal balance of fatty acids.
Leaving aside the criterion of naturalness, however, if (typical) vegan diets are held instead simply to be the "optimum," one would not expect the balance of fatty acids in a typical vegan diet to potentially suppress the synthesis of critically important EFAs, like DHA. The studies discussed above, however, raise serious questions on this point. In addition, it appears that despite flaxseed oil supplementation, for example, it may still be difficult to achieve the proper level of DHA.
The studies above also show a strong link between DHA in the diet and DHA levels in body tissues, with vegans having very low levels of DHA when compared to those whose diet includes preformed DHA. This raises the question (discussed below) of whether the levels of DHA and EPA synthesis are optimal or merely adequate to prevent deficiency symptoms.
Positive evidence suggesting DHA/EPA-synthesizing enzymes may be adequate for complete nutrition. Okuyama [1992] presents limited evidence that the desaturation-elongation enzyme activity is adequate to supply 20-carbon fatty acids like EPA and DHA. Okuyama [1992, p. 170] reports:
Finally, it should be emphasized that strict vegetarians ingest only linoleate and alpha linoleate, and usually do not develop essential fatty acid deficiency (8).
All these data suggest that humans are no exceptions among mammals in the metabolism of linoleate and alpha linoleate, although the desaturation-elongation activities may be relatively lower in most humans probably due to the negative feedback control of the enzyme systems by large amounts of highly unsaturated fatty acids in their diets (9).
