Vitamin E

BACKGROUND

Vitamin E is the name given to a group of water-insoluble, plant-derived substances. There are eight naturally-occurring isomers and a number of semisynthetic or synthetic homologues. The naturallyoccurring d- (or RRR) alpha-tocopherol is the most biologically active form and vitamin E activity is traditionally expressed in terms of equivalents of this isomer (mg alpha-tocopherol equivalents or α-TE). Other tocopherols such as gamma-tocopherol also have vitamin E activity. There are four tocopherol homologues (d-α-, d-β-, d-γ- and d-δ-) and four tocotrienols. Other forms of vitamin E occur in lower amounts in foods and are less active in animal bioassay. The usual form in supplements is synthetic dl- (or all-rac) α-tocopherol that consists of a mixture of active and inactive stereoisomers, because natural vitamin E from wheat germ oil is expensive. The equivalence of the various forms is shown below:

The major role of vitamin E is to protect polyunsaturated fatty acids (PUFA) from oxidation. It acts as an anti-oxidant in the lipid phase of cell membranes. Vitamin E requirements have been reported to increase when intakes of PUFA are increased (Dam 1962, Horwitt 1962) and a ratio of at least 0.4 mg α-tocopherol per gram of PUFA has been recommended (Bieri & Evarts 1973, Horwitt 1974, Witting & Lee 1975). Most dietary sources of polyunsaturated fat are also relatively rich in vitamin E. However supplements of fish oils or other pure n-3 fatty acids may not provide the amount of vitamin E needed. The activity vitamin E complements that of selenium-dependent glutathione peroxidase in protecting the membrane PUFAs from free radical damage. Although vitamin E is mainly located in cells and organelle membranes, its concentration may be very low, suggesting that after its reaction with free radicals it is rapidly regenerated, possibly by other antioxidants such as selenium, ubiquinols and vitamin C (Doba et al 1985, Niki et al 1982, Stoyanovsky et al 1995).

The main source of vitamin E is fats and oils. It is also found in some vegetables, in the fats of meat, poultry and fish and, to lesser degrees, in cereals and dairy foods. About half the tocopherol in wheat germ, sunflower, safflower, canola, olive and cottonseed oils is α-tocopherol but soybean and corn oils contain about 10 times as much γ-tocopherol as α-tocopherol. Most vitamin E is found in foods containing fat. Absorption requires micelle formation and chylomicron secretion in the gut (Muller et al 1974) together with biliary and pancreatic secretions. Efficiency of absorption is low, but the precise rate is unknown.

Vitamin E is transported in the blood by the plasma lipoproteins and erythrocytes. Tocopherols are carried from the gut to the liver in chylomicrons where they are incorporated as chylomicron remnants. Catabolism of chylomicrons takes place in the systemic circulation through the action of cellular lipoprotein lipase. Vitamin E can be transferred to high density lipoprotein (HDL) and then to low density liopoprotein (LDL) and very low density lipoprotein (VLDL). Most α-tocopherol enters peripheral tissues within the intact lipoprotein through the LDL receptor pathway. Although all tocopherol homologues are absorbed similarly, α-tocopherol predominates in blood and tissue as the binding proteins take it up preferentially. Plasma vitamin E and tissue concentrations vary little over a wide range of dietary intake and the brain is particularly resistant to depletion (Bourne & Clement 1991).

The main oxidation product of α-tocopherol is tocopheryl quinone which is conjugated to glucuronate and is excreted in bile or further degraded in the kidneys to α-tocopheronic acid before excretion in bile (Drevon 1991). Some may be excreted through the skin (Shiratori 1974).

Overt deficiency symptoms in normal individuals consuming diets low in vitamin E have never been described. It occurs only as a result of genetic abnormalities, fat malabsorption syndrome (Rader & Brewer 1993, Sokol 1993) or protein-energy malnutrition (Kalra et al 1998, Laditan & Ette 1982). The main symptom is a peripheral neuropathy. Other symptoms include spinocerebellar ataxia, skeletal myopathy and pigmented retinopathy (Sokol 1988).

In epidemiological studies, higher intakes of vitamin E have been related to reduction in cardiovascular disease risk (Gey et al 1991, Rimm & Stampfer 1993, Stampfer et al 1993), diabetic complications (Baynes 1991, Mullarkey et al 1990, Semenkovich & Heinecke 1997), certain cancers (Comstock et al 1997, Eichhlozer et al 1996, Yong et al 1997) and cataracts (Jacques & Chylack 1991, Knekt et al 1992, Leske et al 1991). Not all studies, however, have confirmed a relationship and clinical trials with supplements in high risk groups, have shown little benefit. Further discussion of these trials is given in the 'Chronic disease' section.

Indicators that have been used to estimate vitamin E requirements include lipid peroxidation markers, oxidation products of DNA or proteins, vitamin E metabolite excretion, vitamin E biokinetics, vitamin E deficiency symptoms, plasma α-tocopherol concentration, hydrogen peroxide-induced haemolysis or the relationship of vitamin E to chronic disease status. However, erythrocyte fragility studies have been the most widely used.

The US DRI review in 2000 used the data of Horwitt (1960, 1963). These same data had been used in setting the earlier US RDIs but were interpreted differently in 2000, leading to considerably increased recommendations. In the US DRI review of 2000, the amount of dietary vitamin E required to bring plasma α-tocopherol to a level where per cent haemolysis was low was used to estimate an EAR (Horwitt 1960, 1963). However, the interpretation of these data is problematic in relation to level of plasma α-tocopherol at which adverse effects are seen, as there were no data available for plasma α-tocopherol concentrations between 5 and 12 µmol/L. All four subjects below 6 µmol/L plasma α-tocopherol (range 2–5 µmol/L) had haemolysis of about 80% or above and all 6 subjects with concentrations between 12 and 22 µmol/L, had haemodialysis of 12% or less. There has been disagreement as to whether the 'adequacy' cut off should be midway between these two clusters or at the lowest point showing low haemolysis. The data are dichotomous, not continuous, thus preventing an accurate dose-response analysis. Changing the cut-off point makes a large difference to the estimated requirement. In addition, the authors of the key paper themselves expressed concern about the validity of the technique for assessing vitamin E requirements (Horwitt 1960, 1963, 2001).

Given these uncertainties, an AI rather than an EAR was set for vitamin E based on median population intakes in Australia and New Zealand – both healthy populations with no apparent vitamin E deficiency. Recommendations for infants were based on the median concentration in breast milk of healthy mothers.