2 to 4 mg [4.7 to 9.3 µmol]) of α-tocopherol and 55 g/day of fat (30 g from vitamin E-free lard) for 2.5 years. After the first 2.5 years, serum α-tocopherol levels decreased further when thermally oxidized corn oil with the α-tocopherol removed was substituted for lard. By replacing lard with corn oil, the total intake of polyunsaturated fatty acids (PUFAs) was increased, thereby increasing the oxidant burden on the available vitamin E stores. Subjects were followed on the vitamin E-depleted diet for more than 6 years.
To establish a criterion for estimating the EAR, the biomarker selected was the plasma α-tocopherol concentration that limited hydrogen peroxide-induced hemolysis to 12 percent or less. Differences up to this amount are not significant unless special precautions are taken to age and standardize the hydrogen peroxide solutions (Horwitt et al., 1963). The data in Figure 6-6 comparing long-term vitamin E-depletion in four subjects (depleted for more than 72 months) with six control subjects (Horwitt et al., 1963) show that at some concentration of plasma α-tocopherol between 6 µmol/L (258 µg/dL) and 12 µmol/L (516 µg/dL), an increase in hydrogen peroxide-induced hemolysis above 12 percent was observed in vitro. Averaging the α-tocopherol concentrations in the six subjects with hemolysis values of 12 percent or less in Figure 6-6 results in an average α-tocopherol concentration of 16.2 µmol/L (697 µg/dL). This is higher than the results of Farrell et al. (1977), who suggested that plasma α-tocopherol concentrations of 14 µmol/L (600 µg/dL) are sufficient to prevent hydrogen peroxide-induced hemolysis. Although the exact plasma α-tocopherol concentration that allows hemolysis to take place is unknown, it appeared to be prudent to estimate the lowest known plasma α-tocopherol concentration as that cutoff point where hemolysis would take place in 50 percent of the population. Thus, a plasma concentration of 12 µmol/L (516 µg/dL) was chosen as the concentration of plasma tocopherol associated with normal in vitro hydrogen peroxide-induced hemolysis. Based on NHANES III data (Appendix Table F-2), more than 95 percent of the population surveyed would have plasma concentrations greater than 12 µmol/L (516 µg/dL), thus indicating that the American public is not vitamin E deficient by this criterion.
The effect of vitamin E repletion of depleted subjects has been studied. Horwitt (1960) conducted a study in which each subject received a different amount of vitamin E supplement, ranging from