countered only with very high doses of α-tocopherol and can be corrected by administration of supplemental vitamin K. Similarly, it appears that in vitamin K-deficient humans these effects could occur if sufficiently high doses were obtained to overwhelm the protective effects of vitamin K. However, there is no direct evidence regarding the doses that might put normal apparently healthy humans at risk for such effects; these data show that individuals who are deficient in vitamin K are at increased risk of coagulation defects.
Data Selection. In the absence of human data pertaining to dose-response relationships, the data sets used to identify a no-observed-adverse-effect level (NOAEL) for α-tocopherol include studies showing hemorrhagic toxicity in rats (Abdo et al., 1986; Takahashi et al., 1990; Wheldon et al., 1983).
Identification of NOAEL and Lowest-Observed-Adverse-Effect Level (LOAEL). A LOAEL of 500 mg/kg body weight/day can be identified based on a critical evaluation by Wheldon et al. (1983). They fed all rac-α-tocopheryl acetate to Charles River CD strain rats at levels of 500, 1,000, or 2,000 mg/kg body weight/day for 104 weeks. Hemorrhages from the gut, the urinary tract, the orbit and meninges, and the claws were observed in male rats only by week 15 in the highest-dose group, by week 16 in the intermediate-dose group, and by week 18 in the low-dose group. Prothrombin times were prolonged in males of all three treatment groups by week 4. However, additional vitamin K supplementation of the diets was initiated at week 24 and prothrombin times returned to normal by week 26. Although this was a chronic, 104-week study, the correction of vitamin K levels in the diet at week 24 means that the combined vitamin E-vitamin K effect was evaluated only on a subchronic basis.
Takahashi et al. (1990) supplemented the diets of male Sprague-Dawley rats with 600 or 1,000 mg/kg body weight/day RRR-α-tocopheryl acetate for 7 days. Despite the short duration of the feeding trial, a dose-dependent increase in prothrombin time and partial thromboplastin time was noted in rats receiving 600 and 1,000 mg/kg/day. The number of animals with hemorrhages was similar in both dose groups. No supplementation of vitamin K was attempted in this experiment. This study would yield a LOAEL of