Vitamin E and Mortality in Male Smokers of the ATBC Study: Implications for Nutritional Recommendations

Abstract

The Dietary Reference Intakes (DRI)-monograph (USA/Canada) states that the estimated average requirement (EAR) of vitamin E for men and women of any age is 12 mg/day. The EAR value is based on in vitro hemolysis in young males; a surrogate endpoint without any direct validity. The EAR is then extrapolated to females and older males. The validity of the EAR level is therefore questionable. Total mortality is an outcome of direct clinical relevance. Investigating the effect of long-term dietary vitamin E intake level on mortality in a randomized trial is, however, not feasible. Nevertheless, the effect of dietary vitamin E intake can be investigated indirectly from the effects of a fixed-level vitamin E supplement administered to participants on variable levels of dietary vitamin E intake. If vitamin E intake below the EAR is harmful, then vitamin E supplement should be beneficial to those people who have dietary vitamin E intake level below the EAR. The purpose of this study was to analyze the association between dietary vitamin E intake and the effect of 25 mg/day of vitamin E supplement on total mortality in Finnish male smokers aged 50-69 years in the Alpha-Tocopherol-Beta-Carotene (ATBC) Study. The effect of vitamin E supplement was estimated by Cox regression. Among participants who had dietary vitamin C intake of 90 mg/day and above, vitamin E supplement increased mortality by 19% (p = 0.006) in those aged 50-62 years, but decreased mortality by 41% (p = 0.0003) in those aged 66-69 years. No association between vitamin E supplement effect and dietary vitamin E intake was found in these two groups, nor in participants who had dietary vitamin C intake less than 90 mg/day. There is no evidence in any of the analyzed subgroups that there is a difference in the effect of the 25 mg/day vitamin E supplement on males on dietary vitamin E intakes below vs. above the EAR of 12 mg/day. This analysis of the ATBC Study found no support for the 'estimated average requirement' level of 12 mg/day of vitamin E for older males. Trial registration: ClinicalTrials.gov, identifier: NCT00342992.

Keywords: adverse effect; alpha-tocopherol; ascorbic acid; biomarkers; evidence-based medicine; nutrition policy; nutritional requirements; reference values.

Figure 1

The effect of vitamin E supplement by dietary vitamin E intake. (A) Participants aged 50–62 years at baseline. (B) Participants aged 66–69 years at baseline. The three vitamin E intake groups of Table 2 are plotted at the median vitamin E intake of the respective group. The points indicate the estimates and the vertical lines indicate their 95% CIs. The solid horizontal lines indicate the average effect of vitamin E supplement in the younger (A) and older (B) participants, and the vertical bars on the right-hand side indicate the 95% CI for the average effect. The dashed horizontal lines indicate the no effect level. The arrows indicate the levels of the EAR and the RDA for vitamin E (1).

Figure 2

Relationship between baseline age and the effect of vitamin E supplement in males who had dietary vitamin C intake ≥90 mg/day. The three age groups of Table 2 are plotted at the median age of each group. The points indicate the estimates and the vertical lines indicate their 95% CI. The uniformity of the vitamin E supplement effect over age was tested by adding a dummy variable for vitamin E effect for the 63–65 year and the ≥66 year age groups, allowing each of the three age groups of Table 2 their own vitamin E effect. The regression model was improved by ${\chi }_{(2\text{df})}^2$ = 21.5, p = 0.00002, compared with the model with a uniform vitamin E supplement effect over all ages. The dotted diagonal line shows the Cox regression model with age as a continuous variable. Allowing the slope, the regression model was improved by ${\chi }_{(1\text{df})}^2$ = 12.8, p = 0.0004, compared with the model with a uniform vitamin E supplement effect over all ages.