The questionable association of vitamin E supplementation and mortality--inconsistent results of different meta-analytic approaches.

摘要

Medical research suggests benefits of vitamin E supplementation in treatment or prevention of cardiovascular disease, inflammatory joint diseases and cancer. Regardless of these benefits in a recently published meta analysis the authors drew the conclusion that high dose supplementation may cause a slight increase in mortality of the treated patients. The purpose of the present paper is to re-analyse the association of vitamin E supplementation and mortality. By means of augmented data sources as well as additional methodological approaches the results of the above mentioned meta analysis is to be either confirmed or called into question. In the above mentioned meta analysis 19 clinical trials comprising a total of 135967 participants were included. The dosages of vitamin E supplementation ranged from 16.5 to 2000 IU/d. In the present paper this data source was augmented and 10 additional trials were included (2495 additional participants receiving vitamin E doses from 136 to 5000 IU/d). Moreover in 2 of the originally included trials updated results of mortality at longer periods of follow-up were available. The present paper yields contradictory results regarding the association of vitamin E supplementation and mortality. Hierarchical logistic regression analyses confirm the former results showing an increased mortality of patients receiving high dose vitamin E. Furthermore a traditional methodological approach of meta-regression was applied to the same data source. Contrary to the former result it showed that the increased mortality odds ratio in certain trials is not due to the higher dose of vitamin E supplementation. Rather it can be explained by a higher proportion of male patients that were included in these trials compared to other trials. The causal relationship of vitamin E supplementation and increased mortality is questionable. Different methodological approaches of meta analysis yield contradictory results. Thus none of these results can be regarded to supply evidence in a statistical sense. In particular high dose vitamin E supplementation can not be regarded proved to increase mortality.