close
Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Meta-Analysis
. 2013 Sep 6;8(9):e74558.
doi: 10.1371/journal.pone.0074558. eCollection 2013.

Meta-regression analyses, meta-analyses, and trial sequential analyses of the effects of supplementation with beta-carotene, vitamin A, and vitamin E singly or in different combinations on all-cause mortality: do we have evidence for lack of harm?

Goran Bjelakovic  1 , Dimitrinka Nikolova, Christian Gluud
Affiliations
Meta-Analysis

Meta-regression analyses, meta-analyses, and trial sequential analyses of the effects of supplementation with beta-carotene, vitamin A, and vitamin E singly or in different combinations on all-cause mortality: do we have evidence for lack of harm?

Goran Bjelakovic et al. PLoS One. .

Abstract

Background and aims: Evidence shows that antioxidant supplements may increase mortality. Our aims were to assess whether different doses of beta-carotene, vitamin A, and vitamin E affect mortality in primary and secondary prevention randomized clinical trials with low risk of bias.

Methods: The present study is based on our 2012 Cochrane systematic review analyzing beneficial and harmful effects of antioxidant supplements in adults. Using random-effects meta-analyses, meta-regression analyses, and trial sequential analyses, we examined the association between beta-carotene, vitamin A, and vitamin E, and mortality according to their daily doses and doses below and above the recommended daily allowances (RDA).

Results: We included 53 randomized trials with low risk of bias (241,883 participants, aged 18 to 103 years, 44.6% women) assessing beta-carotene, vitamin A, and vitamin E. Meta-regression analysis showed that the dose of vitamin A was significantly positively associated with all-cause mortality. Beta-carotene in a dose above 9.6 mg significantly increased mortality (relative risk (RR) 1.06, 95% confidence interval (CI) 1.02 to 1.09, I(2) = 13%). Vitamin A in a dose above the RDA (> 800 µg) did not significantly influence mortality (RR 1.08, 95% CI 0.98 to 1.19, I(2) = 53%). Vitamin E in a dose above the RDA (> 15 mg) significantly increased mortality (RR 1.03, 95% CI 1.00 to 1.05, I(2) = 0%). Doses below the RDAs did not affect mortality, but data were sparse.

Conclusions: Beta-carotene and vitamin E in doses higher than the RDA seem to significantly increase mortality, whereas we lack information on vitamin A. Dose of vitamin A was significantly associated with increased mortality in meta-regression. We lack information on doses below the RDA.

Background: All essential compounds to stay healthy cannot be synthesized in our body. Therefore, these compounds must be taken through our diet or obtained in other ways [1]. Oxidative stress has been suggested to cause a variety of diseases [2]. Therefore, it is speculated that antioxidant supplements could have a potential role in preventing diseases and death. Despite the fact that a normal diet in high-income countries may provide sufficient amounts of antioxidants [3,4], more than one third of adults regularly take antioxidant supplements [5,6].

PubMed Disclaimer

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1
Figure 1. PRISMA flow diagram of identification of randomized clinical trials for inclusion.
Figure 2
Figure 2. Trial sequential analysis of 6 trials assessing beta-carotene in a dose at or below 9.6 mg daily versus placebo.
The diversity-adjusted required information size (DARIS = 267,631 participants) was based on a proportion of deaths of 10% in the placebo group; a relative risk reduction of 5% in the beta-carotene group; an alpha of 5%; a beta of 20%; and a diversity of 59%. The blue line represents the cumulative Z-score of the meta-analysis. The green lines represent the conventional statistical boundaries. The red lines represent the truncated trial sequential monitoring boundaries.
Figure 3
Figure 3. Trial sequential analysis of 20 trials assessing beta-carotene in a dose above 9.6 mg daily versus placebo.
The diversity-adjusted required information size (DARIS = 190,906 participants) was based on a proportion of deaths of 10% in the placebo group; a relative risk reduction of 5% in the beta-carotene group; an alpha of 5%; a beta of 20%; and a diversity of 42%. The blue line represents the cumulative Z-score of the meta-analysis. The green lines represent the conventional statistical boundaries. The red inward sloping lines represent the trial sequential monitoring boundaries. The red outward sloping lines represents the area of futility.
Figure 4
Figure 4. Trial sequential analysis of 8 trials assessing vitamin A in a dose at or below RDA (≤ 800 µg) daily versus placebo.
The diversity-adjusted required information size (DARIS = 48,417 participants) was based on a proportion of deaths of 10% in the placebo group; a relative risk reduction of 15% in the vitamin A group; an alpha of 5%; a beta of 20%; and a diversity of 76%. The blue line represents the cumulative Z-score of the meta-analysis. The green lines represent the conventional statistical boundaries. The red lines represent the truncated trial sequential monitoring boundaries. Had we used a relative risk reduction of 5% as planned, the DARIS would have been 456,748 participants and the program could not have drawn the trial sequential analysis due to the fact that the number of randomized patients out of the DARIS is too small. This is why, we post hoc decided to construct the trial sequential analysis with a larger relative risk reduction.
Figure 5
Figure 5. Trial sequential analysis of 4 trials assessing vitamin A in a dose above the RDA (> 800 µg) daily versus placebo.
The diversity-adjusted required information size (DARIS = 415,996 participants) was based on a proportion of deaths of 10% in the placebo group; a relative risk reduction of 5% in the vitamin A group; an alpha of 5%; a beta of 20%; and a diversity of 73%. The blue line represents the cumulative Z-score of the meta-analysis. The green lines represent the conventional statistical boundaries. The red lines represent the truncated trial sequential monitoring boundaries.
Figure 6
Figure 6. Trial sequential analysis of 2 trials assessing vitamin E in a dose at or below RDA (≤ 15 mg) daily versus placebo.
The diversity-adjusted required information size (DARIS = 12,563 participants) was based on a proportion of deaths of 10% in the placebo group; a relative risk reduction of 15% in the vitamin E group; an alpha of 5%; a beta of 20%; and a diversity of 7%. The blue line represents the cumulative Z-score of the meta-analysis. The green lines represent the conventional statistical boundaries. The red lines represent the truncated trial sequential monitoring boundaries. Had we used a relative risk reduction of 5% as planned, the DARIS would have been 119,364 participants and the program could not have drawn the trial sequential analysis due to the fact that the number of randomized patients out of the DARIS is too small. This is why, we post hoc decided to construct the trial sequential analysis with a larger relative risk reduction.
Figure 7
Figure 7. Trial sequential analysis of 44 trials assessing vitamin E in a dose above the RDA (> 15 mg) daily versus placebo.
The diversity-adjusted required information size (DARIS = 110,505 participants) was based on a proportion of deaths of 10% in the placebo group; a relative risk reduction of 5% in the vitamin E group; an alpha of 5%; a beta of 20%; and a diversity of 0%. The blue line represents the cumulative Z-score of the meta-analysis. The green lines represent the conventional statistical boundaries. The red inward sloping lines represent the trial sequential monitoring boundaries. The red outward sloping lines represent the area of futility.
See this image and copyright information in PMC

References

    1. Rosenfeld L (1997) Vitamin e - vitamin. The early years of discovery. Clin Chem 43: 680-685. PubMed: 9105273. - PubMed
    1. Sies H (1997) Oxidative stress: oxidants and antioxidants. Exp Physiol 82: 291-295. PubMed: 9129943. - PubMed
    1. Dwyer JT, Costello RB (2013) Assessment of dietary supplement use. In: Coulston AM, Boushey CJ, Ferruzzi MG. Nutrition in the prevention and treatment of disease. Amsterdam, Boston, Heidelberg, London, New York. Paris, San Diego, San Francisco, Singapore, Sydney, Tokyo: Oxford: Academic Press; . pp. 47-64
    1. Herbert V, Subak-Sharpe G, Hammock DA (1990) The Mount Sinai School of Medicine Complete Book of Nutrition. New York: St Martin’s Press.
    1. Bailey RL, Gahche JJ, Miller PE, Thomas PR, Dwyer JT (2013) Why US adults use dietary supplements. JAMA Intern Med 173: 355-361. doi:10.1001/jamainternmed.2013.2299. PubMed: 23381623. - DOI - PubMed

Publication types