Why Antioxidants in Large Randomized Trials Haven’t Shown Clear Benefits

The Durk Pearson & Sandy Shaw®
Life Extension NewsTM
Volume 7 No. 3 • June 2004


Why Antioxidants in Large Randomized Trials Haven’t Shown Clear Benefits, While Observational Data Have

There has been much discussion concerning why large randomized intervention trials of vitamin E, vitamin C, or the combination have generally shown little or no benefits on cardiovascular disease, cancer, or all-cause mortality, in sharp contrast to the clear and substantial beneficial effects shown in observational studies using cells, animals, or humans, and in many human epidemiological studies. With clear mechanisms that indicate protective effects, why do you not see the expected results in the large intervention trials? Many hypotheses have been proposed, including that doses are too low, follow-up times are too short, or that antioxidants work as systems with other antioxidants, so you need more than just one or two. The authors of a new study1 propose that there are far more confounding factors than have been accounted for.

They analyzed the association of a wide range of indicators of socioeconomic factors, childhood environmental circumstances, and behavioral risk factors with plasma vitamin C and E concentrations from data collected in the British Women’s Heart and Health Study. The study included 4286 women aged 60–79 years randomly selected from 23 British towns. They found that people from poorer socioeconomic status at any time had lower vitamin concentrations. The odds of being in the highest quartile of the plasma vitamin C distribution decreased by 22% for each additional marker of adverse-life-course socioeconomic position. Results for vitamin E were similar. The association with each vitamin was independent of that of the other vitamin. Women who smoked and those who were obese had lower vitamin C and E concentrations. Those who engaged in at least one hour of exercise a week were said to report eating a low-fat diet or high-fiber diet, and those who consumed alcohol daily had higher vitamin concentrations. Women who had longer legs relative to their trunk length had higher concentrations of vitamin C and E, independent of socioeconomic and behavioral risk factors.

The researchers suggest, therefore, that “the conflicting observational and trial findings are probably the result of residual confounding caused by inadequate adjustment for the complexity of social and environmental exposures acting across the life course.” They cite a prospective cohort study that found that cardiovascular death was affected, in a cumulative fashion, by socioeconomic position and behavioral factors throughout the life course. The risk of death from cardiovascular disease was four times greater in those with the most adverse status for all socioeconomic and behavioral factors, compared with those having the most advantageous of those factors.

On the other hand, we and others have noticed that observational studies of hormone replacement therapy that have reported protective effects have largely used estradiol, the major human estrogen, while the large randomized trials that have reported little or no protection or even worsening of risk for cardiovascular disease or Alzheimer’s have largely used conjugated horse (not human) estrogens and synthetic progestins (not the same as natural human progesterone).2 Hence, these trials have not tested the natural human estrogens that provide protection to premenopausal women. Whether the available formulations of natural human estrogens provide the same protection as that enjoyed by premenopausal women has yet to be determined experimentally.

  1. Lawlor et al. Those confounded vitamins: what can we learn from the differences between observational versus randomised trial evidence? Lancet 363:1724-7 (2004).
  2. See, e.g., Shumaker et al. Conjugated equine estrogens and incidence of probable dementia and mild cognitive impairment in postmenopausal women. JAMA 291(24):2947-58 (2004). Espeland et al. Conjugated equine estrogens and global cognitive function in postmenopausal women. JAMA 291(24):2959-68 (2004).

FREE Subscription

  • You're just getting started! We have published thousands of scientific health articles. Stay updated and maintain your health.

    It's free to your e-mail inbox and you can unsubscribe at any time.
    Loading Indicator