Memorable Antioxidants

Sources: Perrig WJ, et al. The relationship between antioxidants and memory performance in the old and very old. JAGS 1997;45:718-724; Sano M, et al. A controlled trial of selegiline, alpha-tocopherol, or both as a treatment for Alzheimer’s disease. N Engl J Med 1997;336:1216-1222.

Oxidative stress mediated by free radicals is thought to contribute to a variety of age-related processes, precipitating recent studies in the use of antioxidant therapy for age-related neurologic disorders. Thus far, however, a paucity of hard data has described their effects in the treatment of neurodegenerative disease and cognitive dysfunction with advancing age. Two recent studies add fuel to this fire.

Swiss investigators Perrig et al studied plasma levels of the antioxidants a-tocopherol (Vit E), ascorbic acid (Vit C), and b-carotene (Vit A) determined at two time points (1971 and 1993) in 442 men and women ranging in age from 65 to 94 years. The correlation between vitamin levels and memory performance was measured by a short, computer-based, cognitive testing battery, controlling for potential confounding variables such as age, education, and gender. Performance in spontaneous recall, recognition memory, and vocabulary correlated significantly with levels of Vit C and Vit A, whereas functions such as priming and working memory did not. Only 6% of all subjects ingested vitamin supplements, indicating that most obtained vitamins from a normal diet. The investigators suggested that a lack of correlation between Vit E and cognitive performance in their study might relate to the already high a-tocopherol levels in their population.

Sano and coworkers studied the association of Vit E and selegiline on the time it took patients with mild-to-moderate Alzheimer’s disease (AD) to reach end points of death, institutionalization, increased functional dependency, or severe dementia. A double-blind, randomized, multicenter, prospective trial compared either 1000 IU Vit E bid or 5 mg selegiline bid alone to both in combination and to placebo. Despite randomization, the control group had a higher average baseline score on the Folstein Mini-Mental State Examination (MMSE) than any of the treatment groups. Baseline MMSE highly predicted outcome so the investigators needed to make statistical adjustments to compensate for this difference. In unadjusted analyses, no difference was detected between the placebo and any of the treatment groups. However, when the baseline differences in MMSE were incorporated in the Cox proportional hazard model, Vit E or selegiline, either alone or in combination, delayed the median time to reach the end points by between 145 and 230 days. The authors conclude, "In patients with mild-to-moderate Alzheimer’s disease, treatment with selegiline or a-tocopherol slows the progression of disease."


The Swiss study, in which most persons did not take vitamin supplements, shows significant correlations of random plasma levels Vit C and Vit A (but not Vit E) with some but not all measures of tested memory function. One cannot assume, however, a causative relationship between plasma vitamin levels and better cognitive function. For example, dietary habits could correlate with these cognitive measures, leading better test performers to have higher vitamin levels independent of any antioxidant effect. The findings are nevertheless encouraging for those who acclaim a potential role for antioxidant therapy in age-related memory problems. Clinically, measurement of vitamin levels is not cost-effective except when one suspects a disorder that would interfere with adequate absorption of orally administered supplements.

The study by Sano and colleagues took a relatively new approach in Alzheimer pharmaceutical trials of using survival analysis and time to reach degenerative end points to assess treatment efficacy. This approach provides more relevant information than methods that focus on statistically significant changes that may or may not have clinical importance. The study has been criticized because its conclusions depend on statistical adjustments, as David Drachman and Paul Leber indicate in an editorial (N Engl J Med 1997;336:1245-1247). The authors’ conclusion that these agents retard disease progression may be misleading, since the investigators studied end points that correlate with progression but are not in themselves disease markers. Although they imply that antioxidants are the first identified treatment to reduce the pathology of AD, there is insufficient evidence to draw this conclusion.

Given the devastating course of Alzheimer’s disease, the low risk and relatively low cost of Vit E make it a reasonable prescription to offer despite the less than convincing evidence that it affects the underlying disease. —nrr