On the Prevention of the Common Cold: No Help from Vitamin C

In 1970, the Nobel-winning scientist Linus Pauling proposed that mega-doses of vitamin C were efficacious in preventing colds and reducing their symptoms. ¹ Subsequent clinical trials have indicated that vitamin C supplementation in large doses has little, if any, effect on an individual’s likelihood of developing a common cold. Even though some believe there is evidence in favor of a small reduction in the duration of symptoms, ^2–4 no major medical authority in the United States currently recommends vitamin C supplementation for the prevention or therapy of the common cold.

Until now the unanswered question has been whether dietary vitamin C, rather than supplementation, could affect the incidence or symptoms of the common cold. Two studies published in this issue of Epidemiology ^5–6 effectively dampen even that hope.

Both studies relied upon self-diagnosis of a cold episode. Since it is one’s own assessment of malaise that largely determines the extent to which one withdraws from economic or normal daily activities, this approach is arguably more relevant than that of seropositivity. From their validation sub-study, Takkouche and colleagues determined self-diagnosis had a sensitivity of 94% and a specificity of 84%. ⁶

In addition to their common finding of dietary vitamin C having no effect on the risk of succumbing to the common cold, each study has other observations worthy of note. The study from Spain involving the faculty and staff of five universities found neither vitamin C nor zinc (also measured in the diet) afforded any protection for subjects under high levels of psychological stress. ⁶ The Finnish study was restricted to males aged 50–69 years who smoked. ⁵ The lack of a protective effect against the common cold was also demonstrated for dietary levels of vitamin E and β-carotene. In the diet-supplementation arm of the Finnish study, the authors found the relative risk of a cold among men aged 65 or more years who took vitamin E (RR = 0.95) had an upper 95% confidence interval of 1.0. The statistical significance of this observation is questionable, however, if it was simply achieved by searching the data and performing multiple tests.

The two studies in this issue are unlikely to face the criticism received by the earliest studies, which found evidence for effects of vitamin C supplementation on the common cold. In 1975 Dykes and Meier ⁷ conducted a particularly incisive review of studies supporting Pauling’s hypotheses. With the exception of Anderson et al’ s first study, ⁸ these studies left serious doubts as to whether their results could be accepted as unbiased. In their second study, Anderson et al⁹ were unable to replicate the results of their first study.

Table 1 summarizes the results of double-blind randomized studies published since the review by Dykes and Meier. ⁷ I excluded letters or abstracts with insufficient details of study methodology, and also those studies dealing only with the issue of therapeutic efficacy. Whether the subjects are schoolchildren, ^10,11 identical twins, ^12,13 or marine recruits, ¹⁴ the results suggest that vitamin C supplementation has as little effect upon the average number of sick days as it does upon the incidence. However, the severity scores (peculiar to each study) are consistently less for ascorbic acid recipients. This may be due to an antihistaminic property of vitamin C. ^10,17

Table 1: Double-Blind Randomized Studies Published Since 1975 on the Relation of Vitamin C Supplementation with Average Number of Colds per Subject, Average Number of Days Sick, and Average Severity

We must be almost at the end of the road with regard to Pauling’s hypotheses. But should we publicly emphasize that there is little effect of vitamin C on the common cold and remove all possibility of placebo benefit? In a clinical trial by Karlowski et al, ¹⁸ the average duration of colds in subjects who guessed correctly that they had been given a placebo throughout the study was 8.6 days. In subjects who guessed correctly that they had been given both prophylactic and therapeutic ascorbic acid, cold duration was only 4.8 days. Similar results (6.9 vs 3.7 days) were obtained by Carr et al, ¹³ except that in this study the longer duration of symptoms was related to a perceived low dose of ascorbic acid and the shorter duration to a perceived high dose. Self-perception of high dose also resulted in a 39% reduction in the incidence.

According to a recent survey, 67% of the general public believes that taking vitamin C reduces cold symptoms. ¹⁹ The results of randomized double-blind studies suggest that it is not the symptoms themselves that are significantly affected but the subject’s perception of them. It might be that, were it not for the public’s perceived benefits from vitamin C, the figure of 250 million restricted activity days per year due to colds ²⁰ would be considerably higher. On the other hand, subjects who return to their normal daily activities too soon increase the probability of infecting others. Until it is clearly demonstrated that the risk of spreading infection is more important than the placebo benefits of vitamin C, perhaps a daily dose of up to 250 mg should not be discouraged, provided it is not substituting for an adequate diet. At the same time, however, the public should be informed that for some people a daily dose of 1 g or more of vitamin C over a long period might have adverse health consequences. ²¹