Non-comparability of the vitamin C trials

by Harri Hemilä

This text is based on pp 34-35 of Hemilä (2006)
These documents have up to date links to documents that are available via the net.

Harri Hemilä
Department of Public Health
University of Helsinki, Helsinki, Finland
harri.hemila@helsinki.fi

Home: http://www.mv.helsinki.fi/home/hemila

These files are at: http://www.mv.helsinki.fi/home/hemila/metaanalysis

Version May 29, 2012

One particular problem in the meta-analysis of vitamin C trials arises from the fundamental difference between vitamin C and ordinary drugs such as antibiotics. It is possible to select a control group which has no intake of an ordinary drug, rendering the interpretation of results relatively simple. It is impossible, however, to select control subjects who have no vitamin C intake, and no vitamin C in their system. Accordingly, all vitamin C trials compare two different intake levels, the lower level being obtained from the diet, and usually not estimated at all, which hampers the comparison of different trials and the generalization of their results.

Since the intake of vitamin C in the diet, and the supplementary dosage have both varied substantially between trials (Table 12), the comparison of vitamin C trials is sometimes quite problematic. For example, while there is some 50-fold difference in the vitamin C intake in the diet by the control group between the Glazebrook and Thomson (1942) and Peters et al. (1993) trials, both are ‘control groups’ of vitamin C trials.

The vitamin dosage in the ‘vitamin C group’ has also varied dramatically (Table 12). At the extreme, Karlowski et al. (1975) administered up to 6 g/day of vitamin C to their subjects, whereas Cowan et al. (1942) administered 25 mg/day as their lowest supplementary dose. There is thus up to a 240-fold difference between the lowest and highest vitamin C supplementary dose used in these trials, yet in many cases such studies are simply labeled ‘vitamin C trial’ without paying any attention to the dosage.

Furthermore, the placebo group of Peters et al. trial (1993) received some 500 mg/day of vitamin C in their diet, whereas the vitamin C group in Baird et al. trial (1979) received only 130 mg/day (diet and supplement together). Thus, vitamin C intake in the placebo group of the former trial was 4 times higher than the vitamin C intake in the vitamin C group of the latter trial.

A further problem in vitamin C doses is the addition of vitamin C to the placebo group in some studies with the apparent rationalization of excluding the possibility that any observed effects of large doses might be explained by incidentally treating marginal deficiency. For example, Carr et al. (1981) administered 70 mg/day and Miller et al. (1977) 50 mg/day to their placebo subjects. When comparing these ‘placebo-group’ vitamin C doses to the 80 mg/day of vitamin C administered to the ‘vitamin C group’ of the Baird et al. trial (1979) or 25-50 mg/day in the Cowan et al. trial (1942), it is clear that the comparison of ‘vitamin C trials’ can be very complex, and extrapolation of the findings of a single trial is limited irrespective of the methodological quality of the particular trial.

The great variation in the vitamin C doses in diet and in supplements probably explains part of the variation in the results of the trials. An interesting example of the possible modification of supplement effect through diet was reported by Anderson et al. (1972, 1975), who found that vitamin C was more beneficial for participants who consumed low amounts of orange juice, which is an important source of vitamin C (Table 13).

The intake of vitamin C in the diet, and the supplementary dosages are rarely considered in discussions of vitamin C trials. For example, they are not considered in the meta-analyses by Chalmers (1975; pp 36-8 in Hemilä 2006) and Kleijnen et al. (1989; pp 38-41 in Hemilä 2006). Dykes and Meier (1975; pp 42-5 in Hemilä 2006) listed the supplementary vitamin C doses used in the trials, but did not consider the variation in dietary intake.

Table 12. Variation in vitamin C intake in diet and supplementary doses in common cold trials

Trial Country Participants	Vitamin C in diet (mg/day)	Vitamin C supplement mg/day)
Glazebrook-1942 UK Schoolboys	10-15	50-300
Peters-1993 South Africa Marathon runners	500	600
Baird-1979 UK Students	50	80
Cowan-1942 USA Schoolchildren	?	25-50
Karlowski-1975 USA Adults	?	3,000-6,000

Table 13. Subgroup differences in the Anderson et al. trials (1972,1975)

	Effect on "total days indors"
Subgroups	(1972) Regular supplementation	(1975) Therapeutic supplementation
Daily juice
0-3 oz	-48%	-33%
4+ oz	-22%	-22%
Contact with young children
Yes	-46%	-40%
No	-17%	-13%
Usual colds
2+	-43%
0-1	-13%

Anderson et al. (1972) administered 1 g/day over the whole study period and 3 g/day for 3 days during cold episodes. anderson et al. (1975) administered 1.5 g/day on the first day of the cold and 1 g/day for the following 4 days. In the latter trial participants were also regularly administered 0.5 g once a week (i.e. 0.07 g/day) on a rationalization that remains obscure.

References

Anderson TW, Beaton GH, Corey PN, Spero L (1975) Winter illness and vitamin C: the effect of relatively low doses. Can Med Assoc J 112:823-6

Anderson TW, Reid DBW, Beaton GH (1972) Vitamin C and the common cold: a double-blind trial. Can Med Assoc J 107:503-8

Baird IM, Hughes RE, Wilson HK, et al. (1979) The effects of ascorbic acid and flavonoids on the occurrence of symptoms normally associated with the common cold. Am J Clin Nutr 32:1686-90

Carr AB, Einstein R, Lai LYC, et al. (1981a) Vitamin C and the common cold: a second MZ cotwin control study. Acta Genet Med Gemellol 30:249-55 * see also: Carr et al. (1981b)

Carr AB, Einstein R, Lai LYC, et al. (1981b) Vitamin C and the common cold: using identical twins as controls. Med J Aust 2:411-2

Chalmers TC (1975) Effects of ascorbic acid on the common cold: an evaluation of the evidence. Am J Med 58:532-6 *** SEE PROBLEMS OF CHALMERS' REVIEW

Cowan DW, Diehl HS, Baker AB (1942) Vitamins for the prevention of colds. JAMA 120:1268-71 CH BM

Dykes MHM, Meier P (1975) Ascorbic acid and the common cold: evaluation of its efficacy and toxicity. JAMA 231:1073-9 * comments in: JAMA (1976);235:1108 and Pauling (1976b, 1976c) * SEE PROBLEMS OF THE DYKES-MEIER REVIEW

Glazebrook AJ, Thomson S (1942) The administration of vitamin C in a large institution and its effect on general health and resistance to infection. J Hyg (Lond) 42:1-19

Karlowski TR, Chalmers TC, Frenkel LD, Kapikian AZ, Lewis TL, Lynch JM (1975) Ascorbic acid for the common cold: a prophylactic and therapeutic trial. JAMA 231:1038-42 CH *** SEE PROBLEMS OF THE KARLOWSKI TRIAL

Kleijnen J, Riet G, Knipschild PG (1989) Vitamine C en verkoudheid; overzicht van een megadosis literatuur [in Dutch]. Ned Tijdschr Geneeskd 133;1532-5
English translation: Vitamin C and the common cold; a review of the megadose literature. In: Food Supplements and Their Efficacy. pp 21-8. Thesis for University of Limburg (1991); Netherlands; ISBN 90 900 4581 3

Miller JZ, Nance WE, Norton JA, et al. (1977) Therapeutic effect of vitamin C: a co-twin control study. JAMA 237:248-51

Peters EM, Goetzsche JM, Grobbelaar B, Noakes TD (1993) Vitamin C supplementation reduces the incidence of postrace symptoms of upper-respiratory-tract infection in ultramarathon runners. Am J Clin Nutr 57:170-4

Copyright: © 2006-2009 Harri Hemilä. This text is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Vitamin C and infections in animals by Harri Hemilä is licensed under a Creative Commons Attribution 1.0 Finland License.
Based on a work at www.mv.helsinki.fi/home/hemila/metaanalysis