Viewpoint: does mega-C do more good than harm, or more harm than good?

Viewpoint: does mega-C do more good than harm, or more harm than good?

Victor Herbert

The recently reported study by.. Enstrom, Kanim and Klein’ demonstrated that a group of individuals with a healthy life-style, manifested by four key healthy life-style markers (“confounding variables” in the jargon of science) that they weigh less (2.2 kg less body fat in the men, 4.4 in the women), smoke less, exercise more, eat more fruits and vegetables (including oranges, grapefruits, tomatoes and their juices), and incidentally also take regular vitamin supplements (averaging a “best guess” megadose of 800 mg of vitamin C/day), live longer than a group who weigh more, smoke more, exercise less, eat less fruits and vegetables and incidentally take no regular supplements.

However, Enstrom et al. ignored two of these four (weight loss and eating fruits and vegetables) and lumped the other two confounding variables, smoking and exercise, with eight other variables including total fat and calories consumed. All of these eight variables were identical in the men who took supplements and those who did not and do not appear to be confounding in their study. The differences in the two were obscured by the non-differences in the other eight. Therefore, Enstrom et al. concluded that the supplements were responsible for the much greater longevity in the men in the supplemented compared to the nonsupplemented groups. Because the women taking supplements ate more fat than the women who did not take supplements, their increase in longevity was less.

Before one accepts their conclusion that the supplements were responsible for greater longevity, one would like to see the results if, instead of using the supplements as the index marker, they reevaluated their data, using the four key healthy life-style markers as principal markers in the same supplement and no supplement groups. One would anticipate that such a reevaluation would show an even greater increment in longevity due to those four markers than shown by using the incidental-to-a-healthy-life-style marker of consumption of vitamin C supplements. Vitamin supplement users are more likely to be from a higher socioeconomic status than non-users. People “into” a healthy life-style often also take vitamin C supplements, because of the relentless hype that vitamin pills are part of a healthy life-style, making vitamin C supplements a surrogate pseudo-marker for a number of unmeasured variables having to do with a healthy life-style and/or better access to medical care and appropriate nutrition.

One suspects that, if Enstrom et al. reevaluate their data using the above four healthy life-style markers to separate those into two groups (>500 mg/day and 50 to 500 mg/day). The group using mega-C (>500 mg/day) supplements will fare less well than the group using less, because of the harms from megadoses of vitamin C.

Promoters of vitamin C supplements claim that megadoses of vitamin C are harmless. Their incessant repetition of this fiction has so fixed it in concrete in the minds of Americans as “rock logic” [2] that even first-rate epidemiologists like Enstrom believe it. Circulated nationwide was his statement to the media, the day his article in Epidemiology [1] was published, that megadoses of vitamin C were harmless.

Nothing could be further from the truth. Vitamin C is a double-edged sword, necessary for health in small amounts and harmful in large amounts. High-dose vitamin C supplements, deceptively represented as “high potency” to convey an aura of increased value, have produced great harm, ranging from serious illness to death.

The representation of vitamin C and [beta]-carotene as antioxidants is both truth and misperception, because both are in fact redox agents and pro-oxidant rather than antioxidant in appropriate circumstances. To quote Repka and Hebbel, [3] and as others [4,5] have also pointed out, “lipid peroxiclation studies show that at physiologic levels ascorbate acts primarily as an antioxidant; however, as pharmacologic levels are reached, its pro-oxidant effects predominate.”

In the presence of iron, vitamin C is one of the most potent prooxidants known.[3-13] It converts iron stores to catalytic iron, one of the most oxidant of substances. [6] About 10% of Caucasians and about 8% of African-Americans are born with a gene for increased iron absorption (heterozygous hemochromatosis), and about 1 in 250 have two genes for enhanced iron absorption (homozygous hemochromatosis). [14,15] Vitamin C supplements, which enhance both iron absorption and the release of iron from body deposits, can act as a second gene for iron overload in those born with 1 gene for enhanced iron absorption. [6] By producing iron overload in these people and releasing catalytic iron from their body stores, vitamin C supplements can maim and kill. In her formal statement [16] supporting the position [17,18] that the FDA-proposed lower U.S. RDIs protect consumers, Margit Krikker of the Hemochromatosis Research Foundation wrote: “Vitamin C, which accelerates iron absorption, has also been responsible for cardiac deaths in at least three athletes, unaware of their predisposition to iron-loading or of the hazards of daily megadoses for years.” [16]

Some pertinent statistics:

1. Twice as many American adult men (1 in 250) have iron overload disease as have iron deficiency (1 in 500), so vitamin C supplements, which enhance iron absorption, are twice as likely to harm them as help them. [14, 19].

2. Almost twice as many Americans (about 10%) have a gene for positive iron balance as are in negative iron balance (about 6%, mainly infants, early adolescents, women in the reproductive years, and pregnant women), so vitamin C supplements, by enhancing iron absorption, if taken nonselectively by all Americans, are likely to do more harm than good.[6]

3. In a 5-year study of more than 1900 Finnish men, published in Circulation in September 1992, Dr. Jukka T. Salonen and his colleagues[20] found that for each 1% increase in serum ferritin there was a more than 4% increase in risk of heart attack. Finnish men with serum ferritin above 200 had 2.3 times as many heart attacks as Finnish men with serum ferritin of 100. High low-density lipoprotein (LDL) cholesterol level per se was not a risk factor. It only became one when there was concurrent high ferritin, which, particularly in the presence of vitamin C, [6,21], releases catalytic iron which, in turn, converts the harmless LDL cholesterol to oxidized LDL cholesterol, which damages the walls of coronary arteries.

Olson and Hodges noted [22] (and provided pertinent literature references for) all of the following harms from excess vitamin C:

Occasional large intakes of vitamin C may cause stomach cramps, nausea and diarrhea in some fasting persons but have no long-term adverse effects.

When daily large doses are ingested routinely for months or years, however, a number of adverse effects may occur, including uricosuria, reduced bactericidal activity of leukocytes, secondary hyperoxalemia (producing metastatic oxalosis) in hemodialysis patients, enhanced mobilization of bone calcium, impaired blood coagulation time, lowered plasma [B.sub.12] levels, interruption of pregnancy, reduced insulin production and interference with anticoagulant therapy …. These and other possible effects of high doses have been thoughtfully reviewed by Barnes and by Hornig and Moser.[22] The extent to which the routine ingestion of very high doses of vitamin C impairs health in a serious and lasting way is unknown. The frequency of reported toxic manifestations is unquestionably low relative to the number of persons routinely ingesting large doses. The mortality rate among health-conscious elderly Californians who routinely ingested large doses of nutritional supplements, including vitamin C, is significantly lower than that of one non-smoking reference population but not lower than that of another health-conscious group. The mortality rate was independent of the reported amount of vitamin C ingested daily.[22]

The above section of their discussion[22] on toxicity of vitamin C is quoted in extenso because, when the Subcommittee on the RDA edited it, in their enthusiasm for the ignis fatuus of vitamin C against cancer,[23] they edited out much of the toxicity section as well as literature references to that toxicity.

Intravenous megadoses of vitamin C can kill within hours that one-eighth of men of Black, Oriental, and Sephardic Jewish and Mediterranean basin-origin born with genetically determined glucose-6-phosphate dehydrogenase deficiency, by instantly oxidatively hemolyzing their red blood cells.[24,25] They can also precipitate acute severe sickle cell crisis in all those with sickle cell disease by causing all their red cells to take the sickle form.[3,24,25]

Luckily, oral megadoses of vitamin C produce less than total hemolysis, less severe sickle cell crises and much less oxalosis than does intravenous mega-C. This is because, while 100% of any size intravenous mega-C dose is absorbed, a genetically calibrated progressively smaller amount of oral mega-C is absorbed the greater the oral dose. The higher the daily dosage, the smaller the percentage absorbed and the higher the amount excreted unchanged in the feces.[26] It is this unabsorbability that produces as a toxic effect the hyperosmotic diarrhea Dr. Linus Pauling proudly states that he gets from the 18 g of vitamin C he takes each day.

It is this same toxic effect that has caused AIDS patients who take mega-C to go into hypovolemic shock, by superimposing the hyperosmotic diarrhea from mega-C on the secretory diarrhea present in about 60% of AIDS patients.[27]

The false claim that mega-C is harmless is so pervasive that the Mount Sinai School of Medicine Complete Book of Nutrition [27] carries in its “Vitamins and Minerals” table a heading entitled “Signs of Overdose.” The listing for vitamin C includes, “Oxalate kidney stones, oxalate deposits in heart, other body tissues. Urinary tract irritation. Diarrhea. Blood destruction.” Other nutrition books seem to follow the practice of not dealing with harms from vitamin C by simply not citing the literature on harms.

A formal request has been sent to the FDA [17] that they require all sellers of vitamin C supplements (and iron supplements), and products with large amounts of additional vitamin C (or iron), to label them as follows:

NOTICE: Before using this product, your serum ferritin should be determined. Millions of Americans, particularly males, but also some females, have a serum ferritin greater than 120 micrograms per liter of blood. If your serum ferritin is greater than 120, this product may be harmful to your health, and you should not take it unless a licensed health professional tells you it is safe to do so.

The cut-off point of 120 is based on the data of Salonen et al. [20] that a serum ferritin level of 200 mg/L or more, more than doubled the relative risk of heart attacks. Additionally, since smokers already have an increased risk of heart attacks, it is likely the Food and Nutrition Board recommendation[28] that they increase their intake of vitamin C to 100 mg daily may further increase their risk of heart attacks, by enhancing iron absorption and increasing release of catalytic iron from ferritin. The latter triggers the oxidation of LDL cholesterol to a form that further narrows coronary arteries already narrowed in response to the nicotine in tobacco smoke.

Much of the commentary[29] on harm from iron supplements was, unfortunately, edited out of the loth RDAs,[28,30] including the reference to a paper reporting that as little as a 100-mg supplement of iron could produce liver cell wall lipid damage with enzyme leakage.

The Finnish study supports the theory, first advanced in 1981 by J. L. Sullivan, [31] that it is not estrogen which protects women in the child-bearing years against heart attacks, but the monthly blood loss, with its concurrent loss of iron.

A good way to bring down a high serum ferritin level in men (when it is due to high iron stores) is to have them donate a pint of blood every few weeks until their serum ferritin level is no longer high to concurrently eat a largely vegetarian diet with a moderate intake of vitamin C and with no supplements of vitamin C or iron. Alternatively, of course, or in addition, iron-chelating agents can be used when appropriate, and when the benefits exceed the harms. [21] Parenthetically, the American Association of Blood Banks says it is up to the individual blood bank director as to whether he/she would use a donation from a person with iron overload.

One suspects the greater fruit and vegetable intake of the supplement group in the Enstrom et al. study is significantly more important to their greater longevity than the supplements of vitamin C.

It is likely that the lesser frequency among vegetarians of heart attacks is in significant part due to the fact that plant iron on average is only about 3% absorbable, whereas animal iron averages 15% absorbable? Thus, vegetarians have substantially lower serum ferritin and iron stores than nonvegetarians.[29]

Each fruit and vegetable is a storehouse of literally hundreds of antioxidants, pro-oxidants, carcinogens, anticarcinogens, mutagens and antimutagens, all of which tend to balance each other out, but with the balance perhaps in favor of the anticarcinogens, in a sensible diet based on the principles of moderation, variety and balance. [27,32,35]

Talalay’s group[34,35] recently isolated and identified the mustard family chemical, sulforaphane, an isothiocyanate not destroyed by microwaving or steaming, from broccoli and other cruciferous vegetables such as Brussels sprouts, cauliflower, and kale, as well as noncruciferous carrots and green onions. It is a highly potent inducer of so-called phase II detoxication enzymes, which are involved in the inactivation of carcinogens; related isothiocyanates also have anticarcinogenic properties.

A variety of chemicals naturally present in foods and preservatives added to food cause cultured cells to generate a variety of enzymes, some of which (phase II detoxication enzymes) bond to and thereby inactivate and flush out carcinogens, others of which (phase I enzymes, such as cytochrome P-450)[36] transform otherwise innocuous chemicals into mutagens and carcinogens. Some chemicals induce both kinds of enzymes to a variable degree, depending on various factors.[27,32-36] An example of creation of a carcinogenesis promoter from a dietary per se nonpromoter is what our liver xanthine oxidase does to the acetaldehyde our livers produce from alcohol, by generating superoxide from it, a process accelerated by iron and vitamin C. [4,5,36] Our genetic blueprint determines our individual potential for maximal generation of each of our enzymes and whether we generate a normal or warped enzyme.[37]

It has been known since the late 1950’s that oral vitamin C dramatically releases iron from body stores, sharply raising the serum iron level.[38] For that reason, physicians started adding intravenous vitamin C to the chelating agents used to lower the high iron stores of patients in whom iron built up because of monthly transfusions for chronic anemias. A number of young people died suddenly while the vitamin C needle was still in the vein, because the vitamin C released into their bloodstreams lethal amounts of free iron, above the ability of the chelating agents to bind and thereby inactivate it. This free iron circulated to their hearts, producing fatal cardiac arrhythmias.[39] Due to this potential lethality, intravenous and oral vitamin C can only be used very cautiously to help lower body iron stores in young people with thalassemia and sickle cell disease.

At the 1992 annual meeting of the American Society for Clinical Nutrition, Elaine Feldman’s group [40] reported that in normal individuals supplements of 1 g daily of vitamin C for 1 month decreased systolic and diastolic blood pressure. In the discussion period, they were responsive to our observation that nitrous oxide, the “endothelial relaxing factor’ our own cells generate, [41] has the same effect, as well as to our suggestion that large doses of vitamin C acted as an oxidant to increase the few seconds of lifespan of nitrous oxide after its manufacture in endothelial cells, allowing more nitrous oxide to attach in the presence of excess vitamin C to its “receptor” iron bound to enzymes, [41] and our suggested methodology for checking this possibility.

At the April 1992 FASEB Symposium on Nutritional Epidemiology of Chronic Disease, a presentation titled “Results of a Metabolic Study of Response to Vitamin C, Population Implications,” based on blood levels, showed once again that if you eat more vitamin C, you get higher vitamin C blood levels, and if you eat less, you get lower blood levels. It promoted vitamin C supplements based on a number of epidemiologic studies. In the discussion, it was pointed out to the author that those epidemiologic studies showed the value of fruits and vegetables, not of vitamin C. A subsequent editorial [42] by the same author produced nationwide head-lines.

In her editorial, [42] accompanying the Enstrom et al. paper, [1] Gladys Block ignored the life-style factors. Her statement, “Their data indicate that those who take supplements containing vitamin C and have a reasonable dietary intake of vitamin C do better than those who simply have a reasonable dietary intake,” was misleading because the supplement takers ate substantially more fruits and vegetables (including substantially more dietary vitamins A and C) than those who did not take supplements. Instead of using creative “water logic” [2] to objectively assess epidemiologic data, the editorial misinterpreted the Enstrom paper to fit the “rock logic” [2] preconception that vitamin C supplements prevent disease and extend life.

The bottom line: Both oxidants and antioxidants are needed in the biochemical economy of human cells. We need to inhale oxygen because it is the fundamental oxidant we use; without it we would die. Cells walk a balance between essential oxidant and essential antioxidant processes. That is the fundamental reason moderate antioxidation is helpful and excessive antioxidation is harmful. Vitamin C is a redox agent, usually antioxidant in the moderate quantities found in food, but often oxidant in the large quantities found in many supplements. [6,21] As a supplement, it can be antioxidant or oxidant, depending on circumstances and is particularly oxidant when taken with iron or when body iron stores are high.


Because of the potential lethality of vitamin C supplements in persons with iron overload, the “1992 Management Protocol for the Treatment of Thalassemia Patients” distributed by the Thalassemia International Federation, states (page 15):


Iron-loaded patients usually become vitamin C deficient, probably because iron oxidizes the vitamin. When this is the case, administration of vitamin C increases excretion of iron in response to Desferal[R]. Vitamin C increases the availability of iron, and so may increase its toxicity if large doses are taken without simultaneous Desferal(R) infusion. Therefore the following precautions are recommended:

a. Start treatment with vitamin C only after an initial month of treatment with Desferal(R).

b. Give vitamin C supplements only if the patient is receiving Desferal(R) regularly.

c. Do not exceed a daily dose of 200 mg. The minimum effective dose of vitamin C is about 2-5 mg/kg (N. Di Palma, A. Piga unpublished data). In general, 50 mg suffice for children under 10 years of age, and 100 mg for older children. Vitamin C should be given only on days when Desferal(R) is taken, ideally when the pump is set up.

The Protocol is available in the United States from the Cooley’s Anemia Foundation, Box CEP, 105 East 22nd Street, New York, NY 10010.

* Desferal(R) is the iron-chelating agent, desferrioxamine.


1. Enstrom JE, Kanin LE, Klein MA. Vitamin C intake and mortality among a sample of the United States population. Epidemiology 1992; 3:194-200.

2. De Bono E. I am right–you are wrong. New York: Viking, 1991.

3. Repka T, Hebbel RP. Hydroxyl radical formation by sickle erythrocyte membranes: role of pathological iron deposits and cytoplasmic reducing agents. Blood 1991;78:2753-8.

4. Herbert V, Jayatilleke E, Shaw S. Alcohol and breast cancer. N Engl J Med 1987;317:1287-8.

5. Shaw S, Herbert V, Colman N, Jayatilleke E. Effect of ethanol-generated free radicals on gastric intrinsic factor and glutathione. Alcohol 1990;7:153-7.

6. Herbert V. Iron disorders can mimic anything, so always test for them. Blood Rev 1992;3:125-32.

7. Sadrzadeh SMH, Eaton JW. Hemoglobin-mediated oxidant damage to the central nervous system requires exogenous ascorbate. J Clin Invest 1988;82:1510-5.

8. Ottolenghi A. Interaction of ascorbic acid and mitochondrial lipids. Arch Biochem Biophys 1959;79:353-63.

9. Barber AA. Lipid peroxidation in rat tissue homogenates: interaction of iron and ascorbic acid as the normal catalytic mechanism. Lipids 1966;1:146-51.

10. Sharma SK, Knshna Murti CR. Production of lipid peroxide by brain. J Neurochem 1968; 151:147-9.

11. Sharma SK, Krishna Murti CR. Ascorbic acid: a naturally occurring mediator of lipid peroxide formation in rat brain. J Neurochem 1976;27:299-301.

12. Zaleska MM, Floyd RA. Regional lipid peroxidation in rat brain in vitro: possible role of endogenous iron. Neurochem Res 1985;10:397-410.

13. Bucher JR, Tien M, Morehouse LA, Aust SD. Redox cycling and lipid peroxidation: the central role of iron chelates. Fundam Appl Toxicol 1983;3:222-6.

14. Herbert V. Prevalence of abnormalities of iron metabolism in the U.S.A. In Serum ferritin: a technical monograph. La Jolla, CA: National Health Laboratories, 1989:3-8.

15. Edwards CQ, Griffen LM, Kushner JP. Disorders of excess iron. Hosp Pract 1991;26(suppl 3):30-6.

16. Krikker MA. A joint statement in support of RDIs replacing US RDAs. Submitted February 24, 1992, to FDA Dockets Management Branch, Docket No. 90N-194.

17. Herbert V. Statement in support of RDIs replacing US RDAs. Submitted February 23, 1992, to FDA Dockets Management Branch, Docket No. 90N-194.

18. Anonymous. Herbert says FDA-proposed US RDIs protect consumers. Food Chem News 1992;March 9:11-2.

19. Herbert V. Introduction and medicolegal considerations: symposium on diagnosis and treatment of iron disorders. Hosp Pract 1991; 26(suppl 3):4-6.

20. Salonen JT, Nyyssonen K, Korpela H, Tuomilehto J, Seppanen R, Salonen R. High stored iron levels are associated with excess risk of myocardial infarction in Eastern Finnish men. Circulation 1992;86:803-11.

21. Herbert V. Everyone should be tested for iron disorders. J Am Diet Assoc 1992;92:1502-9.

22. Olson JA, Hodges RE. Recommended dietary intakes (RDI) of vitamin C in humans. Am J Clin Nutr 1987;45:693-703.

23. Herbert V. The 1989 RDA is mainly the work of the 1980-85 (l0th) RDA Committee, but with 9th RDA numbers for vitamins A and C. FASEB J 1990;4:A374.

24. Herbert V. Vitamin C and iron overload. N Engl J Med 1981;304:1108.

25. Cohen A, Schwartz E. Vitamin C and iron overload. N Engl J Med 1981;304:1108.

26. Marshall CW. In: Barrett S, ed. Vitamins and minerals: help or harm? Mount Vernon, NY: Consumers Union, 1985.

27. Herbert V, Subak-Sharpe G, Hammock D, eds. The Mount Sinai School of Medicine complete book of nutrition. New York, St. Martin’s Press, 1990.

28. Subcommittee on the Tenth Edition of the RDAs (Eds). Recommended Dietary Allowances, 10th Edition. Washington, DC, National Academy Press, 1989.

29. Herbert V. Recommended dietary intakes (RDI) of iron in humans. Am J Clin Nutr 1987; 45:679-86.

30. Stone R. NAS plagiarism fight to go another round. Science 1992;258:19.

31. Sullivan JL. Stored iron and ischemic heart disease: empirical support for a new paradigm. Circulation 1992;86:1036-7.

32. Ames B, Profet M, Gold LS. Dietary pesticides (99.99% all natural). Proc Natl Acad Sci USA 1990;87:7777-81.

33. Ames B, Profet M, Gold LS. Nature’s chemicals and synthetic chemicals: comparative toxicology. Proc Natl Acad Sci USA 1990;87:7782-86.

34. Prochaska HJ, Santamaria AB, Talalay P. Rapid detection of inducers of enzymes that protect against carcinogens. Proc Nail Acad Sci USA 1992;89:2394-8.

35. Zhang Y, Talalay P, Cho C-G, Posner GH. A major inducer of anticarcinogenic protective enzymes from broccoli: isolation and elucidation of structure. Proc Natl Acad Sci USA 1992;89:2399-2403.

36. Liebet, CS. Alcohol, liver, and nutrition. J Am Coll Nutr 1991;10:602

37. Simopoulos A, Herbert V, Jacobson B. Genetic nutrition: designing a diet based on your family medical history. New York: Macmillan, 1993:

38. Zalusky R, Herbert V. Megaloblastic anemia in scurvy with response to fifty micrograms of folic acid daily. N Engl J Med 1961;265:1033-8.

39. Herbert V. Vitamin C and iron overload. N Engl J Med 1981;304:1108.

40. Feldman EB, Gold S, Greene J, Moran J, Xu G, Shultz GG, Feldman DS, Hames CG. Vitamin C administration and blood pressure regulation. Clin Res 1992;40:627A.

41. Hoffman M. A new role for gases: neurotransmission. Science 1991;252:1788.

42. Block G. Vitamin C and reduced mortality. Epidemiology 1992;3:189-91.

Dr. Herbert is chair of the Committee to Strengthen Nutrition of the Mount Sinai School of Medicine of the City University of New York, and chief of the Hematology and Nutrition Research Laboratory of the Research Service, Bronx Veterans Affairs Medical Center, New York.

This review was supported primarily by the V. Herbert Research Fund of the Mount Sinai School of Medicine.

COPYRIGHT 1993 Lippincott/Williams & Wilkins

COPYRIGHT 2004 Gale Group