Resource Logo
AIDS Treatment News

Some Vitamins Associated with Decreased Risk of AIDS and Death


AIDS TREATMENT NEWS Issue #214, January 6, 1995

In December 1993, researchers at Johns Hopkins University published results from a study (funded by the U.S. National Institutes of Health) of 280 volunteers, who were studied for an average followup of 6.8 years, which found that dietary intake of certain vitamins was associated with substantially reduced risk of progression to AIDS.(1) In the second half of 1994, they reported on the association of the same micronutrients with survival, in posters presented at the International Conference on AIDS (Yokohama, August 1994)(2) and at the annual meeting of the American Public Health Association (Washington, October-November 1994).(3) Both results are important because the differences are large; for certain micronutrients, persons who used appropriate amounts had as little as half the risk of progression to AIDS -- and half the risk of dying -- compared to others who used too little (or, in some cases, too much).

These results must be interpreted with caution, for several reasons: (1) First, they are not from clinical trials which randomly assign people to take particular treatments or foods. Instead, they are from an observational study which asked people what they were eating and what supplements they were taking, then followed up to see how well they did during the following years. This kind of study can only show an association -- it cannot prove a cause and effect relationship -- between the nutrient and the result.

In particular, the nutrients studied were in large part obtained from foods, which also contains many other substances which were not looked at in this research. Therefore, it is not certain that taking supplements of the individual nutrients alone would automatically provide the same advantage.

(2) Also, the food questionnaire was given at the beginning of the study, in 1984, and that data was used throughout; if people changed their dietary habits or use of supplements over the years, the change was not considered. (Random changes in food intake or supplement use would tend to reduce the differences found in this study, however, not to increase them. Therefore, we do not think that such changes in food usage could be used to argue that the results of the study are not real.) Note: the food questionnaire was repeated on certain later visits, but the newer food data has not yet been analyzed (3) A third reason for caution is that people might misinterpret the results of the Johns Hopkins study and take megadoses of certain vitamins. Not only could this be dangerous, it is also unnecessary, since the improved outcomes seen in this study were associated with modest doses.

Despite these cautions, the findings on micronutrients and AIDS progression urgently need more attention. We may not have fundamentally better information for many years, since conclusive proof might require clinical trials with survival or AIDS-progression endpoints, which would take years to run; it is possible that such trials will never be conducted, due to increasing uncertainty of government financial support, and lack of pharmaceutical interest in research which does not sell a product. Meanwhile, failure to consider safe, simple strategies which might reduce the risk of AIDS and of death by 40 to 50 percent or more would be tragic.

The Johns Hopkins study is similar in some ways to a major San Francisco study of micronutrients and AIDS,(4) which was reported in 1993 (see summary in AIDS TREATMENT NEWS #181, August 20, 1993). This study also reported that higher intake of certain nutrients may delay the development of AIDS.

The Zinc Controversy There is much disagreement and confusion among experts as to whether zinc supplements may be harmful or helpful for persons with HIV (or possibly harmful in some cases and helpful in others). The Johns Hopkins data suggests that even small amounts of zinc -- close to the U.S. Recommended Daily Allowance of 15 mg -- might be harmful. The researchers did not suggest a possible mechanism.

But many other nutrition experts are skeptical. They cite other studies which found different results -- or which found widespread zinc deficiencies in people with HIV, deficiencies which might interfere with immune functioning.

In this article on the Johns Hopkins results, we will report the findings for zinc along with the other findings, but leave our analysis for later. We may report about zinc in a separate article, based on interviews with experts. However, we suspect that clinical trials might be necessary for resolving this important issue and making an appropriate nutritional recommendation -- trials which may never be done. (One trial we would like to see, which could be run quickly and inexpensively, would be to see if viral load changes when people with HIV start using zinc supplements.) In any event, the confusion about zinc should not distract from the other study results.

Background: The Study Methodology The Johns Hopkins data on micronutrients and AIDS progression came from about 280 patients who enrolled in a major, ongoing national study -- the Multicenter AIDS Cohort Study (MACS), funded by the National Institute of Allergy and Infectious Diseases (NIAID). This study enrolled about 5,000 gay men (both HIV positive and HIV negative) by early 1985, and conducts followup visits every six months for blood tests, physical examinations, and interviews. The 5,000 volunteers are in four cohorts, in Baltimore, Chicago, Los Angeles, and Pittsburgh.

The Baltimore cohort has over 1,000 men, but only those who were HIV positive at their first (baseline) visit, and completed an extensive food questionnaire on their second visit, were included. (Also, anyone who had AIDS by their third visit -- a year after they entered the study -- was excluded. The reason for this is that the researchers wanted to learn how diet influenced the progression of early HIV disease. They did not want their results to be complicated by changes in diet that people may have made as a result of their illness.) Micronutrient data was obtained by asking the men what supplements (vitamin pills, etc.) they used, and also by a food questionnaire. The questionnaire listed 116 foods, and asked participants to rate how often they ate each (choosing one of nine categories from "Never, or less than once per month" to "6+ per day") on the average over the last year. Then special software developed for use with this questionnaire estimated the total amount of each nutrient in the diet, based on those answers.

For each nutrient, data was analyzed by grouping the volunteers into four equal groups based on their total intake of that nutrient (both from foods and from supplements). Usually the top "quartile" (the 25 percent of the volunteers who used the most of the particular nutrient) was compared with the other three quartiles averaged together. But vitamin A/beta carotene was handled differently, because it has shown a U-shaped curve in other studies, with those taking moderate amounts doing better than those taking either too much or too little; for vitamin A, therefore, the two middle quartiles were compared with the two extreme ones. And zinc was also handled differently, with each quartile being compared to the lowest, in order to show the dose response.

The differences in progression to AIDS, or in death, between high and low users of particular supplements was given as "relative hazard." For example, a relative hazard of 0.60 means that one group had only 60 percent of the risk of the other group of progressing to AIDS (or death) within the time frame of the study (6.8 years or more).

All the results given below are statistically significant, unless otherwise stated. This means that it is unlikely that they would have occurred by chance alone.

We report the results in two separate sections, one on the risk of AIDS prevention, the other on survival.

Results: Risk of AIDS Progression The work on micronutrients and AIDS progression was reported in a 14-page paper published a year ago.(1) The last paragraph summarized the findings: "In summary, we have found that high intake of several nutrients (niacin, vitamin C, and vitamin B1) were associated with slower progression to AIDS, after adjustment for confounding variables. Vitamin A may have a U-shaped relation with risk of AIDS, and increasing zinc intake was associated with more rapid disease progression. These results need to be replicated before any firm conclusions can be drawn about their relevance to the natural history of HIV-1 infection." We discuss some of the numbers below. Note that the nutrient amounts given in this article are NOT recommendations, for many reasons. One is that these figures for nutrient intake include both food and supplements, not supplements alone. Also, there is no adjustment for body weight, or for individual medical conditions. And the situation with vitamin A is more complex, because the amounts are given in international units of vitamin A, while in fact, the relation found was between beta carotene intake and AIDS (beta carotene is converted to vitamin A in the body, and is safer to take than vitamin A itself); no relation was found between vitamin A intake and AIDS. Professional assistance will be required to develop recommendations for individuals, or to prepare general guidelines, based on this micronutrient study and other information.

For vitamin C, the volunteers in the highest quartile had a relative hazard of 0.55 of progression to AIDS -- meaning that they were scarcely more than half as likely as those who used less vitamin C to progress during the study. Those in this upper quartile took 715 mg (0.715 grams) or more of vitamin C per day, in food and supplements combined.

For vitamin B1, those in the highest quartile had a relative hazard of 0.60 of progression to AIDS, compared to those who took less. Those in this quartile took greater than 4.9 mg of vitamin B1 per day, in food and supplements combined.

For niacin, those in the highest quartile had a relative hazard of 0.52. Those in this quartile took greater than 61 mg of niacin per day total (in their food and supplements combined).

For vitamin A, those who did best were in the two middle quartiles, with total intake (mostly from beta carotene) from food and supplements combined being equivalent to between 9,062 IU (international units) and 20,268 IU. Their relative hazard of progression to AIDS was 0.55, compared to the those in the lower quartile (who consumed less than 9,062 IU per day). The upper quartile (who consumed a total of more than 20,268 IU) had a relative hazard of 0.94 (not statistically significant), compared to the lower quartile -- meaning that the risk of progression for these two groups was about equal.

For zinc, those in the highest quartile (consuming more than 20.2 mg per day in food and supplements combined) had a relative hazard of 2.06 of progression to AIDS, compared to those in the lowest quartile (less than 11.7 mg per day). [This is a cause for concern in that the U.S. RDA is only 15 mg per day -- and many, probably most, of the multivitamin/mineral pills in common use contain the 15 mg RDA. Yet taking only 20.2 mg of zinc per day, from food and supplements combined, was found to be associated with double the risk of progression to AIDS, compared to those who used less zinc.] These particular numbers depended on how the statistics were analyzed. The numbers above were from the analysis which the researchers chose to use in their abstract. These numbers (a "single nutrient model") look at each nutrient alone, statistically adjusted for age, symptoms, T-helper count, lymphocyte count, energy intake, use of antiretrovirals, and use of pneumocystis prophylaxis. A different analysis -- which statistically adjusted with some of the nutrients together, not alone -- did not change the relative hazards much, except for zinc, where the relative hazard increased, from 2.06 to 2.97 for the highest quartile (those consuming over 20.2 mg per day).

Two other nutrients barely missed being statistically significant for risk of AIDS progression. They are vitamin B2 (relative hazard 0.61, for those consuming more than 5.9 mg in food and supplements together, vs. those consuming less), vitamin B6 (relative hazard 0.60 for those consuming greater than 5.7 mg per day, vs. those consuming less then 2.0 mg). However, both of these were significant in the analysis of relative hazard of death, below.

Results: Survival The full results on micronutrient intake and probability of survival have not yet been published. The most complete data made public so far was presented at a poster at the annual meeting of the American Public Health Association, October- November 1994 in Washington. D.C.(3) For beta carotene, the third quartile of intake was associated with increased survival (relative hazard of death, 0.58), and increasing amounts of zinc were associated with poorer survival. After the data was statistically adjusted for beta carotene and zinc (as well as for age, symptoms, T- helper count, lymphocyte count, energy intake, and treatment), the following were statistically significant: vitamin B1 (relative hazard 0.62), vitamin B2 (relative hazard 0.61), vitamin B6 (relative hazard 0.47), and niacin (relative hazard 0.59).

The above relative hazards relate to total intake of the nutrient, from both foods and supplements. If one looks at supplements alone, the results are sometimes different. Those who took vitamin B6 in supplements at more than twice the RDA had improved survival (relative hazard of death, 0.63). Any intake of zinc supplements was associated with poorer survival (relative hazard, 1.52, meaning that those who took zinc supplements had a .52 greater chance of dying; the median amount of zinc supplement taken was 15 mg per day, which is the same amount as the RDA). Use of vitamins B1 and B2 at five times the RDA or more was associated with better survival, but this result was not statistically significant. Information on niacin from supplements alone is not available, because the software did not break out this number. And while total vitamin C may have been associated with increased survival, no relationship was found between vitamin C from supplements alone and survival.

Comment This observational study cannot prove that the supplements are causing the changes in progression to AIDS and in survival. To get definitive proof will require clinical trials -- a different large trial for each nutrient -- which, as we pointed out before, will take a long time, if they are done at all.

It might be possible to get some insights into what is happening by doing small, short, exploratory trials, which randomize patients to nutritional supplements and look at changes in viral load, T-cell subsets, tests of immune functioning, and other markers. If the results of these trials parallel the results of the observational study -- for example, if the optimal amount of vitamin A leads to improvement in the markers, while too much leads to worsening -- this would help to confirm the observational results, and open doors to further rapid testing of combinations of nutrients, and to the use of nutrients to enhance drug therapy. But no one knows in advance which markers, if any, would be appropriate, since we do not know the possible mechanisms of action of the various nutrients.

Still, the fact remains that certain safe, inexpensive nutrients may be associated with a greater long-term reduced progression of HIV disease, and improved survival, than that of any known anti-HIV drug. The key question, of course, is whether there is cause and effect -- whether improving nutrient level will improve disease outcome.

Or could it be that persons with HIV who tend to take certain levels of nutrients also tend to do better, for other reasons? Perhaps those who eat well and use reasonable levels of supplements also tend to have better access to health care, or take better care of their health in other ways. Or maybe those who are healthier anyway are more likely to eat well. If theories like these account for the associations found in the micronutrient study, then changing one's nutrient intakes to match those associated with favorable outcome might not be of any benefit.

But it seems unlikely that these theories could explain the results observed. How would they explain why some nutrients were associated with improvement in the study, while zinc was consistently and unexpectedly associated with worse outcome? How would they explain the U-shaped curve of the response to vitamin A/ beta carotene (also seen in other studies), where a moderate amount seemed to be better than either too much or too little? We will be living with uncertainty about micronutrients and HIV disease for years to come. We need to make the best decisions possible now and at each future time, instead of waiting for perfect information. What we know now strongly suggests that rational strategies for improving micronutrient intake can improve disease outcome. And since the strategies suggested are generally quite safe, the cost of adopting them in error is far less than the cost of ignoring the strategies if, in fact, they do work.

For too long the micronutrient work outlined in this article has received little attention, partly because people are afraid to make nutritional recommendations, due to the possibility that they might be wrong. What we need now is for medical and scientific experts to examine all the information available, and develop guidelines (by consensus when possible) for use by physicians and patients.


A new book which should be out this Spring is POSITIVELY WELL: LIVING WITH HIV AS A CHRONIC, MANAGEABLE, SURVIVABLE DISEASE, by Lark Lands (Irvington Publishers Inc., New York, 1995). Copies can be ordered by calling 800/542-8102, 9-5 Eastern time, to be shipped when the book is available.

References 1. Tang AM, Graham NMH, Kirby AJ, McCall LD, Willett WC, and Saah, AJ. Dietary micronutrient intake and risk of progression to acquired immunodeficiency syndrome (AIDS) in human immunodeficiency virus type 1 (HIV-1)-infected homosexual men. AMERICAN JOURNAL OF EPIDEMIOLOGY December 1993; volume 138, number 11, pages 937-951.

2. Tang AM, Graham NMH, and Saah AJ. The effect of micronutrient intake on survival in HIV-1 Infection. Tenth International Conference on AIDS, Yokohama, August 7-12, 1994 [abstract PB0894].

3. Tang AM, Graham NMH, and Saah AJ. The effect of micronutrient intake on survival time in HIV-infected gay and bisexual men. American Public Health Association 122nd Annual Meeting, Washington, DC, October 30 - November 3 [abstract book, page 2082].

4. Abrams B, Duncan D, and Hertz-Picciotto I. A prospective study of dietary intake and acquired immune deficiency syndrome in HIV-seropositive homosexual men. JOURNAL OF ACQUIRED IMMUNE DEFICIENCY SYNDROMES August 1993; volume 6, number 8, pages 949-958.


Copyright © 1995 -AIDS Treatment News, Publisher. All rights reserved to AIDS Treatment News (ATN), Email AIDS Treatment News .

Information in this article was accurate in January 6, 1995. The state of the art may have changed since the publication date. This material is designed to support, not replace, the relationship that exists between you and your doctor. Always discuss treatment options with a doctor who specializes in treating HIV.