Letters To The Editor

oxidative stress but also to HIV replication and disease progression. Although animal studies are exciting, results from these studies cannot be used as claims of efficacy in the treatment of human diseases unless proper clinical trials are performed. For example, vitamin E may slow cancer development in murine AIDS (5), but in human smokers (6), another oxidatively stressed population with antioxidant deficiencies, supplementation with vitamin E and !'I-carotene actually increased risk for lung cancer. What about the HIV population? Observational studies (7, 8) suggest that over-thecounter multivitamin and antioxidant vitamin supplements may have some benefits and slow the progression to AIDS. On the basis of our study (4), I agree with Watson that such supplements may be used in an attempt to overcome antioxidant deficiencies. However, it is not yet proven that these supplements given to humans infected with HIV will result in the same efficacy as reported in in vitro or animal studies. So far, few trials have addressed this issue. In HIV-infected patients, a trial investigating the effect of !'I-carotene supplementation on immune function was negative (9) and another with selenium and 13-carotene produced mixed results (IO). Therefore, from our point of view, although antioxidant supplementation has been proven to be of benefit in animal models, the data on supplementation in humans are unfortunately lacking. We hope to be able to remedy this in the near future; we have now documented increased oxidative stress and antioxidant micronutrient deficiencies in humans infected with HIV. As for "the mechanisms underlying the increased oxidative stress in the HIV population," we agree with Watson that it is likely these are related mostly to activated polymorphonuclear leukocytes and cytokine production, as mentioned in the introduction of our article (4).