Several factors have been demonstrated to influence the progression of HIV-1 disease, including the biologic properties of HIV-1, coinfection with other viruses, genetic susceptibility, age, and the use of antiretrovirals(1-5). Altered nutritional status has also been shown(6,7) to influence the course of HIV-1 disease. These results are consistent with the occurrence of severe malnutrition and marked wasting that frequently characterize advanced disease(8,9). Although nutritional factors are not likely to be the most important etiologic determinants, they may have a significant impact on HIV-1 disease expression and related morbidity and mortality.
Our research studies(6,10) indicate that alterations in nutritional status are prevalent among HIV-1-infected individuals, even before the development of symptoms. Decreased plasma levels of zinc and vitamins B6, B12, A, and E are particularly widespread and appear to be functionally relevant in maintaining the integrity of immune responses, including the peripheral blood lymphocyte response to the mitogens phytohemagglutinin and pokeweed and natural killer cell activity(6,11). Moreover, development of nutrient deficiency has been associated with faster disease progression in vitamin A- or vitamin B12-deficient, HIV-1-infected men(6) and with increased mortality in vitamin A-deficient, HIV-1-seropositive drug users(7). Restoration of vitamin A, vitamin B12, and zinc status has been associated with slowing of HIV-1 disease progression in HIV-1-seropositive homosexual men(6).
To determine whether nutritional status affects survival, this longitudinal study examined the impact of specific micronutrient deficiency in combination with immune function and use of antiretroviral medication on mortality in HIV-1-infected men and women.
The HIV-1-seropositive participants in this study were drawn from a cohort of 230 persons who were recruited from a street drug-using population and engaged in high-risk behaviors for HIV-1 transmission; the group has been followed for 9 years. Participants had their HIV antibody status determined by standard HIV enzyme-linked immunosorbent assay antibody testing with Western blot confirmation. The 125 HIV-1-seropositive participants in this study were seen every 6 months in a community-based study clinic for 3.5 years. Blood was drawn at each evaluation for immune function and biochemical evaluation of nutritional status. A history, which included information on antiretroviral treatment (zidovudine), and physical examination were conducted at each visit. Mortality and causes of death during the study were confirmed by death certificates or by medical examiner reports.
Informed consent was obtained from all study subjects. The investigation was approved by the Human Studies Committee of the University of Miami School of Medicine.
Biochemical determinations of specific nutrient levels that are known to influence immune function and be altered in HIV-1-infected individuals were evaluated as previously described(10). Deficiency of each nutrient was based on clinically recognized standards. Briefly, total plasma vitamin B12 levels were established by radioisotope dilution assay, and deficiency was defined as plasma levels <200 pg/ml. An enzymatic assay was used to determine vitamin B6 status, expressed as an activity coefficient (AC), with an AC >1.85 considered to be evidence of deficiency. Total serum vitamins A and E were determined by high-performance liquid chromatography; measurements below 0.3μg/ml and <5 μg/ml for vitamins A and E, respectively, were considered inadequate. Selenium levels were determined by a standardized fluorometric method, with selenium reference obtained from the National Institute of Standards and Technology (Standard Reference Material 1598); using bovine serum; measurements <85 μg/L were considered deficient. Zinc analysis was performed using flame atomic absorption spectrophotometry, and plasma levels <0.75 μg/ml were considered deficient. Levels of prealbumin, which are a sensitive measure of subclinical malnutrition(12) were determined by radial immunodiffusion method, with levels <3.0 mg/100 ml considered inadequate.
Flow cytometry and monoclonal antibodies(13) were employed to quantitate numbers of circulatory helper/inducer T cells, defined as CD4+, CD3+, CD14-, and CD45+ cells.
Cox proportional hazards regression models were used to investigate the effect of micronutrient deficiency on survival of HIV-1-infected drug users. The end point was HIV-1-related death. Survival time was defined from the date of the first blood draw at baseline to date of death or the date when the study was closed (April 30, 1996). Because micronutrient levels changed over time, as did the deficiency status, micronutrient deficiency was treated as a time-dependent covariate in the Cox models. The association between a CD4 count <200/mm3 at baseline, CD4 count over time, antiretroviral treatment, overall malnutrition(i.e., low prealbumin), micronutrient deficiency, and mortality were first examined in individual univariate Cox models. To quantify the independent effect of micronutrient deficiency on survival, CD4 count over time and baseline CD4 count <200/mm3 were treated as covariates in the Cox models. Because gender, age, and antiretroviral treatment(categorical and time-dependent) were not shown to be associated with survival in the univariate Cox model, these variables were not included as covariates in the multivariate analysis.
The impact of poor overall nutrient status (i.e., low prealbumin levels) and deficiency of each micronutrient (i.e., vitamins A, B6, B12, and E; minerals selenium and zinc) on survival was first evaluated in a Cox model, controlling for a baseline CD4 cell count <200/mm3 and CD4 count over time individually. In a subsequent analysis, micronutrient deficiency variables that were significantly associated with survival in these models were then included in an overall model to investigate the joint effect of these variables on mortality and to identify micronutrient deficiency variables that are independently associated with survival.
The micronutrient deficiency variables that ceased to show significant contributions to the model were deleted, yielding a final model. All p values reported are two-sided.
The participants in this study included 82 men and 43 women who were between 27 and 62 years of age (mean of 42 ± 6 years). Most participants were African American(88%), with a smaller proportion of Hispanics (8%), and Caucasians (4%). The major drug of abuse was cocaine, with positive cocaine urine toxicologic results observed for 77.6% of the group. The mean CD4 cell count at baseline was 428 ± 322 (range 1528 to 2). Twenty-seven percent of the subjects had a CD4 cell count <200/mm3; 35% had CD4 counts between 200 and 500/mm3, and 38% had CD4 cell counts >500/mm3.
Nutritional Status and Mortality
Over the course of the 3.5-year follow-up, 21 participants died of HIV-1-related causes. Separate Cox proportional hazards models, which controlled for baseline CD4 cell counts <200/mm3 and CD4 cell counts over time, were used to determine the relation between specific micronutrient deficiency and mortality. These analyses, as shown in Table 1, indicated that impaired nutritional status (i.e., overly low levels of prealbumin) and deficiency of vitamin A, vitamin B12, zinc, and selenium over time were significantly associated with mortality. The risk ratio for a low prealbumin level was 4.0(p < 0.007); for vitamin A deficiency, 3.2 (p < 0.03); for vitamin B12, 8.3 (p < 0.009); and for zinc, 2.9 (p < 0.04). It was highest (19.9) for selenium deficiency over time (p< 0.0001). Neither vitamin B6 nor vitamin E deficiency was associated with an increased risk for mortality. The use of zidovudine and other antiviral regimens was not significantly associated with mortality.
Because multiple nutrient deficiencies tend to occur simultaneously and are highly correlated, a multivariate Cox model was used, adjusting for baseline CD4 cell counts and time-dependent CD4 counts, to identify independent nutritional factors associated with survival. All the nutrients that had significantly contributed to mortality in separate Cox proportional hazards models were included in this multivariate model in addition to CD4 counts <200/mm3 at baseline and CD4 counts over time. As indicated in Table 2, the results of this analysis reveal that only CD4 counts over time (RR = 0.69, p < 0.04) and selenium deficiency (RR = 10.8, p < 0.002) over time was significantly associated with mortality. Selenium deficiency was the only independent predictor of survival, even when controlling for CD4 counts <200/mm3 at baseline, CD4 counts over time, overall poor nutritional status, and deficiencies of vitamin A, vitamin B12, and zinc.
The results of our investigation reveal a dramatic, elevated risk for mortality among HIV-1-infected men and women with nutritional deficiencies. Subclinical malnutrition(i.e., low prealbumin levels) and individual deficiencies of vitamin A, vitamin B12, zinc, and selenium over time were shown to be associated with HIV-1-related mortality, independent of CD4 counts <200/mm3 at baseline and CD4 counts over time.
Because multiple nutrient deficiencies tend to occur simultaneously, the joint effect of deficiencies that singly predict HIV-1-related mortality was investigated. Our findings demonstrate that only a deficiency of selenium, an essential trace element that is part of the antioxidant defense system, was profoundly associated with decreased survival in HIV-1 disease. A significant effect for selenium deficiency was shown, even when controlling for deterioration of overall nutritional status, CD4 counts <200/mm3 at baseline, and CD4 counts over time. Antiretroviral treatment was not significantly associated with mortality, probably because of a very low frequency of use (1.7% to 12.8% of the population, mostly in the form of monotherapy) over the course of the study.
These findings extend those of earlier studies of associations between nutritional deficiencies, faster HIV-1 disease progression(6,14), and increased HIV-1-related mortality(7). The significant effect of selenium on disease progression does not appear to be related to drug use, because selenium deficiency has been associated with immune dysfunction(15) and decreased survival in a cohort of HIV-1-infected pediatric patients(16) and has been predictive of HIV-1-related prognosis in drug users and nondrug users(17). Although selenium deficiency in healthy humans is relatively rare(18,19), several studies have documented a decline in plasma selenium levels and decreased glutathione peroxidase activity in the general HIV-infected population(20-24), emphasizing the importance of maintaining optimal selenium status in all HIV-1-infected cohorts.
Our findings of the independent effect of selenium on HIV-1-related mortality support the suggestion that inadequate antioxidant function may be an important factor in AIDS patients(25-27). Erythrocyte oxidative stress has been associated with the progressive development of HIV disease(28), and evidence indicates an increased production of free radicals or a failure of antioxidant mechanisms in asymptomatic HIV-infected adult patients(29,30) and children (31). Antioxidant defense mechanisms primarily involve glutathione and glutathione peroxidase intracellularly, with α-tocopherol as the major hydrophobic membrane antioxidant(32). Selenium, an integral part of the enzyme glutathione peroxidase, increases the oxygen consumption of cells and appears to protect cells and membranes against peroxidative damage.
Despite the importance of other nutrients, including tocopherols, in immune function and as protection against oxidative stress, only selenium deficiency was significantly associated with HIV-related mortality. This finding may be related to selenium's action in gene regulation(33,34). At the genetic level, selenium exerts its effects by means of selenoproteins that incorporate selenocysteine at the messenger RNA codon UGA, which serves to terminate protein synthesis in other contexts(35-37). Data consistent with the possibility of viral selenoproteins have been reported(35,36,38). Several genomic features required for the expression of potential selenoprotein genes in HIV-1 have been experimentally confirmed(39,40). Moreover, in vitro studies have demonstrated that selenium inhibits viral cytotoxic effects and the reactivation of HIV-1 by hydrogen peroxide and that it protects against activation of HIV-1 by tumor necrosis factor(34). Several selenium compounds specifically inhibit HIV-1 cytopathic effects at micromolar concentrations, comparable to plasma selenium levels in healthy individuals(41).
Selenium status may have an important role in modulating viral expression, with adequate selenium preventing the replication of HIV and slowing the progression of HIV-1 disease. A decline in selenium levels would be expected to increase viral replication and accelerate disease progression. Our data indicate that selenium deficiency is a significant predictor of HIV-1-related mortality, independent of CD4 counts over time, CD4 counts<200/mm3 at baseline, and antiretroviral treatment. These findings are of particular concern in light of selenium's influence on immune function, viral replication, and survival. Because investigations(42) have demonstrated that supplementation with selenium may help to increase the enzymatic defense systems in HIV-1-infected patients, further studies to determine possible mechanisms and clinical trials to evaluate the effect of selenium supplementation on HIV-1 disease progression are essential.
Acknowledgments: We thank K. Bridbord, M.D., for his enthusiastic support of this work and N. Rodriguez for excellent technical assistance. This study was supported by NIMH grant #5R01MH50239-05 (M.K.B.), NIH and Fogarty International Center #5D43TW00017-08 (M.K.B.), and NIMH grants P50MH42455 (M.K.B.) and 1R01-1MH50240 (J.B.P).
1. Polk BF, Fox R, Brookmeyer R, et al. Predictors of the acquired immunodeficiency syndrome developing in a cohort of seropositive homosexual men. N Engl J Med
2. Mann D, Murray C, Yarchoan R, Nlattner WA, Goedert JJ. HLA antigen frequencies in HIV-1 seropositive disease-free individuals and patients with AIDS. J Acquir Immune Defic Syndr
3. Cheng-Meyer C, Seta D, Tateno M, Levy JA. Biological features of HIV-1 that correlate virulence in the host. Science
4. Blaxhult A, Granath F, Lidman K, Giesecke J. The influence of age on the latency period to AIDS in people infected by HIV through blood transfusion. AIDS
5. Des Jarlais DC, Friedman SR, Marmor M, et al. Development of AIDS, HIV seroconversion, and potential cofactors for T4 cell loss in a cohort of intravenous drug users. AIDS
6. Baum MK, Shor-Posner G, Lu Y, et al. Micronutrients and HIV-1 disease progression. AIDS
7. Semba RD, Graham N, Caiaffa WT, Margolick JB, Vlement L, Vlahov D. Increased mortality associated with vitamin A deficiency during human immunodeficiency virus type 1 infection.Arch Intern Med
8. Serwadda D, Murgewa D, Sewankamboo NK, et al. Slim disease: new disease in Uganda and its association with HTLV-III infection.Lancet
9. Malebranchi R, Arnoux E, Guerin JM, et al. Acquired immunodeficiency syndrome with severe gastrointestinal manifestations in Haiti. Lancet
10. Beach RS, Mantero-Atienza E, Shor-Posner G, et al. Specific nutrient abnormalities in asymptomatic HIV-1 infection. AIDS
11. Baum MK, Mantero-Atienza E, Shor-Posner G, et al. Association of vitamin B6
status with parameters of immune function in early HIV-1 infection. J Acquir Immune Defic Syndr
12. Shetty PS, Jung RT, Stampler M, et al. Rapid turnover transport proteins: an index of subclinical protein-energy malnutrition. Lancet
13. Fletcher MA, Gjerset GJ, Hassett J, et al. and the Transfusion Safety Study. Effect of age on HIV-1 induced changes in lymphocytes populations among persons with congenital clotting disorders.Blood
14. Graham NMH, Sorensen D, Odaka N, et al. Relationship of serum copper and zinc levels to HIV-1 seropositivity and progression to AIDS. J Acquir Immune Defic Syndr
15. Bologna R, Indacochea F, Shor-Posner G, et al. Selenium and immunity in HIV-1-infected pediatric patients. J Nutr Immunol
16. Campa A, Shor-Posner G, Scott GB, et al. Mortality risk in selenium deficient HIV+ children. AIDS
17. Constans J, Pellegrin JL, Sergeant C, et al. Serum selenium predicts outcome in HIV infection [letter]. J Acquir Immune Defic Syndr
18. Cohen HJ, Brown M, Hamilton D, Lyons-Patterson J, Avissar N, Liegey P. Glutathione peroxidase and selenium deficiency in patients receiving home parenteral nutrition: time course for development of deficiency and repletion of enzyme activity in plasma and blood cell. Am J Clin Nutr
19. Lockitch G. Selenium: clinical significance and analytical concepts.Crit Rev Clin Lab Sci
20. Dworkin BM, Rosenthal WS, Wormser GP, et al. Abnormalities of blood selenium and glutathione peroxidase in patients with acquired immunodeficiency syndrome and AIDS-related complex.Biol Trace Elem Res
21. Cirelli A, Ciardi M, De Simone C, et al. Serum selenium concentration and disease progress in patients with HIV infection. Clin Biochem
22. Mantero-Atienza E, Sotomayor MG, Shor-Posner G, et al. Selenium status and immune function in asymptomatic HIV-1 seropositive men. Nutr Res
23. Staal FJT, Ela SW, Roederer M, Anderson MT, Herzenberg LA. Glutathione deficiency and human immunodeficiency virus infection.Lancet
24. Allavena C, Dousset B, May T, Dubors F, Canton P, Nelleville F. Relationship of trace element, immunological markers, and HIV-1 infection progression. Biol Trace Elem Res
25. Papadopulos-Eleopulos E. Reappraisal of AIDS-is the oxidation induced by the risk factors the primary cause? Med Hypotheses
26. Revillard JP, Vincent CMA, Favier AE, et al. Lipid peroxidation in human immunodeficiency virus infection. J Acquir Immune Defic Syndr
27. Sappey C, Favier A, Leclercq P, Micoud M, Faure P, Pelen F. Relative decrease in antioxidant status during evolution of HIV infection: effect on lipid peroxidation [Abstract 7502]. Presented at the VIIIth International Conference on AIDS. Amsterdam, The Netherlands, July 19-24, 1992.
28. Repetto M, Reides C, Gomez Carrelero ML, Costa M, Griemberg G, Llesuy S. Oxidative stress in blood of HIV-infected patients.Clin Chim Acta
29. Pace GW, Leaf CD. The role of oxidative stress in HIV disease.Free Radic Biol Med
30. Sonnerberg A, Cailin G, Akerlund B, Jarstrand C. Increased production of malondialdehyde in patients with HIV infection. Soc J Infect Dis
31. Malvy DJ-M, Richard M-J, Arnaud J, Favier A, Amedee-Manesme O. Relationship of plasma malondialdehyde, vitamin E and antioxidant micronutrients to human immunodeficiency virus-1 seropositivity.Clin Chim Acta
32. Halliwell B. Oxidants and human disease: some new concepts.FASEB J
33. Shu-Yu Y, Lu XP, Liao SD. Selenium stimulates c-fos
and suppresses c-myc
gene expression in human hepatoma cells. In: Collery P, Priorier LA, Manfait M, Etienne JC, eds. Metal ions in biology and medicine
. Paris: John Libbey Eurotext, 1990:487.
34. Sappey C, Legrand-Poels S, Best-Belpomme M, Favier A, Rentier B, Peitte J. Stimulation of glutathione peroxidase activity decreases HIV type 1 activation after oxidative stress. AIDS Res Hum Retroviruses
35. Taylor EW, Ramanathan CS, Jallurl RK, Nadimpalli RG. A basis for new approaches to the chemotherapy of AIDS: novel genes in HIV-1 potentially encode selenoproteins expressed by ribosomal frameshifting and termination suppression. J Med Chem
36. Senkevitch G, Bugert JJ, Sisler JR, Koonin EV, Darai G, Moss B. Genome sequence of a human tumorigenic poxvirus: prediction of specific host response-evasion genes. Science
37. Bock A, Forchammer K, Heider J, et al. Selenocysteine: the 21st amino acid. Mol Microbiol
38. Ramanathan CS, Taylor EW. Computational genomic analysis of hemorrhagic fever viruses: viral selenoproteins as a potential factor in pathogenesis. Biol Trace Elem Res
39. Battigello JMA, Cui M, Roshong S, Carter B. Enediyne mediated cleavage of RNA. Biol Org Med Chem
40. Taylor EW, Nadimpalli RG, Ramanathan CS. Genomic structures of viral agents in relation to the biosynthesis of selenoproteins. Biol Trace Elem Res
41. Taylor EW, Ramanathan CS, Nadimpalli RG, Schinazi RF. Do some viruses encode selenoproteins? Assessment of the theory in the light of current theoretical, experimental and clinical data. Antiviral Res
42. Delmas-Beauvieux M-C, Peuchant E, Coucouron A, et al. The enzymatic antioxidant system in blood and glutathione status in human immunodeficiency virus (HIV)-infected patients: effects of supplementation with selenium or β-carotene. Am J Clin Nutr