|Home | About | Journals | Submit | Contact Us | Français|
Our objective was to describe the prevalence of low concentrations of retinol, β-carotene, and vitamin E in a group of HIV-infected Latin American children and a comparison group of HIV-exposed, uninfected children. Our hypothesis was that the rates of low concentrations of these micronutrients would be higher in the HIV-infected group than those in the HIV-exposed, uninfected group. This was a cross-sectional substudy of a larger cohort study at clinical pediatric HIV centers in Latin America. Serum levels of micronutrients were measured in the first stored sample obtained after each child’s first birthday by high-performance liquid chromatography. Low concentrations of retinol, β-carotene and vitamin E were defined as serum levels below 0.70 μmol/L, 0.35 μmol/L and 18.0 μmol/L, respectively. The population for this analysis was 336 children (124 HIV-infected, 212 HIV-exposed, uninfected) aged ≥ 1 to < 4 years of age. Rates of low concentrations were 74% for retinol, 27% for β-carotene, and 89% for vitamin E. These rates were not affected by HIV status. Among the HIV-infected children those treated with antiretrovirals were less likely to have retinol deficiency, but no other HIV-related factors correlated with micronutrient low serum levels. Low concentrations of retinol, β-carotene and vitamin E are very common in children exposed to HIV living in Brazil, Argentina and Mexico, regardless of HIV-infection status.
In the second decade of the human immunodeficiency virus (HIV) epidemic, research related to nutrition in HIV/AIDS, a condition associated with increased oxidative stress, began to focus on micronutrients, largely because of their role as immunomodulators and antioxidants (1, 2). Adults in all stages of HIV infection have been found to have low circulating concentrations of micronutrients including vitamin E, vitamin A and carotenoids (3). Severe vitamin E deficiency early in life has been associated with stunted cognitive development (4). Retinol, β-carotene and vitamin E deficiencies increase the incidence and severity of infections by impairing T- and B-cell function (5). Retinol may inhibit HIV-associated apoptosis of peripheral blood lymphocytes (6), and modulate cellular HIV expression (7). Lower concentrations of micronutrients have been associated with increased mortality in HIV-infected adults (8–10).
Low concentrations of vitamins A and E have been reported in French children with AIDS (11). Micronutrient deficiencies have been found in North American HIV-exposed infants (12) and HIV-infected children (13), in HIV-infected Italian children (5), and in African children with AIDS (14). By contrast other studies have not detected micronutrient deficiencies in North American HIV-infected children (15, 16) nor an association between vitamin A levels and morbidity or mortality (16). Information regarding micronutrient status in HIV-infected children in Latin America is scarce. HIV-exposed, uninfected infants may also be susceptible to micronutrient deficiencies given the epidemiological and social characteristics of HIV-infected Latin American women (17).
The aim of this study was to describe the frequency, characteristics and correlates of low serum concentrations of retinol, β-carotene, and vitamin E in a cohort of HIV-infected children living in Brazil, Argentina and Mexico, ≥ 1 to < 4 years of age, and compare the findings to those seen in a comparison cohort of uninfected children of the same age born to mothers who were HIV-infected enrolled in the same cohort study. Our hypothesis was that, due to factors reviewed above, rates of low concentrations of these micronutrients would be higher in the HIV-infected group than those in the HIV-exposed, uninfected group. Our findings have important implications for the appropriate surveillance for and management of nutritional deficiencies in these settings.
The NICHD (Eunice Kennedy Shriver National Institute of Child Health and Human Development) International Site Development Initiative (NISDI) pediatric protocol is an ongoing prospective cohort study enrolling HIV-infected and HIV-exposed uninfected children at multiple clinical sites in Latin America. A description of this protocol and the cohort has been published (18). Briefly, a total of 1240 children were enrolled in the cohort from its inception in 2002 until May 31, 2005, the cut off point for the use of repository specimens for this analysis. When enrollment began in 2002, 2 groups of subjects who were receiving care at participating urban sites (11 in Brazil and 2 each in Mexico and Argentina) were eligible: 1) infants (≤ 12 months of age) who were born to women diagnosed with HIV infection either prior to or during pregnancy or within 1 month postpartum, and 2) HIV-infected infants, children, and adolescents (≤21 years of age). For this analysis children enrolled in the cohort study up to May 31, 2005, who had a repository plasma sample obtained between ≥1 and <4 years of age were eligible. The first available sample from each eligible child after the age of 12 months was analyzed. The protocol was approved by the ethical review boards of each clinical site enrolling subjects, by the sponsoring institution (NICHD), the data management and statistical center (Westat), and the Brazilian National Ethics Committee (CONEP). Informed consent was obtained from either parents or guardians.
The following data were collected in a standardized fashion and available at the time of repository plasma sample collection as part of twice a year study visits in the NISDI pediatric protocol: medical history (including HIV disease classification, demographic data including habits such as tobacco and alcohol use during pregnancy, pregnancy history, medication history), physical examination, laboratory evaluations (including hematology, flow cytometry, and standard biochemical assays), and growth parameters (including birth weight, gestational age, weight, height and head circumference). Laboratory results were classified as abnormal (Grade ≥ 1) according to the Division of AIDS toxicity table (19). The CDC SAS program was used for categorizing subject’s height and weight for age (20). CDC HIV disease classification (Clinical category C – severely symptomatic; B moderately symptomatic; A mildly symptomatic and N non symptomatic) (21) and virologic assessments were performed in the HIV-infected group.
History of substance use during pregnancy was obtained by maternal/care provider report, with yes/no answers. Micronutrient supplementation intake history was based on maternal/care provider report and/or medical record abstraction. Supplementation of retinol, β-carotene and vitamin E was defined as reported intake of these vitamins (alone or as part of a multivitamin) regardless of dose or duration of supplement use.
If any antiretroviral (ARV) was received during pregnancy, the mother was considered to have been exposed to ARVs. ARV regimens received by the HIV-infected children at the time the repository sample was obtained were recorded as part of the regular protocol visit.
Sites were instructed to separate and process plasma within 4 to 6 hours of blood collection but no longer than 30 hours after collection. Plasma samples were stored locally at −70°C and then shipped frozen in batches approximately twice a year to a central repository where they were also stored at −70°C until analysis. Retinol, β-carotene, and vitamin E (alpha-tocopherol) were quantified in plasma samples using high performance liquid chromatography (HPLC) following the gradient reversed-phase procedures, at a flow rate of 2 ml/min, described by Arnaud et al. (22), with appropriate controls, and were run in a single batch. Briefly, plasma was collected and deproteinized with ethanol and N-hexane. After centrifugation at 1000g for 10 min, N-hexane was removed and carefully dried under nitrogen. The sample was then resuspended in 0.5 mL of the mobile phase for plasma, and 100 μl was injected into the chromatograph. The HPLC mobile phase was acetonitrile, methanol and dichloromethane (70:10:20). A 25-cm × 0.46-cm Shim-pack CLC-ODS column and a 1-cm × 4-mm Shim-pack CLC-G-ODS precolumn (Shimadzu) were used. The chromatographic system consisted of a Rheodyne injector apparatus (model 7161, Rheodyne, L.P., Cotati, CA, USA), a Shimadzu LC-9A infusion bomb, and a Chromatopac Shimadzu C-R6A recorder. The reagents obtained from Merck (Darmstadt, Germany) were HPLC grade. A photodiode array detector was set to monitor the absorbance of these groups of compounds at 450, 292 and 325 nm for β-carotene, alpha-tocopherol and retinol, respectively. The peak areas were calibrated against known amounts of external standards. Low serum concentrations were defined as the following: retinol <0.70 μmol/L (23, 24), β-carotene <0.35 μmol/L (25), and vitamin E <18 μmol/L (26). Values below the limits of detection were expressed as 0.0 μmol/L.
Proportions of subjects with low micronutrient concentrations were calculated, and the exact binomial method was used to estimate 95% confidence intervals. Bivariate associations were evaluated using the chi square test and Fisher’s exact test (27) to examine correlates of low serum concentrations. Odds ratios and corresponding 95% confidence intervals were calculated. All variables that were marginally associated with low concentrations of micronutrients at alpha ≤0.2 were considered candidate variables for the logistic regression models. Both stepwise selection and backward elimination strategies were applied to determine whether both procedures arrived at the same parsimonious model (using a 5% significance level).
As of May 31, 2005, a total of 1240 children were participating in the cohort study. Six hundred seventy were ≥1 and < 4 years of age, and 336 of these 670 had at least one study visit and had at least one repository sample obtained during that period and thus comprised the study population for this analysis. One hundred twenty-four (37%) were HIV-1 infected and 212 (63%) were uninfected. As shown in the Table 1, the HIV-uninfected children were younger than the infected, and most (88.4%) of the subjects were enrolled in Brazil. There was no significant difference in the use of tobacco, alcohol, marijuana, or crack/cocaine by the mothers of these children during pregnancy. Frequency of birth weight < 2500g was similar in HIV-infected and HIV-uninfected children. Twenty-two percent of the HIV-infected group had height for age percentile < 5%, which differed significantly from 4.9% in the uninfected group. Most children had normal values of plasma albumin regardless of HIV status. However, anemia was a more common finding among the HIV-uninfected children and was reported in almost one-third of that group. A higher proportion of the HIV-uninfected children reported a history of micronutrient supplementation compared to the HIV-infected children.
Most of HIV-infected children (97%) acquired HIV by vertical transmission. 30.6% were severely symptomatic, 28.9% were moderately symptomatic and 40.5% were mildly or non symptomatic. Two-thirds (66.4%) of the HIV-infected children had HIV-RNA levels ≥ 1000 copies/ml, and 95.8% had CD4+ cell counts > 500 cells/mm3. Most (79%) were on ARVs.
Samples were drawn and stored a median (range) of 1.8 years (0.8–3.4) before they were analyzed. The median concentrations of the 3 micronutrients studied did not differ between the 2 groups based upon HIV status. Median (range) concentrations for the HIV-infected and uninfected groups respectively were 0.53 (0.17–2.12) μmol/L and 0.57 μmol/L (0.17–1.25) for retinol, 1.84 μmol/L (0.0–6.06) and 1.05 μmol/L (0.0–8.03) for β-carotene, and 12.42 μmol/L (3.34–44.75) and 11.38 μmol/L (3.66–27.65) for vitamin E. In addition, there were no significant differences in any micronutrient concentration based upon history of supplementation intake or country of origin (data not shown). Despite the lack of differences based upon HIV status and history of supplementation intake there were high rates of low micronutrient concentrations overall. As shown in the Figure, the majority of both HIV-infected and uninfected subjects had low retinol and vitamin E concentrations while over a quarter of both groups had low β-carotene concentrations.
We further analyzed the data to determine factors associated with low micronutrient concentrations among the 124 HIV-infected children. Maternal factors such as years of formal education or ARV or substance use during pregnancy were not associated with low levels of β-carotene or vitamin E (data not shown). However, children on any ARV regimen were less likely to have low retinol concentrations (OR=0.16, 95% CI: 0.04 – 0.74) than children who were not on ARVs, regardless of supplementation. This association was observed for regimens consisting of nucleoside reverse transcriptase inhibitors (NRTI) alone (OR=0.17, 95% CI: 0.03 – 0.95), NRTI plus protease inhibitors (PI) regimens (OR=0.18, 95% CI: 0.04 – 0.85) and NRTI plus non-nucleoside reverse transcriptase inhibitors (NNRTI) regimens (OR=0.11, 95% CI: 0.02 – 0.62). No other variables including CD4 count, CDC-HIV classification, and HIV viral load met inclusion criteria for multivariable modeling.
Our findings show that low concentrations of retinol, β-carotene and vitamin E are very common in children living in Brazil, Argentina and Mexico exposed to HIV, regardless of HIV-infection status thus disproving our original hypothesis. The rates of low concentrations of retinol and vitamin E were very high (74% and 89%, respectively).
The lack of an effect of HIV status on rates of micronutrient low serum levels was surprising. However, this study does not provide final answers to this question and raises several additional ones. Because we did not examine a group without exposure to HIV, the rates of low serum concentrations we observed could be due to factors related to in utero HIV exposure, including possible genetic defects in vitamin transfer proteins, fat metabolism disorders and oxidative stress which could alter vitamins requirements (28). Cunningham-Rundles et al. (12) reported that 70% of North American infants perinatally exposed to HIV were retinol deficient in the first months of life whereas infants with various other disorders had normal retinol levels. Another hypothesis is that in utero or postnatal exposure to ARVs results in the low micronutrient concentrations we observed. Evidence against this though includes our finding that HIV-infected children treated with ARVs had a lower risk of retinol deficiency; this effect was not seen with β-carotene or vitamin E so may represent a chance observation. Nevertheless, a study in HIV-infected adults on highly active ARV therapy showed low rates of retinol and vitamin E deficiencies (29).
A more likely explanation for the high rates of low serum levels regardless of HIV status are maternal factors such as socioeconomic and health status. Published reports and experience demonstrate that the HIV epidemic in Latin America is characterized, as in many places, by poverty and social marginalization (17, 30). Although we did not formally examine these factors, approximately half of the mothers of children from both groups had a limited number of years of formal education, a condition associated with poverty in these countries. In addition, a high rate of anemia was found in both groups, especially among the HIV-exposed/uninfected, which may reflect overall insufficient nutritional intake (31).
For the individual micronutrients we examined, the overall rates of low serum concentrations were in the range of those previously reported in pediatric populations without HIV exposure or infection. Although definitions of retinol deficiency vary across studies, reports of retinol deficiency in pre-school children show rates up to 74.5% in Brazil (32), 30% in Argentina (33), and 46.3% in Mexico (34). These studies support the idea of poor intake as the main cause for the high prevalence of retinol deficiency. Similarly, while definitions of vitamin E deficiency vary across studies, ranging from 7 to 28 μmol/l without adjustment for serum cholesterol (26), reports of vitamin E deficiency show rates as high as 69% in apparently healthy Latino children in the United States (35) and 70% in Mexican preschoolers (36). In addition, vitamin E deficiency was found in 70% of HIV- infected and uninfected South African women and no significant difference by HIV status was observed (37). The rate of β-carotene deficiency we observed was considerably lower than the rate of retinol deficiency. Although guidelines regarding the serum values for carotenoid adequacy have not been validated, definitions of β-carotene deficiency have been used, ranging from 0.35 to 4.5 μmol/l (25, 38, 39). The explanation for this relative preservation of β-carotene in the face of retinol deficiency is not known. Possible explanations, none of which we were able to assess in this study, include a plant-based diet, which has been shown to have a significant impact on β-carotene serum levels but no significant effect on retinol levels (40, 41), the presence of another deficiency, such as zinc, that negatively affects retinol mobilization from the liver (42) or the negative impact of the acute phase response on retinol levels as has been shown in children with malaria (43), and in HIV-1 seropositive subjects (44).
The impact of the inflammatory response related to HIV and its impact on micronutrient concentrations is unclear. Jahoor et al. have shown that AIDS elicits an acute phase response that is different from bacterial infections, as the higher concentrations and faster rates of synthesis of the positive acute phase proteins are not accompanied by lower concentrations and slower rates of synthesis of most of the negative acute phase proteins, including retinol binding protein (45). In addition, clinically stable HIV – infected children in South Africa were evaluated for vitamin A and vitamin E and the results failed to demonstrate a correlation between these micronutrients and C-reactive protein concentrations (14). We were unable to address the potential association between low retinol and inflammation since we did not include measurements of inflammatory markers, such as C-reactive protein. However, we did not find an association between HIV disease parameters, such as clinical status, viral load, and CD4 count, and micronutrient concentrations, suggesting that inflammation related to HIV disease itself did not play a role in our findings. However, other causes of inflammation, common to both groups, may well have played a role that we were not able to assess.
The mean age of the HIV-uninfected children corresponds to infants and younger children whose nutritional status might still be reflecting maternal nutritional status. It is known that pregnancy increases the risk of vitamin deficiency and this extends to the newborn. The prenatal dietary intake of vitamin A is frequently considered to be insufficient to meet increased requirements during pregnancy, particularly in developing countries (46). In a Brazilian study, maternal vitamin A deficiency was strongly associated with infant vitamin A deficiency and low birth weight (47). We did not collect maternal dietary intake or maternal plasma levels of vitamin A and vitamin E so were unable to assess their association with the plasma levels of their infants.
Other limitations of this study include the possibility of incomplete recording of nutritional supplements and no dietary intake history, which could explain, at least in part, the lack of significant associations between history of supplement intake and rates of micronutrient low serum levels. Since this study examined stored samples, processing and storage factors may have affected the results. However, retinol, β-carotene, and vitamin E appear to be stable for at least 15 years at −80°C (48) and repeated freezing and thawing, which was not the case in our study, do not appear to affect the concentrations of these vitamins in serum (49).
In summary, our findings demonstrate that low concentrations of vitamin A and vitamin E are extremely common among HIV-infected and HIV-exposed/uninfected children living in Brazil, Argentina and Mexico and similar to rates in other pediatric populations in comparable settings. Poor nutritional intake is likely the main factor contributing to the low serum values of micronutrients in these children. Factors associated with HIV disease itself did not appear to play a role, but other inflammatory processes common to both the HIV-exposed/uninfected and the HIV-infected groups were not assessed.
Source of Financial Support: NICHD Contracts #N01-HD-3-3345 and #N01-DK-8-0001.
Each author has participated sufficiently in the work to take public responsibility for the content of the paper and approved the final version of the manuscript. We declare that there is no conflict of interest among authors and coauthors. We thank the first principal investigator of the NISDI pediatric study, Leslie Serchuck. We also thank Jennifer Graf for her thoughtful review of the manuscript
Principal investigators, co-principal investigators, study coordinators, data management center representatives, and NICHD staff include: Brazil: Belo Horizonte: Jorge Pinto, Flávia Faleiro (Universidade Federal de Minas Gerais); Caxias do Sul: Ricardo da Silva de Souza, Nicole Golin, Sílvia Mariani Costamilan (Universidade de Caxias do Sul/Serviço Municipal de Infectologia); Nova Iguacu: Jose Pilotto, Beatriz Grinsztejn, Valdilea Veloso (Hospital Geral Nova de Iguacu – HIV Family Care Clinic); Porto Alegre: Ricardo da Silva de Souza, Breno Riegel Santos, Rita de Cassia Alves Lira (Universidade de Caxias do Sul/Hospital Conceição); Ricardo da Silva de Souza, Mario Peixoto, Elizabete Teles (Universidade de Caxias do Sul/Hospital Fêmina); Ricardo da Silva de Souza, Marcelo Goldani, Margery Bohrer Zanetello (Universidade de Caxias do Sul/Hospital de Clínicas de Porto Alegre); Regis Kreitchmann, Debora Fernandes Coelho (Irmandade da Santa Casa de Misericordia de Porto Alegre); Ribeirão Preto: Marisa M. Mussi-Pinhata, Maria Célia Cervi, Márcia L. Isaac, Bento V. Moura Negrini (Hospital das Clínicas da Faculdade de Medicina de Ribeirão Preto da Universidade de São Paulo); Rio de Janeiro: Ricardo Hugo S. Oliveira, Maria C. Chermont Sapia (Instituto de Puericultura e Pediatria Martagão Gesteira); Esau Custodio Joao, Maria Leticia Cruz, Claudete Araujo Cardoso, Guilherme Amaral Calvet (Hospital dos Servidores do Estado); São Paulo: Regina Celia de Menezes Succi, Daisy Maria Machado (Federal University of São Paulo); Marinella Della Negra, Wladimir Queiroz (Instituto de Infectologia Emilio Ribas); Mexico: Mexico City: Noris Pavía-Ruz, Patricia Villalobos-Acosta, Elsy Plascencia-Gómez (Hospital Infantil de México Federico Gómez); Peru: Lima: Jorge Alarcón (Instituto de Medicina Tropical “Daniel Alcides Carrión”-Sección de Epidemiologia, UNMSM), Maria Castillo Díaz (Instituto Nacional de Salud del Niño), Mary Felissa Reyes Vega (Instituto de Medicina Tropical “Daniel Alcides Carrión” - Sección de Epidemiologia, UNMSM); Data Management and Statistical Center: Yolanda Bertucci, Laura Freimanis Hance, René Gonin, D. Robert Harris, Roslyn Hennessey, Margot Krauss, James Korelitz, Sharon Sothern, Sonia K. Stoszek (Westat, Rockville, MD, USA); NICHD: Rohan Hazra, Lynne Mofenson, Jennifer Read, George Siberry, Carol Worrell (Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda, Maryland).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.