| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Strategies to reduce the burden of malaria morbidity and mortality in young children include early identification and appropriate treatment of febrile episodes, effective treatment of uncomplicated and severe malaria cases in health facilities, increased use of insecticide impregnated bednets, and intermittent preventive treatment.5 Although children who live in malaria-endemic areas are often exposed to nutritional deficiencies, and the contribution of such deficiencies to the susceptibility to malaria has been suspected for several decades,6 limited evidence from randomized trials is available to propose the inclusion of nutritional interventions as potentially useful strategies to decrease childhood malaria. Recent studies indicate that supplementation of preschool children with vitamin A7 or zinc8 could reduce the incidence of clinical malaria by P. falciparum, whereas supplementation with iron to iron-sufficient children from highly endemic areas might increase severe illness and mortality.9 Evidence on the role of other nutrients is, however, lacking. Deficiencies of thiamin,10 folate,11 and antioxidants including vitamin E12 have been noted in patients with acute malaria, but there is a paucity of prophylactic or therapeutic intervention studies that could support a causal link between the nutrient status of children and their risk of malaria.
Using data collected in a randomized controlled trial of vitamin supplementation to HIV-infected pregnant and lactating women in Tanzania, we examined the effect of multi-vitamins (B, E, and C) and vitamin A plus ß-carotene on the incidence of malaria in the women’s babies during the first 2 years of age.
|
METHODS |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
At recruitment, research nurses obtained information on age, education, socioeconomic and marital status, obstetric history, and anthropometry. Study physicians performed a complete medical examination and collected blood specimens that were later used for assessment of malaria parasites in peripheral blood and other tests.
Follow-up was carried out monthly at the Muhimbili National Hospital antenatal clinic, where all pre- and postnatal visits, childbirth, and study procedures took place. We provided information on the risks and benefits of infant feeding options among HIV-infected women, following the World Health Organization (WHO) and the Tanzanian Ministry of Health guidelines; almost all women chose to breastfeed. At each monthly postnatal visit, study physicians carried out complete physical examinations of the mothers and children. Trained research nurses inquired about the health and breast-feeding status during the preceding period, assessed compliance with the study regimen, took the children’s axillary temperature, and obtained anthropometric measurements. At 6 months of age and every 6 months thereafter, all children received an oral dose of vitamin A (200,000 IU; 100,000 IU for age < 1 year) as per the standard of pediatric care in Tanzania, irrespective of treatment assignment. Antiretroviral medications were unavailable in this setting at the time of the study.
The children’s survival status was assessed during clinic visits or through home visits when the mothers and children did not attend the clinic. If the mother was not at home, her neighbors or relatives were asked about vital status of the child. To approximate the cause of death of deceased children, we completed verbal autopsy questionnaires by interviewing the primary care provider, reviewing medical records, or both. The cause of death was ascertained in a blinded fashion.
HIV status of the children was assessed from blood samples collected at birth, 6 weeks, and every 3 months thereafter. We defined HIV infection as a positive PCR test at any age, or a positive ELISA, confirmed by Western blot, in children 
18 months of age. The time of transmission was estimated as the midpoint between the last negative and the first positive samples.
Blood films were obtained from the children every 3 months. Women were encouraged to attend the study clinic at any time when they or their children had febrile episodes or felt ill otherwise. Blood films were also requested during such interim clinic visits at the physicians’ discretion. The presence of malaria parasites in peripheral blood was assessed by thick and thin smears stained with Giemsa. One trained laboratory technician carried out readings of each blood smear in three different fields of the slide. Similarly to all other study personnel, investigators, and patients, the microscopist was blinded to the study regimens. Parasite density per cubic millimeter was estimated from the number of parasites per 200 leukocytes and a leukocyte count of 8000/µL.15 Virtually all infections were by P. falciparum.
Supplementation of the mothers with the experimental regimens continued throughout the 2 years post-delivery that we consider in this study. Compliance with the study regimen by 2 years, defined as the number of tablets absent from the returned bottles at monthly visits divided by the total number of tablets the individual should have taken, was 83% on average.
Data analysis.
Of 1,078 women randomized, 3 were found to be non-pregnant, 6 died before delivery, and the date of delivery or outcome of pregnancy was unknown in 42 (Figure 1
). We included 829 live-born, singleton children who had at least one assessment of malaria during their first 2 years of age. Baseline characteristics or treatment assignment in this subset did not differ from the original group of randomized women.
|
1/µL), 2) high parasitemia ( 5,000/µL), 3) clinical malaria (parasitemia 5,000/µL with an axillary temperature 37.5°C, in accordance with previous studies,16,17) and 4) death from malaria, when malaria was assigned as the primary cause of death in the verbal autopsy. Malaria events that occurred within 20 days of each other were counted as a single outcome.
To verify the randomization assumption in the group included for malaria analyses, we compared the distribution of baseline characteristics across treatment groups using Kruskal-Wallis tests for continuous variables and 
2 tests for proportions. We followed the intent-to-treat principle to examine the effect of maternal vitamin supplementation on the incidence of child malaria endpoints. Incidence rates were estimated as the number of observed events over the person-months that children contributed to follow-up in each treatment arm. For parasitemia and clinical malaria endpoints, incidence rate ratios and 95% confidence intervals (CI) were estimated from Poisson regression models for correlated data within child, with the log link function, by comparing the rates in each of the three active treatment groups against those in the placebo arm. For death from malaria, we obtained hazard ratios with 95% CI from a Cox proportional-hazards model with time-to-death as the outcome and indicator variables for treatment assignment. The main effects of multivitamins were examined by comparing children whose mothers received multivitamins (alone or in combination with vitamin A/ß-carotene) versus women who did not receive multivitamins (those in the vitamin A/ß-carotene-alone and placebo arms). An analogous approach was followed to assess the main effects of vitamin A/ß-carotene. We examined interactions between multivitamins and vitamin A/ß-carotene by testing the significance a cross-product term (Wald test) in the models. We also explored whether the effect of supplements on malaria differed between children who became HIV-infected and those who remained uninfected during follow-up, by introducing cross-product terms between regimen assignment and child HIV status as a time-varying covariate in the models.
Ethical considerations. All participants gave informed consent to be included in the study. The institutional review boards of Muhimbili University College of Health Sciences, the Tanzanian National AIDS Control Program, and Harvard School of Public Health approved the study protocol.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Babies born to mothers who were assigned to multivitamins had significantly greater birth weight and a lower incidence of intrauterine growth retardation, compared with women who did not receive multivitamins, as we have previously reported.18 During the first 2 years of life comprising the study duration, children contributed 11,933 child-months to the follow-up of malaria endpoints, with a median 17.1 months per child (mean = 14.4, SD = 8.8), irrespective of treatment assignment. One hundred sixty-eight of the 829 children died at or before 24 months of age. Breastfeeding was adopted universally (> 99%); the proportion of breastfeeding children by 12, 18, and 24 months was, respectively, 94%, 74%, and 22%. Information on HIV status was available for 818; among them, 29.6% became HIV-infected during the first 2 years of life. HIV transmission occurred more frequently among children whose mothers received vitamin A/ß-carotene (33.5%) compared with those who did not receive vitamin A/ß-carotene (25.6%) (P = 0.01), consistent with our previous report.19 Multivitamins were not associated with transmission.
|
1/µL; however, multivitamins (B, C, and E) resulted in a nonstatistically significant 30% reduction of high parasitemia (95% CI = –2%, 52%; P = 0.06) (Table 2). This reduction appeared to be larger when multivitamins were administered together with vitamin A/ß-carotene (43% reduction, P = 0.04), although there was not a statistically significant interaction between the two regimens on the high parasitemia outcome (P for interaction = 0.97). Maternal supplementation with multivitamins alone resulted in a statistically significant reduction of 71% in the risk of clinical malaria (P = 0.02), whereas vitamin A/ß-carotene alone was related to a nonstatistically significant reduction in clinical malaria by 63% (P = 0.06). There was a significant interaction between the two regimens on clinical malaria, by which their joint administration reduced the protective effect of each one individually (P for interaction = 0.05). Both regimens reduced the incidence of malaria-related deaths, but the effects were not statistically significant. The children’s HIV status did not significantly modify the effect of vitamins on the malaria outcomes considered. The results did not change after adjustment for potential clustering by recruitment clinic.
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Research on the potential efficacy of vitamin supplements to prevent child malaria has focused primarily on vitamin A. In a randomized clinical trial conducted in Papua New Guinea, the three-monthly administration of 200,000 IU pre-formed vitamin A to children of age 6–60 months significantly decreased the incidence of fever with high parasitemia over a 13-month follow-up period.7 Whether supplementation with other vitamins could also be beneficial is less clear. Results from early animal models had suggested that riboflavin (vitamin B2), pyridoxine (vitamin B6), and vitamin E could exacerbate experimental malaria and that deficiency of those nutrients could in some cases be protective.6 One study of 5-to 14-year-old children from The Gambia reported that a supplement consisting of iron, thiamin, riboflavin, and vitamin C administered twice a week during a 3-month period had no effect on clinical malaria, compared with placebo.20 In a study of 207 children of age 5 to 36 months from Bagamoyo, Tanzania, supplementation 3 times/week for 5 months with a preparation of vitamins A, B, C, D, and E plus iron increased average hemoglobin concentrations but did not affect the incidence of malaria.21 It is not possible to assess the impact of the vitamins alone in these trials because of the inclusion of iron. More recent evidence from observational studies in humans lends some support to the potential beneficial role that vitamins other than A could play against malaria. Thiamin deficiency was positively associated with uncomplicated and severe malaria in adults from Thailand.10 Among Ugandan children acutely ill with uncomplicated malaria, the serum concentration of vitamin E and other antioxidants was severely depressed12; these concentrations increased significantly as parasitemia cleared. An early study of children from Niger showed that those with cerebral malaria had lower concentrations of folate compared with children with uncomplicated malaria.11 Conclusions from these observational studies, however, are limited by the possibility of reverse causality because malaria infection could be the cause of acute nutrient deficiencies rather than the consequence and by confounding due to common causes of both vitamin deficiencies and malaria.
It is not possible to determine to what extent the apparent benefits of multivitamins (B, C, and E) could be attributed to the actions of individual nutrients in our study; the potential mechanisms that mediate the vitamins’ effects on clinical P. falciparum malaria are speculative. We have previously reported that the administration of supplements to these lactating mothers significantly increased the concentration of nutrients in the children22; therefore, the benefits could be mediated through improvements in the vitamin status of the children and subsequent enhancements of specific aspects of immunity. We have shown that multivitamins significantly increased the number of circulating CD4 cells in these children,23 and CD4 cells play a fundamental role in the immune response to malaria, as helpers in the humoral response of B cells and by regulating the innate parasitocidal activity and the level of the inflammatory response.24 In addition to the immunomodulatory potential of vitamins, some in vitro studies suggest that high-dose riboflavin25 and vitamin C26 could exhibit a direct parasitocidal effect.
Vitamin A and ß-carotene supplementation decreased the risk of clinical malaria in this study; although the association did not reach statistical significance, it was in the same direction as that reported by Shankar and others in Papua New Guinea.7 The effect of vitamin A on clinical malaria could be mediated in part by increased phagocytosis of infected red cells and decreased proinflammatory activity.27 Given that vitamin A and ß-carotene increased the risk of mother-to-child transmission of HIV in this cohort,19 we cannot advocate daily supplementation with these nutrients to HIV-infected lactating women as a strategy to prevent malaria in children. Instead, direct supplementation of vitamin A to children older than 6 months of age is likely to be a safer public health intervention.
The nature of the negative interaction by which the joint administration of multivitamins and vitamin A/ß-carotene seemed to decrease the effects of each individual treatment on clinical malaria is uncertain. Biochemical nutrient-to-nutrient interactions could be an explanation, because some of the nutrients present in one arm have been reported to decrease the bioavailability of others. We previously found that multivitamins (B, C, and E) decreased the concentrations of retinol (vitamin A) in the infants at 6 weeks of age.22 This effect is likely attributable to the 
-tocopherol (vitamin E) component of the multivitamins, as suggested by a study of well-nourished men in whom serum retinol concentrations were significantly decreased after 16 weeks of daily high-dose -tocopherol supplementation.28
Our study has a few limitations. Malaria surveillance was hospital-based and scheduled at relatively wide time intervals; in addition, diagnosis of clinical malaria depended in part on the guardian’s attendance to the clinic during episodes of illness. These factors could have resulted in an underestimation of the incidence of clinical malaria. If this underestimation was differential by treatment arm, the estimates of effect could be biased in an unpredictable direction. However, the randomized design makes it unlikely that such underestimation was differential by treatment arm. Each blood film was examined by a single microscopist. This could have introduced some random variation in the diagnoses of malaria parasitemia; however, given that the microscopists were blinded to the study regimen, and considering the randomized nature of the trial, we do not expect that this factor was a significant source of bias for the treatment effects. The vitamins could have changed the relation between parasite density and fever (pyrogenic) threshold, which might in part explain the observed results. We had limited statistical power to formally examine this possibility; however, Shankar and colleagues did not find that differences in pyrogenic threshold could account for the effect of vitamin A in their study of children from Papua New Guinea.7 We have previously reported that multivitamins improved the health status of these HIV-infected mothers13 and cannot rule out that the observed effect of these supplements on child malaria could be caused by provision of better infant care by less ill women. Despite these limitations, our results suggest a protective potential for maternal multivitamin supplements on malaria in children. Future investigations need to address whether the direct administration of multivitamins to infants offers similar benefits against malaria and whether the results could also be generalized to children born to HIV-uninfected women. It is also relevant to assess the effect of periodic multivitamin supplementation to children on malaria susceptibility after 2 years of age.
Received December 11, 2006. Accepted for publication March 13, 2007.
Acknowledgments: We are grateful to the women and children who participated in the study. The authors thank the field teams—including nurses, physicians, midwives, supervisors, lab staff, and the administrative staff—who made the study possible. The authors also thank the authorities at Muhimbili University College of Health Sciences, Muhimbili National Hospital, the City of Dar es Salaam Regional Health Authority, and the Tanzanian National AIDS Control Program for their institutional support.
Financial support: This study was supported by the National Institute of Child Health and Human Development (NICHD R01 32257).
* Address correspondence to Eduardo Villamor, Department of Nutrition, Harvard School of Public Health, 665 Huntington Ave., Boston, MA 02115. E-mail: evillamo{at}hsph.harvard.edu 
Authors’ addresses: Eduardo Villamor and Wafaie W. Fawzi, Department of Nutrition, Harvard School of Public Health, 665 Huntington Ave., Boston, MA 02115, Telephone: +1 (617) 432-1238, Fax: +1 (617) 432-2435, E-mails: evillamo{at}hsph.harvard.edu and mina{at}hsph.harvard.edu. Gernard Msamanga, Department of Community Health, Muhimbili University College of Health Sciences, Dar es Salaam, Tanzania, E-mail: gmsamanga{at}muchs.ac.tz. Elmar Saathoff, Department of Tropical Medicine, University of Munich, Germany, E-mail: saathoff{at}lrz.uni-muenchen.de. Maulidi Fataki and Karim Manji, Department of Pediatrics and Child Health, Muhimbili University College of Health Sciences, Dar es Salaam, Tanzania, E-mail: mfataki{at}muchs.ac.tz, kmanji{at}muchs.ac.tz.
Reprint requests: Eduardo Villamor, Department of Nutrition, Harvard School of Public Health, 665 Huntington Ave., Boston, MA 02115, Telephone: +1 (617) 432-1238, Fax: +1 (617) 432-2435, E-mail: evillamo{at}hsph.harvard.edu.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Dinerman Malaria in Children Following International Travel AAP Grand Rounds, July 1, 2009; 22(1): 8 - 8. [Full Text] [PDF]
|
|
|
|
|
|
M. Dinerman Effects of Maternal Vitamin Supplements on Malaria in Children of HIV-Infected Women AAP Grand Rounds, October 1, 2007; 18(4): 43 - 44. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
