| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The recommendation of universal iron supplementation of children in areas in which iron-deficiency anemia is common16 is currently being challenged by two parallel studies performed in Pemba, Tanzania and Nepal.8,17 The supplementation arms of the study that contained iron were stopped early because of an increased risk of experiencing an adverse event or hospitalization in the iron supplemented groups in Pemba, an area in which Plasmodium falciparum malaria is endemic. No effect of iron on morbidity or mortality was noted in the Nepalese study, an area without significant risk of malaria, indicating that the differing results may be a result of local malaria endemicity.
This report describes a community-based, randomized, double-blind, placebo-controlled trial in which children received supplements containing iron, zinc, iron plus zinc, or placebo. The primary hypotheses were that daily zinc and combination of iron plus zinc supplementation would have a protective effect against malaria compared with the placebo, and that daily iron supplementation would result in higher hemoglobin concentrations than the placebo, zinc, and combination iron plus zinc groups. Because iron and zinc supplementation may have additional health effects, morbidity from diarrhea and acute lower respiratory infections (ALRIs) was also examined, as were anthropometric measurements at the beginning and end of the trial.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The study enrolled 855 children 0.5–15 years of age who did not have a chronic illness (absence of congenital diseases or major illness requiring medical care and/or medication determined by the physician at baseline evaluation) or severe malnutrition (weight for height z-score less than two standard deviations below the National Center for Health Statistics [Hyattsville, MD] reference population20 or clinical signs of marasmus or kwashiorkor) after an initial interview during which parental written consent was obtained. The age range was intentionally broad because the Iquitos situation is one of an emerging or re-emerging disease and many of the older children had not acquired protective immunity. Given that iron and zinc supplementation may be advocated for areas with varying levels of malaria transmission, it is important to examine the effects in this less than hyperendemic transmission zone.
Randomization and masking. After the baseline evaluation, all records were sent to Lima to the Insituto de Investigación Nutricional (IIN) for data entry and randomization. Using an algorithm in SAS version 6 (SAS Institute, Cary, NC), we randomized children meeting the entry criteria in blocks of four into four supplement groups with stratification on age (three groups: < 5, 5–9, and > 9 years of age) and presence of anemia. A fixed dose of 15 mg iron (as iron sulfate), 15 mg of iron plus 20 mg of zinc (as zinc sulfate), 20 mg of zinc, or placebo was used for efficiency in implementation; therefore, young children received a higher dose per kilogram of weight. The zinc and iron dose and schedule used in this protocol were 1–2 times the recommended dietary allowances in 1997 when this trial was designed because of a higher prevalence of deficiency resulting from diarrhea morbidity and low bioavailability of these nutrients in the diet. The doses were considered to be well within the safety range for preschool and school age children.20–24
The supplements were similar in appearance and taste, bottled in similar containers, and labeled with a supplement code. Each dose was 2.5 mL of a red strawberry-flavored syrup made in Lima, Peru by a licensed Peruvian pharmaceutical laboratory (Instituto Quimioterápico SA, Lima, Peru). Supplements were administered using a spoon. The participants and study personnel were both masked throughout the study to the supplement contents. The data analyst was masked for the seminal analyses and was unmasked for additional analyses and sub-analyses. No data were excluded or altered after unmasking.
Intervention. After the baseline cross-sectional survey and randomization to a supplement group, daily supplementation was carried out for seven months from February 7 to September 8, 1998. Study personnel observed supplement administration at daily (Monday through Saturday) home or school-based visits, and supplementation on Sundays was unsupervised. The study was reviewed and approved by The Joint Committee on Clinical Investigations of the Johns Hopkins University School of Medicine and the Ethical Committee of the IIN.
Measurement of outcomes.
Baseline weight, height, hemoglobin status, zinc status, and other characteristics were measured at the time of enrollment and at the end of the study (seven months) at the project’s outpatient health center. Hemoglobin was measured immediately after blood was drawn using a portable B-hemoglobin photometer (HemoCue AB, Angelholm, Sweden). As a quality control measure for the field hemoglobin measurement, two standardized microcuvettes provided by the manufacturer were tested at approximately 20-sample intervals. Thick and thin blood smears were also prepared from venous blood at the time of drawing. The venous samples were transported in cold boxes to the Hospital Regional of Iquitos where the plasma was separated and stored at –20°C. The samples were then transported to the IIN for plasma ferritin and zinc analyses. Plasma zinc was analyzed by atomic absorption spectrophotometry at the IIN, and ferritin was determined in duplicate by an enzyme-linked immunoadsorbent assay kit (Ramco Laboratories, Stafford, TX). Standard quality control measures were taken with the laboratory measurements (e.g., for zinc, a sample of known concentration was included every 12 samples). For blood slides, species-specific parasite density per 200 white blood cells was determined on the thick and thin blood smears, and the slide was considered negative if no parasites were observed in 100 thick film fields. The slides were read by an experienced laboratory technician who had been tested and standardized prior to the survey. Stunting was defined as height for age z-score more than two standard deviations below the reference population,20 and wasting was defined as weight for height z-score more than two standard deviations below the reference population. Zinc deficiency was defined as plasma zinc concentration less than 9.9 µmol/L.25 Although using plasma zinc concentration is not an ideal individual level indicator of zinc status because of homeostasis and the impact of infection on serum zinc, this indicator has been demonstrated to be a good population level indicator of compliance with zinc supplementation.9,26 Anemia was defined as a hemoglobin concentration less than 110 g/L in children less than five years of age, less than 115 g/L in children 5–11 years oaf age, and less than 120 g/L in children more than 11 years of age.16 Iron deficiency was defined as a plasma ferritin concentration less than 12 µg/dL in children less than five years of age and less than 15 µg/dL in children 
5 years of age.27 Compliance with supplementation was calculated as the number of days on which supplementation was taken divided by number of days in the trial x 100 to obtain a percentage.
Twice a week, the study personnel gathered information on morbidities, including the consistency and number of stools, presence of fever, vomiting, coughing, and a variety of other symptoms. Three main morbidity outcomes were considered: malaria, diarrhea, and ALRI. Malaria was defined as the presence of malaria parasites in blood and fever within the past 72 hours. Children who were diagnosed with malaria were provided with free treatment through the Ministry of Health’s malaria program. To avoid the inclusion of recrudescent malaria in defining new episodes, the child must not have been diagnosed with malaria in the previous 28 days to be considered a new episode. In addition, 20 episodes of malaria in which P. vivax and P. falciparum were both diagnosed were excluded from the analysis because interaction between these infections could limit interpretation of a malaria parasite-specific effect. Diarrhea was defined as more than three loose stools or one bloody stool within the past 24 hours, and episodes were separated by at least two diarrhea-free days. An ALRI was defined by the presence of cough and rapid respiration (both reported by the parent), as well as no positive test result for malaria in the previous 28 days, and episodes were separated by at least seven ALRI-free days.
Data management and analysis. All forms were checked by the field supervisor for consistency and completeness. After the data were entered, range and consistency checks were performed and study personnel obtained additional information and verified questionable values. The primary goals of analysis were to identify differences between treatment groups with regards to malaria, diarrhea, and ALRI morbidity, side effects experienced, and biochemical indices. Of particular interest were differences in these outcomes between single and combination nutrient supplementation. Analysis was based on intention to treat.
Initially, the data were analyzed as a two-factor analysis with interaction, and when the interaction was found to be significant (P < 0.15), an analysis of the four supplement subgroups was performed. This approach was also applied to sub-group analysis based on child age and sex because the literature supports the possibility of age and sex-specific effects, and because children were randomized within age strata. Children from 1 to 14 years of age are not expected to experience the same degree of morbidity, or to respond similarly to supplementation. In addition, there are substantial age-related differences in susceptibility to infections and in immunologic maturation. Thus, differences in the impact of nutritional supplements might be expected to vary with age and age subgroups were analyzed separately. The age subgroups were selected based on expected similarity in morbidity experience while taking sample size into consideration. Furthermore, because target beneficiaries of current programming in iron and zinc interventions internationally are children less than five years of age, it would be important to consider this age group explicitly.
Chi-square tests were used to compare categorical variables, and analysis of variance tests were used to compare continuous variables among the four supplement groups. If the overall test result was significant (P < 0.05), pairwise comparisons were performed, correcting for multiple comparisons using the Bonferroni correction. Statistical analyses were performed using Stata version 7.0 (Stata Corporation, College Station, TX). Z-scores were calculated using Epi-Info version 3.2.2 (Centers for Disease Control and Prevention, Atlanta, GA). A Cox proportional hazards model was used to analyze time to first P. vivax malaria episode, and Nelson-Aalen cumulative hazard estimates were used to display the risk in the four different supplement groups over time. For variables not normally distributed (e.g., hemoglobin concentration, plasma zinc concentration), the original values are presented, but statistical tests were performed on log-transformed values.
Comparison of morbidity outcomes among supplement groups, taking into account the time contributed to the denominator and correlation between measurements, was done using generalized estimating equations (GEEs).28 This method allows more robust estimation of the variance because it does not assume independence of events. The time contributed by the children was divided into 14-day child periods, and for the child-period to be included in the analysis, the child must have contributed at least seven days (50% of the child period) of morbidity information to that period. The GEE model for incidence used the occurrence of a new episode within the child period as the dependent variable and used binomial family, log link, and exchangeable correlation.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Fifteen children were interviewed but not enrolled in the trial, 11 of whom refused the blood draw, 2 parents refused consent to enter the trial, and 2 had congenital diseases and therefore were not eligible for entry into the trial. A total of 855 children were randomized to receive one of the four treatments, but 19 children had no days of follow-up and therefore did not contribute data to the longitudinal analysis. Overall, 836 (97.8%) children were included in the morbidity analysis, and the 748 (87.5%) children who had complete follow-up were also included in the comparison of anthropometric and biochemical indices. There were no deaths during the study. The numbers of child periods and days of follow-up in the study, as well as the number of morbidity episodes detected by supplement type, are shown in Table 1. Baseline hemoglobin concentrations were measured in all 855 children who began the study, and follow-up hemoglobin concentrations were measured in the 748 children who completed the study. Baseline and final plasma ferritin concentrations were measured in a random sample of 387 (45.2%) children, and baseline and final zinc concentrations were measured in a random sample of 395 (46.2%) study participants. Compliance with supplementation ranged from 88.7% to 89.8%, and did not vary by type of supplement used (P = 0.90).
|
|
), both among the children with complete follow-up and those who were lost to follow up. Of the 855 children randomized to a treatment group, 53.0% were classified as stunted, 46.4% of the children were anemic, and 46.0% of the children for whom serum zinc status was collected were zinc deficient at baseline. Only 10.4% of the children were found to be iron deficient as defined by low concentrations of plasma ferritin.
|
. When examining final hemoglobin concentrations, two-factor analysis of variance did not show significant interaction between iron and zinc supplementation (P = 0.84). At the final assessment, the mean hemoglobin concentration in the two groups that received iron (alone and combined with zinc) (119.1 g/L) was significantly higher than in those who did not receive iron (116.8 g/L; P = 0.004), whereas the main effect of zinc was not statistically significant (117.8 g/L versus 117.9 g/L; P = 0.94).
|
Malaria.
There were 268 malaria episodes over the course of the study. Both P. vivax and P. falciparum are present in the population, and the incidence data for the two types of malaria were analyzed separately. Plasmodium vivax predominated in this trial with 208 episodes (Table 1
). Initial results from GEEs indicated opposite effects of supplemental iron and zinc on risk of malaria (risk ratio for iron [alone or with zinc] = 1.24, P = 0.14 and for zinc [alone or with iron] = 0.68, P = 0.008), and the interaction by supplement type was significant (P = 0.12). Therefore, the supplement groups were analyzed separately. When we used a Cox proportional hazards model to compare the time to the first P. vivax malaria episode among supplement groups, the iron group experienced greater risk (hazard ratio [HR] = 1.46, 95% confidence interval [CI] = 0.99, 2.17, P = 0.056) of malaria morbidity compared with placebo, although this was not statistically significant (Figure 2). Considering multiple malaria episodes and correcting for correlation within children using GEEs resulted in similar findings that were statistically significant (Table 2). Among children in all age groups, iron supplementation increased the risk of a P. vivax malaria episode (incidence rate ratio [IRR] = 1.49, 95% CI = 1.03, 2.15, P = 0.034). The inclusion of the mixed malaria episodes (both P. falciparum and P. vivax) also demonstrated a statistically significant > 50% increased risk of malaria in the iron supplemented group (IRR = 1.54, 95% CI 1.06, 2.24, P = 0.03). Zinc and iron plus zinc did not statistically significantly affect P. vivax incidence in all children (P > 0.36).
|
5 years of age, iron increased the risk of a P. vivax episode (IRR = 1.76, 95% CI 1.14, 2.70, P = 0.010). No statistically significant effect was observed for zinc or iron plus zinc on P. vivax in children 5 years of age. In addition, no statistically significant interaction was observed between supplement type and baseline zinc or anemia status (P > 0.20). There were fewer P. falciparum episodes than were expected in this population (60 episodes). The main effects of iron and zinc, as well as their interaction, were not statistically significant (P > 0.23).
Acute lower respiratory infection.
Over the course of the study, 141 episodes of ALRI occurred. The majority of episodes (90) were observed in the less than five years of age group, and only 10 episodes were observed in children more than nine years of age. As shown in Table 3, no significant effect was found with either of the supplements or their interaction on incidence of ALRI (P > 0.40).
Diarrhea. There were 1,737 episodes of diarrhea in the four treatment groups. The interaction between iron and zinc was statistically significant (P = 0.03). When we controlled for a correlation using GEEs, children who received zinc were 23% less likely to experience a diarrhea episode within a child period than children who received placebo (IRR = 0.77, 95% CI 0.61, 0.96, P = 0.02). The effect of iron and iron plus zinc on diarrhea morbidity was not statistically significant (P > 0.50). Interaction with age was found to be statistically significant (P = 0.05), and subgroups of age were analyzed separately. Children less than five years of age who received zinc alone were 30% less likely to experience diarrhea within a child period than were those who received placebo (IRR = 0.70, 95% CI 0.54, 0.92, P = 0.01). Children less than five years of age who received iron or iron plus zinc did not have significantly different diarrhea morbidity incidence than the children who received placebo (iron: IRR = 0.97, 95% CI 0.78, 1.21, P = 0.78 and iron plus zinc: IRR = 0.89, 95% CI 0.70, 1.12, P = 0.32). Among children more than nine years of age, iron and iron plus zinc supplementation significantly increased the likelihood of experiencing a diarrhea episode within a child period (iron: IRR = 1.72, 95% CI 1.06, 2.79, P = 0.03 and iron plus zinc: IRR = 1.99, 95% CI 1.18, 3.34, P = 0.009). No statistically significant interaction was observed between supplement type and baseline zinc or anemia status (P > 0.24).
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
5 years of age in Papua New Guinea, which found no protective effect of zinc supplementation against health center visits for P. vivax episodes, although visits for P. falciparum episodes were reduced by 38%.31
Other findings of this study are novel or do not confirm previous research. This study is among the first to demonstrate an exacerbative effect of iron on P. vivax in children. Previous studies have documented an exacerbative effect of iron on P. falciparum malaria,6,7 and a study in Thailand indicated that iron plus folate supplements may exacerbate P. vivax in pregnant women.32 Iron alone increased the incidence of P. vivax malaria in all children; this increased risk was primarily in children 5 years of age. One potential explanation for the increased risk is that P. vivax preferentially invades young erythrocytes, and anemic erythrocytes containing elevated zinc protoporphyrin IX can inhibit parasite growth.33 Iron supplementation may therefore lead to conditions that are more favorable for parasite growth. Variations in the physiologic response to iron supplementation in older children (i.e. 5 years of age) may underlie the age-dependent effects observed here and warrants further investigation.
Iron plus zinc supplementation was protective against P. vivax in children less than five years of age, but had no effect on older children. It is of interest that zinc offset the exacerbative effects of iron on P. vivax in older children because, if confirmed, it may help overcome concerns related to iron supplementation in P. vivax malaria-endemic areas. These results are quite different from the findings in the Pemba study, which found that supplementation of children 1–35 months of age with iron and folic acid with or without zinc resulted in increased risk of serious adverse events related to P. falciparum malaria in a holoendemic area.7 The contrasting results may be due to species-specific differences in the effect of iron supplementation on P. falciparum and P. vivax malaria and/or on the effect of different levels of endemicity of malaria. Additional research into these questions is necessary to clarify the future of iron supplementation programs not only in hyperendemic regions, but also in regions with a variety of transmission levels.
Previous trials have demonstrated a decrease in the incidence of respiratory infection in children who receive zinc supplements,29,34,35 but no such effect was observed in this trial, perhaps due to the relatively small number of ALRI episodes observed. Another factor that may have contributed to the finding of no effect is the non-specific ALRI definition used that was based on reported symptoms. In other studies, the protective effect of zinc has been shown more clearly for clinically confirmed ALRI/pneumonia episodes.29,34
Iron plus zinc was not found to significantly affect diarrhea morbidity, unlike a recent study in Bangladesh that found iron plus zinc supplementation was associated with lower diarrhea morbidity in 799 six-month old children.36 The findings of no significant effect in our study could be due to the limited sample size available in the younger age group.
A key aim of the study was to compare iron and zinc supplementation alone and in combination for biochemical as well as morbidity outcomes. Iron (with and without zinc) increased hemoglobin concentration with no significant interaction between iron and zinc. Regarding final zinc concentration, the interaction term was significant, but when the individual supplement groups were compared, no difference was noted between the zinc and iron plus zinc supplements. Both of these findings indicate that the combination supplements will improve biochemical indicators of iron and zinc status in areas where deficiencies in both nutrients are common. These findings differ from the results of a trial in Indonesian infants, which found the combination supplement to be less effective than the individual supplements at decreasing iron and zinc deficiency.14 Although the combined supplement was determined to be less effective than iron or zinc supplementation alone, the investigators in the Indonesian infant study still recommended that the combined supplement be used in situations where deficiencies in both nutrients exist.
One curious finding at the final evaluation was that the group that received iron supplementation had a higher average plasma zinc concentration than did the placebo group. Iron supplementation has not been previously reported to improve zinc status15 and additional investigation of this unexpected finding is warranted.
There are a variety of routes to micronutrient sufficiency, including supplementation, fortification, dietary diversification, and decreased consumption of absorption-inhibiting foods. Although the ultimate goal is to provide complete nutrition to children through their daily diet, zinc and iron supplementation in P. vivax malaria–endemic areas where deficiency of these micronutrients is highly prevalent may be beneficial in children less than five years of age. However, the local endemicity and type of malaria should be considered when designing a supplementation program that includes iron, and the program should include heightened surveillance and treatment of malaria. If co-administration of iron reduces the benefits of zinc for infectious disease protection, as seen in this study for diarrhea, zinc can be administered singularly or possibly in fortified food where the adverse interactions of iron and zinc may be reduced.34 This study provides additional insight into the complicated relationships between micronutrient supplementation and infectious disease. Additional research is needed to clarify the roles of age, endemicity, and species-specific effects of iron on malaria.
Received November 4, 2005. Accepted for publication February 23, 2006.
Acknowledgments: We thank the parents and children who generously participated in the study; the study team, in particular Dr. Manuel Ríos for medical care of children, Sofia Madrid for field supervision, and Jorge Quintana for laboratory support; and the staff of Hospital Regional de Apoyo Iquitos and the Ministry of Health authorities in the region of Loreto, Peru for their assistance. We also thank Elizabeth Johnson and Ping Chen for statistical support and guidance.
Financial support: This study was supported by the United Nations Children’s Fund (New York) and the Johns Hopkins Family Health and Child Survival Cooperative Agreement with the United States Agency for International Development.
Disclosure: None of the authors had a personal or financial conflict of interest in this study.
* Address correspondence to Anuraj H. Shankar, Department of International Health, The Johns Hopkins University Bloomberg School of Public Health, 615 North Wolfe Street, Room E8527, Baltimore, MD 21205. E-mail: ashankar{at}jhsph.edu 
Author’s addresses: Stephanie A. Richard and Laura E. Caulfield, Center for Human Nutrition, Department of International Health, The Johns Hopkins University Bloomberg School of Public Health, 615 North Wolfe Street, Room W2041, Baltimore, MD 21205 Telephone: 410-955-2786, Fax: 410-955-0196, E-mails: cyq9{at}cdc.govlcaulfie{at}jhsph.edu. Nelly Zavaleta, Instituto de Investigacion Nutricional, Avenida la Universidad 685, La Molina, Lima 18, Peru, Telephone: 51-14-349-6023, Fax: 51-14-349-6025, E-mail: nzavalet{at}iin.sld.pe. Robert E. Black, Richard S. Witzig, and Anuraj H. Shankar, Department of International Health, The Johns Hopkins University Bloomberg School of Public Health, 615 North Wolfe Street, Room E8527, Baltimore, MD 21205, Telephone: 410-955-3934, Fax: 410-955-0196, E-mails: rblack{at}jhsph.edu, rwitzig{at}tulane.edu, and ashankar{at}jhsph.edu.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. M. Prentice Iron Metabolism, Malaria, and Other Infections: What Is All the Fuss About? J. Nutr., December 1, 2008; 138(12): 2537 - 2541. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Kelly, M. Katubulushi, J. Todd, R. Banda, V. Yambayamba, M. Fwoloshi, I. Zulu, E. Kafwembe, F. Yavwa, I. R Sanderson, et al. Micronutrient supplementation has limited effects on intestinal infectious disease and mortality in a Zambian population of mixed HIV status: a cluster randomized trial Am. J. Clinical Nutrition, October 1, 2008; 88(4): 1010 - 1017. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Kiedaisch, A. Akel, O. M Niemoeller, T. Wieder, and F. Lang Zinc-induced suicidal erythrocyte death Am. J. Clinical Nutrition, May 1, 2008; 87(5): 1530 - 1534. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
