点击显示 收起
1 From the Departments of Nutrition (KZL) and Epidemiology (EH), Harvard School of Public Health, Boston MA; the Center for Childhood and Adolescent Health, Secretary of Health, Mexico City, Mexico (YM); the Hospital Infantil de Mexico "Federico Gomez", Mexico City, Mexico (JIS); the School of Natural Sciences, Universidad Autónoma de Queretero, Queretero, Mexico (JLR)
2 Supported by the Instituto de Nutricion Danone, the National Council of Science and Technology of Mexico, and National Institutes of Health grant K01 DK06142-02. 3 Address reprint requests to KZ Long, Department of Nutrition, Harvard School of Public Health, 1633 Tremont Street, Boston MA, 02115. E-mail: klong@hsph.harvard.edu.
ABSTRACT
Background: The efficacy of micronutrient supplementation in improving childhood health and survival in developing countries may be specific to the micronutrient used and health outcome measured.
Objective: We evaluated the effect of vitamin A and zinc supplementation on overall rates of childhood diarrheal disease and respiratory tract infections and rates stratified by household and personal characteristics.
Design: A double-blind, randomized, placebo-controlled trial was carried out in which 736 children aged 6–15 mo living in a periurban area of Mexico City were assigned to receive vitamin A every 2 mo, zinc daily, vitamin A and zinc together, or placebo. Children were followed for 12 mo to determine overall counts of diarrheal episodes and respiratory tract infections.
Results: Vitamin A supplementation was associated with a 27% increase in diarrheal disease [risk ratio (RR): 1.27; 95% CI: 1.10, 1.45; P < 0.001] and a 23% increase in cough with fever (RR: 1.23; 95% CI: 1.02, 1.47; P = 0.02), whereas zinc had no effect on these outcomes. Vitamin A supplementation decreased diarrhea in children from households with dirt floors but increased diarrhea in children from households with nondirt floors, piped water, and indoor bathrooms. Zinc supplementation decreased diarrhea in children from households with dirt floors and whose mothers were more educated. Vitamin A supplementation increased cough with fever in children from less-crowded households that lacked indoor bathrooms and in children of less-educated mothers.
Conclusions: Vitamin A increases diarrheal disease and respiratory tract infections in young children in periurban areas of Mexico City. Vitamin A and zinc have more heterogeneous effects in different subgroups of children.
Key Words: Vitamin A supplementation • zinc supplementation • diarrhea • respiratory tract infection • Mexico
INTRODUCTION
The extensive research concerned with the effect of vitamin A and zinc supplementation on childhood health and survival in developing countries has established that the effect of these 2 micronutrients depends on which health outcome is being analyzed. Vitamin A supplementation reduces mortality and the severity of infectious disease episodes in young children (1–4) but appears to have no consistent effect on the incidence or duration of diarrheal disease and may even increase respiratory tract infections (5–8). Zinc supplementation, in contrast, has been found to reduce the incidence and duration of diarrheal disease and respiratory tract infections (9–11). It remains unclear why supplementation with each micronutrient produces these differences because no single hypothesized mechanism can explain these findings.
These specific effects of vitamin A and zinc on diarrheal diseases and respiratory tract infections may reflect an underlying heterogeneity in both the etiology of these 2 broad disease groups and the differential regulation of the immune response by micronutrients. Diarrheal disease and respiratory tract infections are caused by an extensive and diverse group of pathogens (12–14). The prevalence of these pathogens can vary between communities because of differences in such factors as water and sanitation facilities and maternal education, all of which can modify pathogen transmission (14–16). The differential regulation of the pathogen-specific immune response by vitamin A and zinc may produce divergent clinical outcomes for each pathogen (17, 18). This differential effect coupled with the differences in pathogen prevalence between communities may underlie the inconsistent effects of supplementation reported in the literature. Trials that examine how household and community conditions modify the effects of supplementation on childhood health outcomes may provide insight into whether such epidemiologic and immunologic mechanisms are responsible for such inconsistent results.
We carried out a placebo-controlled trial of the effect of vitamin A and zinc supplementation on diarrheal disease in a periurban community located on the eastern periphery of Mexico City to evaluate the effect of frequent low doses of vitamin A or zinc on these rates. In this article we address the hypothesis that children supplemented with zinc will have reduced rates of diarrhea and respiratory tract infections when compared with children who only received vitamin A or placebo, whereas vitamin A will have no consistent effect. We then explore whether these hypothesized effects are modified within strata of characteristics related to household water and sanitary facilities and hygiene.
SUBJECTS AND METHODS
Study population
This study involved young children living in La Magdalena Atlicpac, a community with 36 581 inhabitants made up of recent immigrants to Mexico City, Mexico. The 9 neighborhoods or colonias that make up this community differ in their degree of access to potable water, sanitary facilities, and overall development; greater access was found in older communities and less access among more recently founded communities. Approximately 12% of the children were classified as having a height-for-age z score below –2 in surveys carried out by the Mexican Ministry of Health in this region. The study was approved by the ethical review committee from the National Center for the Health of Infants and Adolescences of Mexico (CENSIA).
This double-blind randomized trial used a factorial design in which children 6–15 mo of age were assigned to 1 of 4 groups: 1) a vitamin A group that received 20 000 IU retinol every 2 mo for children aged 1 y or 45 000 IU for children aged >1 y, 2) a group that received a daily dose equivalent to 20 mg elemental Zn as zinc methionine, 3) a group that received both the zinc supplement and the vitamin A supplement as above, and 4) a placebo group. This design would allow the evaluation of the effect of frequent low doses of vitamin A or zinc given independently or in combination on overall and subgroup rates of infection. Zinc methionine was used as a zinc supplement because of its better absorption relative to other chemical forms of zinc (19).
Community health workers identified eligible children by carrying out a census of all children aged < 2 y living within the community. Eligible children were those between 6 and 15 mo of age whose parents had given informed consent. Children were excluded from the study if they had diseases causing immunosuppression and any congenital or acquired alteration of the digestive tract that could alter the absorption of micronutrients. Children who were taking vitamin supplements were also excluded from the study. Once enrolled, children were randomly assigned to 1 of 4 groups and followed for 1 y. The randomization sequence was generated by using a random-number table by project personnel from CENSIA, a division of the Mexican Ministry of Health.
Micronutrient supplements
The vitamin A, zinc, and vitamin A + zinc supplements were prepared by personnel at the National Institute of Nutrition in 5-mL solutions that were similar in taste and appearance. Children in the placebo group received the same solution without any supplement. These solutions were packaged in consecutively numbered, color-coded, opaque plastic droplet bottles to ensure that field personnel and the principal investigator were blinded. During the 12-mo follow-up period, field personnel administered the solution containing the respective supplements twice a week to all children in the 3 treatment arms and placebo group and left bottles for the remaining days of the week with instructions on how to administer the solution.
Household surveillance for morbidity
At the first household visit, the mothers were interviewed to determine their children's morbidity experience in the previous 2 wk. Information was also collected at this baseline visit regarding the number of household members and their ages, education levels, income, household construction materials, the source and quality of household water, the type of household sanitation facilities, and household possessions. The mothers or child caretakers were also interviewed regarding the feeding and weaning patterns of the children. A previously validated questionnaire was used to collect these data by project personnel who had received training in its application.
After enrollment, each child was visited twice a week for 12 mo. Standardized, previously validated morbidity questionnaires were used during these visits to interview mothers regarding the presence of the following symptoms in their children: diarrhea, the number and consistency of stool evacuations, the presence of blood and mucus in stools, fever, cough, and difficulty breathing. Project supervisors accompanied 5% of all household visits to ensure the quality of data collection. Children were referred to the study physician for diagnosis and treatment when the fieldworker or caregivers were concerned about their children's health status.
Trained personnel used calibrated methods and standardized techniques to obtain weight and height measurements from each child at the first household visit and once a month thereafter to evaluate the nutritional status of each child. Children were weighed with the use of a calibrated Salter scale, to a precision of 1 g, while undressed or wearing light clothing. Supine length was measured within 1 mm with the use of a locally constructed rigid measuring board. Training exercises at the beginning of the study and at repeated intervals thereafter were used to standardize these personnel.
Sample size
Epidemiologic surveillance reports from the Mexican Ministry of Health led us to assume that the study population would have a diarrheal disease rate of 3 episodes per child per year. It was calculated that a sample size of 800 children, 200 per group, was required to detect a 20% reduction between the control and treatment groups with a power of 80% and 95% significance level and an expected loss to follow-up of 20%.
Data analysis
The endpoints for the study were the prevalence of diarrheal episodes and respiratory tract infections during the 1-y follow-up of each child. A diarrheal disease episode was defined as the passage of 3 liquid stools by the child followed by 3 symptom-free days as reported by the mother or caretaker of the child. Mutually exclusive subgroups of diarrhea were defined: acute diarrhea was defined as an episode that lasted between 1 and 14 d, whereas persistent diarrhea was defined as an episode that lasted >14 d. Acute diarrhea was classified as dysentery if blood or pus was present in the stool or as watery diarrhea if neither blood nor pus was present. Respiratory tract infection outcomes were defined as the occurrence of cough alone, cough and fever, or cough and rapid respiratory rate as reported by the mother. A new episode was defined if there was an interval of 14 d with no symptoms.
Anthropometric indicators were calculated by using sex- and age-specific standards developed by the US Centers for Disease Control and Prevention (20). Children with length-for-age z scores below –2 were considered stunted, those with weight-for-length z scores below –2 were wasted, and those with both of these measurements were considered stunted and wasted. Children with z scores at or above –2 for both of these measurements were classified as nonstunted or nonwasted.
The rate ratio (RR) and 95% CIs for risk of diarrheal disease and respiratory tract infections were estimated by fitting Poisson regression models to the total number of diarrheal disease episodes and respiratory tract infections in the different treatment groups during the follow-up period (21). The primary analysis of the main effects of vitamin A and zinc for each outcome was based on the factorial design of the study: comparisons were made between groups that received zinc (zinc and zinc + vitamin A) and those that did not receive zinc (vitamin A or placebo) and groups that received vitamin A (vitamin A and vitamin A + zinc) and groups that did not receive vitamin A (zinc and placebo). An additional analysis for each outcome was then carried out in which a term for the vitamin A–zinc interaction was included in the model. The results were then presented by using the original 4 study arms when an interaction was found. Statistical significance was set at a probability level of <0.05, or <0.1 for interactions. Separate models were also run within strata of childhood and household characteristics in an effort to determine how the effects of vitamin A and zinc were modified within these different strata. The 3-way interactions between vitamin A, zinc, and each characteristic were formally tested by using the multivariate score test. The variables of special interest in this analysis were those that are indirect measures of pathogen prevalence and transmission, such as water source, availability of indoor toilets, presence or absence of dirt floors, and maternal education (6 y, >6 y). Data were analyzed by using the PROC GENMOD procedure in the Statistical Analysis System software (version 8.2; SAS Institute, Cary, NC).
RESULTS
In all, 800 children were recruited for the study after the consent of their mothers. Seven children migrated from the area with their families immediately after being randomly assigned, so they never entered the study. A total of 93 children were lost to follow-up or were excluded; 736 children remained in the study and were followed for 12 mo: n = 181 in the zinc group, 192 in the vitamin A + zinc group, 180 in the vitamin A group, and 183 in the placebo group (Figure 1). The duration of follow-up in the 3 treatment arms and the placebo group was not significantly different (vitamin A group, 13.1 mo; no–vitamin A group, 13.05 mo; zinc group, 13.2 mo; and no-zinc group, 12.96 mo; P = 0.62). There were no significant differences in compliance rates for scheduled supplements between children in the vitamin A (87.5%), zinc (87.3%), vitamin A + zinc (88.4%), and placebo groups (89%), respectively.
FIGURE 1.. Trial profile.
The distribution of sociodemographic characteristics of study children and households was not significantly different between the 4 groups (Table 1). The average age of children at recruitment was 9.8 ± 4.7 mo. The proportion of households with access to piped water within each group ranged from 73% to 78%, between 46% and 56% of the mothers in the study had >6 y of schooling, and 7–13% of the children were stunted.
View this table:
TABLE 1. Characteristics of the children at enrollment by treatment group
Vitamin A had no significant effect on diarrhea in the factorial analysis. However, there was a significant interaction between vitamin A and zinc in this analysis for overall diarrhea and persistent diarrhea (P = 0.05 and 0.01, respectively). In the 4-group analysis, vitamin A was associated with a significant 27% increased prevalence of diarrhea (RR: 1.27; 95% CI: 1.10, 1.45; P < 0.001, Table 2) and a significant increase in the prevalence of persistent diarrhea (RR: 3.70; 95% CI: 2.06, 6.65; P < 0.01). Zinc had no effect on these outcomes. Vitamin A and zinc also had no effect on dysentery.
View this table:
TABLE 2. Risk of diarrheal disease, dysentery, and persistent diarrhea in 736 children in the vitamin A and zinc diarrheal disease trial according to treatment group1
A significant 3-factor interaction was found in the analysis of diarrhea stratified by household characteristics and so the analysis of the original 4 study arms was carried out. In this analysis, children in the 3 treatment arms from households with dirt floors had a significantly lower prevalence of diarrheal disease than did children in the placebo group: vitamin A group (RR: 0.64; 95% CI: 0.48, 0.84; P = 0.04), zinc group (RR: 0.67; 95% CI: 0.51, 0.88; P = 0.5), and vitamin A + zinc group (RR: 0.46; 95% CI: 25, 0.82; P = 0.01) (Table 3). Vitamin A supplementation was associated with a 46% increase in the prevalence of diarrheal disease among children from households with tile or mosaic floors (RR: 1.46; 95% CI: 1.07, 2.00; P = 0.01). Finally, vitamin A was associated with an increases in diarrhea among children from households with indoor bathrooms (RR: 2.04; 95% CI: 1.00, 4.13; P = 0.04).
View this table:
TABLE 3. Risk of diarrheal disease in children in the vitamin A and zinc diarrheal disease trial according to treatment group and household socioeconomic factors1
Vitamin A + zinc supplementation was not associated with significant effects on overall cough or cough with difficulty breathing. However, vitamin A supplementation was associated with a significant increase in cough with fever in the factorial analysis (RR: 1.23; 95% CI: 1.02, 1.47; P = 0.02, Table 4). Neither vitamin A nor zinc had significant effects on fever alone in the factorial or 4-group analysis. No significant 3-factor interaction was found in the analysis of respiratory tract infections stratified by household characteristics. In the 2-factor stratified analysis, vitamin A–supplemented children from less crowded households had a 30% greater prevalence of cough with fever (RR: 1.3; 95% CI: 1.06, 1.6; P < 0.01, Table 5), whereas there was a 32% decrease in the prevalence of cough with fever among supplemented children from more crowded households (RR: 0.68; 95% CI: 0.42, 1.10; P = 0.11, P for interaction = 0.01). Significant increases in cough with fever were also found among children of less-educated mothers (RR: 1.25; 95% CI: 1.03, 1.52; P = 0.02) and children from households with no access to piped water (RR: 1.22; 95% CI: 1.01, 1.42; P = 0.03) or that lacked indoor bathrooms (RR: 1.35; 95% CI: 1.09, 1.66; P < 0.01).
View this table:
TABLE 4. Risk of overall cough, cough with difficulty breathing, and cough with fever in 736 children in the vitamin A and zinc diarrheal disease trial according to treatment group1
View this table:
TABLE 5. Risk of respiratory tract infections in children in the vitamin A and zinc diarrheal disease trial according to treatment group and household socioeconomic factors1
DISCUSSION
We found that vitamin A supplementation was associated with an increase in overall diarrheal disease and cough with fever infections among young children who live in periurban communities of Mexico City, whereas zinc had no significant effects on any of these or other outcomes. We found more complex associations between supplementation and specific outcomes among subgroups of children stratified by household and personal characteristics. These subgroup effects of vitamin A and zinc on health outcomes suggest that such factors as household flooring and access to both potable water and sanitation facilities are indeed modifying the response of this population to supplementation.
These results partly agree with previous findings regarding the effect of vitamin A supplementation on diarrheal disease and respiratory tract infections but differ from outcomes reported in zinc supplementation trials. Studies carried out in different regions of the world have reported that vitamin A supplementation has an inconsistent effect on diarrhea and may increase respiratory tract infections (7, 8, 22–24). Many trials have also reported either positive or negative effects of supplementation on diarrhea and respiratory tract infections among children stratified by age, nutritional status, and feeding mode when no clear effects were found on the overall main outcomes (4, 25). In contrast, zinc supplementation has consistently produced beneficial effects on these outcomes reported in previous studies (9, 26–28). A previous study of zinc supplementation among children living in a rural area of Mexico showed a 34% decrease in diarrhea incidence (29). In this current study a similar reduction occurred only in children from households with dirt floors and in children of more educated mothers.
Baseline differences in vitamin A and zinc status among study children may explain these effects with deficient children benefiting more and nondeficient and marginally deficient children benefiting less. In our study it was not possible to address whether baseline differences in status or the change in status may have contributed to these results because we had no biochemical indicators of the vitamin A and zinc status of study children. The National Nutrition Survey of Mexico has reported that 33–34% of children aged <5 y were zinc deficient (serum zinc concentrations <65 µg/dL) and <5% had deficient plasma retinol concentrations (<10 µg/dL) (30). However, children from urban areas in Mexico such as La Magdalena are less at a risk of being zinc deficient than rural children because the bioavailability of zinc is greater in more refined urban rural diets than in rural diets that consist primarily of maize (31, 32).
These findings may also be due to the differential effects that vitamin A and zinc have on the pathogen-specific immune response resulting in distinct clinical outcomes for specific pathogen infections. Vitamin A deficiency induces an up-regulation of the T helper subset 1 cell (Th1)–mediated response, whereas vitamin A supplementation up-regulates the T helper subset 2 cell (Th2) response (18, 33). In contrast, zinc deficiency is associated with a reduced Th1 response, whereas zinc supplementation leads to a more dominant Th1 response (18, 34). The lack of an overall significant effect of zinc supplementation in our study may have been due to the reduced prevalence of such pathogens as Salmonella spp. and S. typhoid after implementation of sanitary measures and hygiene education programs in Mexico (35). Zinc up-regulates the Th1 response, which protects against these pathogens (36).
The increase in diarrhea in our study may also reflect the relatively greater prevalence of such gastrointestinal pathogens as rotavirus. Supplementation with vitamin A may result in children being less protected against this pathogen because vitamin A down-regulates the Th1 response, which protects against infections by this pathogen (37, 38). Similarly, the adverse effects of vitamin A on specific respiratory tract pathogens may underlie the adverse effects of vitamin A on cough with fever in our study, because vitamin A supplementation is associated with adverse outcomes in children infected with respiratory syncytial virus (39).
The effect of vitamin A and zinc on diarrhea and respiratory tract infections among children stratified by household socioeconomic factors may be due to differences in the prevalence of specific pathogens. Epidemiologic studies have reported that disposal of feces, access to water, degree of crowding, household floor construction, and maternal education can determine transmission and, thus, the prevalence of distinct types of diarrheal and respiratory tract pathogens (40, 41). As a result, children from households with poorer measures of socioeconomic status (SES) may be exposed to different groups of enteric and respiratory tract pathogens than children from households with better SES measures. However, not all pathogens may be more prevalent among households with poorer SES measures given the distinct transmission patterns for each of these pathogens. Improved water access, for example, may increase the exposure of children to such water transmitted parasites as Giardia lamblia if the quality of the piped water is low. Maternal education may also modify the transmission of pathogens in unexpected ways. A study in Malaysia has shown that improvements in access to water benefits infants of more educated mothers who take advantage of these improvements, whereas improved sanitation benefits more infants of less educated mothers (15).
Overall, we found that vitamin A and zinc supplementation have distinct effects on different health outcomes among children in periurban communities of Mexico City. We have suggested that these distinct effects may reflect the effect of vitamin A and zinc on specific pathogen infections. If future research establishes that micronutrients indeed have distinct effects on health outcomes associated with specific pathogenic infections, the design of micronutrient supplementation programs should take into account what pathogens are causing the episodes of morbidity. The targeting of subgroups of children would be one obvious strategy that could potentially enhance the effectiveness of such supplementation programs in improving the overall health and survival of children in Mexico.
ACKNOWLEDGMENTS
We thank Cristina Barrgan and Matilde Juarez of the Clinical Microbiological Laboratory, Hospital La Perla, Secretary of Health, Cda Netzahuaycoytl, Mexico, for their assistance in the laboratory.
KZL, JIS, and JLR were involved in the design of the project, procurement of funds, and project supervision. YM was involved in the coordination of the project. KZL and EH were involved in the data analysis. KZL was responsible for the preparation of the article, with input from JIS, JLR, and EH. None of the authors had a conflict of interest. KZL had full access to all the data in the study and had final responsibility for the decision to submit the article for publication.
REFERENCES