QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Online Supplemental Material Free Access Article All Versions of this Article: 139/11/2154S    most recent jn.109.105312v1 Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Savy, M. Articles by Prentice, A. M. PubMed PubMed Citation Articles by Savy, M. Articles by Prentice, A. M.

---

Supplement: Landscape Analysis of Interactions between Nutrition and Vaccine Responses in Children

Landscape Analysis of Interactions between Nutrition and Vaccine Responses in Children^1–3,

⁵ Medical Research Council (MRC) International Nutrition Group, ⁶ Infectious Diseases Epidemiology Unit, ⁷ Immunology Unit, London School of Hygiene and Tropical Medicine, London, UK, WC1E 7HT and ⁸ MRC International Nutrition Group, Keneba, The Gambia

^* To whom correspondence should be addressed. E-mail: Mathilde.Savy{at}lshtm.ac.uk.

	ABSTRACT

TOP ABSTRACT Introduction Methods Results Discussion LITERATURE CITED

 
The world's poorest children are likely to be malnourished when receiving their childhood vaccines. It is uncertain whether this affects vaccine efficacy and whether the coadministration of nutrient supplements with vaccines has beneficial or detrimental effects. More recently, a detrimental interaction between vitamin A (VA) supplementation (VAS) and the killed diphtheria-tetanus-pertussis vaccine given in early childhood has been suggested. This report provides a critical review of the published interactions between nutritional status and/or supplementation and vaccine responses in children. Due to an absence of evidence for most nutrients, this analysis focused on protein-energy, vitamins A and D, and iron and zinc. All vaccines were considered. Both observational studies and clinical trials that led to peer-reviewed publications in English or French were included. These criteria led to a pool of 58 studies for protein-energy malnutrition, 43 for VA, 4 for vitamin D, 10 for iron, and 22 for zinc. Our analysis indicates that malnutrition has surprisingly little or no effect on vaccine responses. Evidence for definitive adjunctive effects of micronutrient supplementation at the time of vaccination is also weak. Overall, the paucity, poor quality, and heterogeneity of data make it difficult to draw firm conclusions. The use of simple endpoints that may not correlate strongly with disease protection adds uncertainty. A detailed examination of the immunological mechanisms involved in potential interactions, employing modern methodologies, is therefore required. This would also help us understand the proposed, but still unproven, negative interactions between VAS and vaccine safety, a resolution of which is urgently required.

	Introduction

TOP ABSTRACT Introduction Methods Results Discussion LITERATURE CITED

In contrast to vaccination, undernutrition is an area of little progress. Today, undernutrition is still the underlying cause of more than one-third of deaths in children under 5 y old, representing 3.5 million child deaths every year. Current forecasts indicate that, due to slow progress in South Asia and sub-Saharan Africa, the UN Millenium Development Goal target for reducing underweight by 2015 will not be met. The recent Lancet series on maternal and child malnutrition highlighted that several nutritional interventions were highly effective in reducing stunting, micronutrient deficiencies, and child deaths (2). Counseling about breast-feeding and supplementation of key micronutrients, including vitamin A (VA) and zinc, has been identified as having the greatest potential to reduce child morbidity and mortality.

Full vaccine efficacy requires a full set of immunological responses, starting with antigen recognition and presentation through to memory cell immortalization and antibody production, or priming of cellular responses. Each of these steps is sensitive to numerous indirect effects, such as the coadministration of adjuvants, or variations in the state of background immunostimulation in the recipient. Nutritional status may also play an important role in immune responses to vaccination. However, over the years, studies have produced conflicting evidence. The nutritional status of a child influences its immune responses in varied and complex ways that are far from fully understood (3,4). The specific effects on vaccine efficacy have not been extensively reviewed and there is a need to gather all the available data together to inform future policy making.

The particular stimulus for this review was a controversy concerning suggested detrimental interactions between supplemental VA administration and the killed DTP vaccines given in early childhood. Some studies have suggested that the coadministration of VA and DTP vaccine might be associated with increased child mortality, especially in girls (5–7). The hypothesis is that VA may interact negatively with killed vaccines by amplifying their nonspecific effects on childhood mortality (8).

This report provides a critical review of published knowledge, gaps, and promising opportunities for learning related to the interactions among nutrition, nutritional interventions, and vaccine responses. It also includes a review of the potential adverse outcomes of simultaneous administration of nutritional supplements and vaccines. In particular, this will include a summary of the current controversy regarding potential downstream effects on mortality of negative interactions between VA supplementation (VAS) and DTP vaccination in infancy.

	Methods

TOP ABSTRACT Introduction Methods Results Discussion LITERATURE CITED

 
Databases and keywords

The databases used to identify studies investigating possible interactions between nutrition and vaccine responses were PubMed, Medline, and Google scholar databases. The keywords used for this search are summarized in Table 1.

The research was restricted to clinical trials and observational studies conducted in humans, which led to publications written in English or French, with no limit regarding publication dates. A decision was made not to include unpublished studies or datasets not fully analyzed because of the absence of peer review. Although the major target group was children, we included studies relating to individuals of any age, because we considered that this might help in understanding any relationship between nutrition and vaccine responses. In practice, almost all studies were based on children. Studies with extremely small sample sizes (<10 participants per group) were not reviewed but are mentioned to make the reader aware of all the data available in the literature. We identified a few studies in which the target group was not relevant for this article, e.g. HIV-positive or hemodialysed individuals or unhealthy elderly individuals. These studies are mentioned but not formally reviewed. We have advisedly included papers from Chandra et al. (9) in this review [see cautionary note (10)]. After the identification and inclusion processes, we scanned the references cited in each study to check for possible omissions.

Nutrients reviewed

We focused on vitamins A and D, iron, zinc, and protein-energy, because initial searches indicated a virtual absence of studies relating to other micronutrients. Deficiencies in the nutrients we focused on have known effects on the immune system (Table 2).

As no search restriction on vaccines was used for this review, the vaccines against the main target diseases in developing and/or developed countries were included: diphtheria, pertussis, tetanus, Hib, pneumococcal and meningococcal diseases, poliomyelitis, measles, mumps, rubella, hepatitis A and B, rotavirus infection, influenza, varicella, tuberculosis, yellow fever, and Japanese encephalitis. Summary information on these diseases is available as Supplemental Table 1. Vaccination schedules designed for children in the USA and UK (used as representatives for other countries in the developed world) and in developing countries are also available as supplemental material (Supplemental Tables 2 and 3).

Vaccines use different technologies to induce a protective response to the disease they are targeted against. The 4 types of traditional vaccines include: live attenuated vaccines, inactivated or killed vaccines, toxoid vaccines, and subunit vaccines. A number of innovative vaccines are also in development or in use, including polysaccharide vaccines, conjugate vaccines, recombinant vaccines, and DNA vaccines. Summary information on these vaccines is available as Supplemental File 1. The different types of vaccines induce different immune responses (11) and some of them may need to be combined with adjuvants to enhance this response (Table 3). A comprehensive review of immune responses is beyond the scope of this paper, but a brief summary tha provides a background against which to interpret the possible interactions between nutrition and vaccine efficacy is available as supplemental material (Supplemental File 1). Further information can be found in several excellent reviews and textbooks (12–15). Factors that can influence the nature and strength of the response in different individuals are also detailed in Supplemental File 2, including those that are directly or indirectly related to nutrition, such as prematurity and gender, but also nonnutritional factors such as smoking and stress (Fig. 1).

View larger version (26K): [in this window] [in a new window]   FIGURE 1  Factors affecting vaccine responses.

Once identified, studies were reviewed by the main researcher (M.S.) and data were extracted into 2 databases. The first database encompassed detailed information about the studies, including the methodology used (design, sample size, inclusion and exclusion criteria, intervention characteristics, nutritional status assessment, vaccine response outcomes, and method of assessment), the main results on vaccine response and on adverse events when available, and some elements of discussion/interpretation. These files also included studies that did not assess vaccine response but only adverse events. These tables are available upon request from the first author. The second database consisted of summary tables that condensed the main findings. All database files were then examined by a multidisciplinary Expert Advisory Board of senior academics from the London School of Hygiene and Tropical Medicine, including experts in nutrition, immunology, and vaccines (listed as coauthors). No exclusions were made, but benchmarks were determined for studies with poor quality (e.g. clinical trials without randomization, very small sample size, etc.) or with irrelevant information.

Endpoints

Immunogenicity (measured as the proportion of recipients who mount an antigen-specific response) is sometimes used as a surrogate of protection induced by vaccination. For vaccines that induce antibodies, then, these are usually the outcome measurement (either serological or mucosal). A variety of methods is available and particular tests are preferred depending on the vaccine and antibodies studied (e.g. ELISA or hemagglutination tests). The quantity of antibodies detected can be expressed as a concentration or titer. Seroconversion and seroprotection rates (proportion above some threshold thought to be protective) may also be used with appropriate cutoffs that vary according to the vaccine studied (16). The variety of serological parameters and cutoffs used in the literature reflects both methodological differences and uncertainty over the significance and implications of such parameters in relation to protection and can make comparison among studies difficult, especially among populations with different epidemiological backgrounds.

In some situations, the type of antibodies induced by vaccination must also be considered. With many bacterial pathogens, it is important to develop antibodies that kill the pathogen or enhance phagocytosis. Vaccination can sometimes induce the production of antibodies detectable by ELISA but which do not have opsonic or bactericidal properties. This explains, e.g., the low efficacy of unconjugated polysaccharide in infancy, at least for Meningococcus Group C. Similarly, antibody avidity may be an important factor in assessing vaccine efficacy. This affinity generally increases with repeated exposure to the antigen, because B cells with higher affinity antigen receptors are selected to produce larger clones of antibody-secreting plasma cells. Unfortunately, this information is rarely available for studies examining vaccine efficacy. The action of most vaccines depends also on a cellular immune response, if only to help B cells to produce antibodies. Certain vaccines, such as the bacille Calmette-Guerin (BCG) vaccine, predominantly induce a cellular immune response. Parameters of cellular immune responses, including T cell proliferation, cytokine secretion, and delayed-type hypersensitivity (DTH) tests, can thus be used as indicators of the immunogenicity of vaccines.

Vaccine efficacy can best be assessed by comparing the incidence of the target disease between vaccinees and nonvaccinees, ideally in a trial setting. The major advantage of such an approach is that the actual efficacy is measured rather than a surrogate, avoiding the problem of interpreting presence or absence of immunological markers. Observational studies must be very carefully designed to avoid a variety of potential biases. Such biases may come from differences between the vaccinee and the nonvaccinee groups, unequal exposure and risk of disease between these 2 groups, problems in the case definition, or misclassification in vaccination status and disease records. These points have been detailed in a review from Oreinstein et al. (17). Among them, difference in disease exposure between the groups to be compared is particularly challenging. In addition to the possible influences of prior disease and subclinical infections on vaccine-induced response, community protection through the "herd" effect may also have an impact on vaccine efficacy measurement (17,18).

Meta-analyses

Meta-analyses were performed when sufficient datasets were available (at least 2 studies with comparable outcomes). These were performed using Review Manager version 5 (19). Data from different studies were pooled using a Mantel-Haenszel method (dichotomous data) and a pooled odds ratio with 95% CI was calculated for each parameter. Heterogeneity between studies was tested using the chi-square statistic. Heterogeneity was considered significant when the test P-value was >0.10. Because this test is poor at detecting true heterogeneity among studies, we also used the statistic I² provided by the software. This statistic quantifies the effect of heterogeneity by providing the degree of inconsistency in the studies' results. A value of 0% indicates no observed heterogeneity and greater values indicate increasing heterogeneity. I² values >50% reflect a moderate heterogeneity and values >75% reflect a high heterogeneity (20). A fixed effects model was used to pool the data when there was no significant heterogeneity and a random effects model was used in cases of significant heterogeneity. In the latter case, further analyses using subgroups were performed to assess possible causes of heterogeneity. Funnel plot analyses were used to evaluate publication bias.

VA and mortality/adverse events

A special section has been dedicated to the controversy regarding a possible effect of VA administered with a vaccine on subsequent mortality and other adverse events. The publications that were selected and reviewed with regard to vaccine response have also been screened for results on mortality and adverse events. An additional literature search was performed using the following keywords: mortality, morbidity, adverse events, side-effects, VA, retinol, vaccine, and immunization to identify specific studies that addressed this issue.

	Results

TOP ABSTRACT Introduction Methods Results Discussion LITERATURE CITED

 
Evidence of the effect of nutritional status on immunity has been reviewed elsewhere; therefore, the present paper provides only a brief summary and relevant references for each of the nutrients considered (Table 2). When possible, a specific focus is made on the effect of each nutrient on immune functions that might affect the vaccine response. The influence of nutritional status and supplementation on vaccine responses is much less known and is comprehensively reviewed in the following sections. Our literature search led to a final set of 137 studies, of which 58 concerned protein-energy malnutrition (PEM), 43 concerned VA, 4 vitamin D, 10 iron, and 22 zinc.

Protein and energy

    PEM and immunity. Many studies have shown that individuals with severe PEM present functional abnormalities in 1 or more host defense systems. Excellent reviews have been published on this topic (21–23). In addition to a depression in nonspecific defenses and cell-mediated immunity, PEM is usually associated with thymic atrophy and diminished numbers of T helper (Th) and T cytotoxic lymphocytes (24,25). This results in a decrease in normal T cell functional capacity, potentially leading to impairment of all immune mechanisms requiring T cell involvement. Studies have shown that PEM was associated with a decrease in Th1-type cytokines, such as interferon- and interleukin (IL) 2, whereas the effect on Th2-type cytokines (IL-4, IL-5, or IL-10) is less clear (26,27). DTH test is also depressed (28,29). In contrast, B cells and antibody responses seem less affected in PEM, as reviewed in the following section.

PEM and supplementation and vaccine responses

Details of the studies reviewed are available in Appendix 1.

    BCG. A relatively large number of studies investigating the interactions between PEM and vaccine responses have focused on BCG vaccine. Immune response induced by BCG vaccine is mainly cellular immunity. Skin reaction to the vaccine (scar) and response to purified protein derivative (PPD) of Mycobacterium tuberculosis (an immunogenic component of BCG) are used as measures of BCG vaccine response. The latter test is also known as the tuberculin skin test and is a measure of DTH. The scar induced by BCG and the reaction induced by PPD are both correlated with BCG vaccination but do not directly assess the efficacy of BCG vaccine. The presence of a BCG scar means that there has been inflammation, whereas DTH estimates T cell responses.

All the studies we reviewed were observational studies, generally using small sample sizes, and conducted 20–40 y ago. They all used the tuberculin sensitivity skin test. Although most of these studies used 5 tuberculin units (TU) of PPD, some used 2 TU only, therefore making the comparison between studies difficult. Criteria and growth references used to define malnutrition were also widely variable. Despite these differences, most of the studies we reviewed found an effect of PEM on DTH in infants and children (28,30–35). On the whole, malnourished children, mostly determined by their weight-for-age, were found to be less likely to have a positive response to the tuberculin test, as defined by an induration diameter of >5 mm. In these studies, the mean size of induration was also generally lower in malnourished children than in well-nourished controls. Most of these studies specified that the children were negative to the skin test before BCG vaccination. These findings were observed in severely but also in moderately malnourished children, except in 1 study where the response to the tuberculin test was depressed in severely malnourished children only (30). Only 4 studies did not corroborate such results (36–39). In Iran, a very small study (n = 26) showed that malnourished children had similar responses to the tuberculin test to well-nourished children, the cutoff for a positive response having been fixed at 6 mm rather than 5 mm (38). The mean size of induration did not significantly differ between the groups either. Another study conducted in India found that 1- to 5–y-old children suffering from mild, moderate, or severe PEM had similar rates of positive reaction to the skin test (induration size 8 mm) compared with well-nourished children. On the other hand, in the same study, the percentage of positive reactions, as well as the mean diameter of induration, were both significantly lower in children with kwashiorkor compared with the other groups (39). In older Indian children (5–12 y), nutritional status was not associated with the mean size of induration after the tuberculin test; however, conversion rates were not presented (37). In Nigeria, Greenwood et al. (36) found no significant differences in both the percentage of skin test conversion (induration 5 mm) and mean diameter of induration between the well-nourished and malnourished children, although in this study, only 3 children were severely malnourished.

    OPV. Four observational studies have assessed the effect of malnutrition on responses to OPV in young children. In Nigeria, the mean increase in antibody titers after 3 doses of OPV did not significantly differ between infants severely or moderately malnourished (according to their weight-for-age; n = 31) and well-nourished infants (n = 10) (36). In addition, no correlation was found between the antibody response to OPV and serum albumin levels, used as an indicator of nutritional status. Similarly, a study conducted in 62 Indian children aged 1–12 mo showed no difference in seroconversion rates after 5 doses of the trivalent OPV between underweight and nonunderweight children (40). However, the analyses were not adjusted for any variables, not even the prevaccination antibody levels, which were found to be seropositive for some of the children. Chandra et al. (41)¹⁰ also observed that 10 malnourished Indian boys had a similar seroconversion rate and serum antibody levels to 10 control healthy boys, after receiving a single dose of the oral vaccine. In this study, malnutrition was diagnosed on the basis of a history of reduced dietary intake, loss of subcutaneous fat, hair changes, and weight and height between 50 and 70% of the mean on a reference growth curve. However, protein and albumin levels did not significantly differ between malnourished and healthy children. These findings were not in agreement with the 3rd study that was conducted in 11- to 23-mo-old Egyptian children (42). Children suffering from kwashiorkor and marasmus were matched with healthy children of the same age and sex and they all received a single dose of the OPV. The authors reported that antibody titers to the vaccine were significantly lower in children with kwashiorkor or marasmus than normal controls, suggesting a depression in antibody production against poliovirus vaccine in children with PEM.

    Measles. We also only found observational studies dating back to the 1970s and 1980s that examined PEM and measles vaccine (MV) response. Globally, those studies did not show any association between malnutrition and the immune response to MV (36,41,43–51). Children with kwashiorkor, marasmus, or different degrees of PEM defined by weight-for-age all had high seroconversion rates to the vaccine, as did the well-nourished children. In these studies, antibody titers to measles have been mainly assessed by the hemagglutination method. It is generally accepted that a 4-fold increase in antibody titers indicates good seroconversion after immunization. Unfortunately, for comparative purposes, the criteria used to define seroconversion varied widely between studies.

Only 4 studies have shown that immune responses to MV may be depressed in malnutrition, in particular with kwashiorkor. In Zimbabwe, children with kwashiorkor had a significantly lower conversion rate to the vaccine than well-nourished controls when assessed 2 wk after vaccination (52). However, only 12 malnourished and 11 well-nourished children completed this study and children from the former group were much older than the other children. Because children were undergoing nutritional rehabilitation during the study, antibody levels were not measured later than 2 wk postvaccination. The study therefore does not exclude the possibility that malnourished children may simply have a delayed immune response to measles vaccination. This possibility is validated by a South African study that observed a significant delay in anti-measles antibody production in children with PEM compared with well-nourished children (53). However, by d 21 after vaccination, both groups had good seroconversion rates. Salimonu et al. (25) observed that Nigerian children suffering from kwashiorkor had reduced anti-measles antibody levels measured 10 and 21 d after vaccination compared with well-nourished children. The increase in T lymphocyte percentage postimmunization, assessed by the fluorescent-activated cell sorting technique, was also lower in kwashiorkor children compared with the controls, but no difference was detected in B lymphocyte percentage. The 4th study was conducted in Egypt, where 27 infants with kwashiorkor or marasmus and 10 well-nourished control children received MV. Fewer infants in both the kwashiorkor and marasmic groups were found to have specific antibodies against measles above the seroconversion cutoff than infants in the control group (42).

    Smallpox and cowpox vaccines. One observational study compared the successful immunization rates to smallpox vaccine in children with different nutritional status (from normal to severely malnourished). The authors reported that among children with no previous smallpox scar, the overall success rate was 97% and did not differ between children with different nutritional status (49). Another observational study investigating smallpox vaccination in malnourished children has been identified; however, the authors examined only the appearance of the local reaction after vaccination (54).

    Yellow fever. One observational study enrolling 14 Ugandan children has been identified, but is not reviewed here because of the extremely small sample size (<10 participants in each group) (55).

    Inactivated vaccines DTP. Four of the 6 studies reviewed did not report any effect of PEM on the response to vaccination to either diphtheria or tetanus toxoids in children (25,28,36,56). In these 4 studies, the mean antibody titers, seroconversion, or seroprotection rates were not different between malnourished (moderate or severe) and well-nourished children after each of the 3 doses of DTP vaccine. Seroconversion rates were reported to be high whatever the nutritional status of the children. One study assessed the affinity of antibodies to tetanus toxoid after vaccination in 21 undernourished (marasmus or kwashiorkor) and 21 well-nourished children (56).¹¹ The mean affinity was significantly lower in undernourished children than in well-nourished children, especially after the first injection of the vaccine, but the results were only presented graphically with P-values for the differences. Nevertheless, in the same study, the percentage seroconverting did not differ between kwashiorkor, marasmic, and well-nourished children.

In Ecuador, a large observational study has tested the serum antibody response to tetanus antigen in children who had received 3 doses of the DTP vaccine, as indicated on their vaccination cards (57). Children were classified as malnourished if their height-for-age or weight-for-age were less than –2 Z scores. Mean IgG and IgM antibody titers to tetanus toxoid were significantly lower in stunted than in nonstunted children as well as in underweight vs. nonunderweight children, whereas no differences were detected regarding antibody titers to diphtheria toxoid. However, the proportion of children with antibody titers below the protective level did not differ (<5% for tetanus toxoid and <15% for diphtheria toxoid in all groups). In addition, a small study in Egypt described low diphtheria toxoid antitoxin levels in children suffering from kwashiorkor (n = 13) or marasmus (n = 12) compared with control children (n = 10) assessed 21 d after vaccination (42).

    Cholera. Two very small studies (n = 29 and 23, respectively) have investigated the association between PEM and the immune response to cholera vaccine in children. One study conducted in Egypt reported a significant rise in vibriocidal antibody titer after vaccination with cholera antigen in healthy children as well as in children with kwashiorkor or marasmus and no difference between these 3 groups (58). In the other study, the B-subunit cholera vaccine was administered to 1–8 y undernourished and well-nourished Bangladeshi children (59). The serum IgG and IgA antitoxin levels to immunization were not altered in the malnourished children.

    Rabies. In The Gambia, Moore et al. (3) studied the effect of nutritional status on the immune response to human-diploid cell rabies vaccine in 7- to 9-y-old children. They showed that anti-rabies antibody titers were positively associated with mid-upper arm circumference (MUAC) and BMI-for-age Z score after the first dose of rabies vaccine, even after adjustment for gender, age, and month of the study. However, none of the nutritional indices were associated with immune response after the second dose of the vaccine. The authors also measured plasma leptin concentration, the hormone derived from adipocytes, to test whether low leptin levels could be responsible for this impaired immune response as it had been previously suggested (60). Leptin concentration was associated with body fat mass as assessed by MUAC and BMI for age and was also significantly higher in girls than in boys, even after adjustment for anthropometrics. Leptin concentration was not related to antibody titers after the first or second dose of rabies vaccine in either males or females or in both sexes combined.

    Typhoid. The effect of malnutrition on typhoid vaccine response has been investigated in 2 observational studies in 88 and 45 Nigerian infants, respectively (36,61). In both studies infants were classified as malnourished or well nourished according to several indices, including weight-for-age, height-for-age, MUAC, and/or presence of edema. In the Dossetor et al. (61) study, it must be noted that all the malnourished children had had a measles rash and were not recovering normally, whereas all the well-nourished children had recovered normally from measles. All children in both studies received a single dose of killed vaccine against Salmonella typhi and serum antibodies to S. typhi O antigen were measured by bacterial agglutination using a microtiter system. The findings indicated that the increase in antibody titers after vaccination was similar in the malnourished and control groups. In the Greenwood et al. (36) study, it was additionally reported that antibody responses to the vaccine were not associated with prealbumin and albumin levels in children. The other studies we identified, 2 observational studies and 2 clinical trials, had smaller sample sizes (<16 participants per group) and produced conflicting findings. The 2 observational studies observed that specific typhoid antibody levels measured 8–10 d after vaccination were significantly depressed in malnourished children compared with nonmalnourished children (62,63). However, one of these studies additionally reported that this difference was no longer significant when antibody levels were measured 25 d after vaccination (63). The randomized controlled trial (RCT) conducted in South Africa showed that children with kwashiorkor receiving protein and multivitamin supplements, with or without pyridoxine (vitamin B-6), and well-nourished children had similar anti-O and anti-H levels after vaccination with the killed typhoid vaccine (64). In India, 32 young children suffering from kwashiorkor received a diet that provided either 50 or 30 g protein/d. The results suggest that antibody levels to O antigen after a primary and booster of killed typhoid vaccine were significantly higher in children who received the 50-g protein diet compared with those who received 30 g. However, from the paper, it is not clear whether children were randomly assigned to the different diets (65). Another clinical trial conducted in a school in New Guinea used a vaccine prepared with flagellin from Salmonella adelaide (66). Children aged 9–11 y with retarded growth were randomly assigned to receive either 25 g/d protein supplement for 5 d/wk in the form of skim-milk powder or the usual school diet that provided 8–10 g/d protein. The flagellin vaccine was injected subcutaneously in all children after 7 mo of supplementation. Serum specific antibodies to flagellin (total and IgG) were measured by hemagglutination before immunization and at 2 and 6 wk thereafter. Serum antibody titers reached a peak at 2 wk and fell slightly at 6 wk in both groups. At both time points, mean total antibody titers were slightly, but significantly, higher in the supplemented group than in the control group. However, specific IgG antibody levels were not significantly different between the groups at either time point.

Polysaccharide vaccines

    Meningococcal vaccine. The few observational studies available on the effect of PEM on immune response to meningococcal vaccine are conflicting. Two observational studies in Nigerian infants <32 mo old (61) and children >4 y old (67) have shown that the increases in antibody titers against meningococcal polysaccharide groups A and C were significantly lower in malnourished children than in well-nourished controls. However, 2 other studies also conducted in Nigerian children found no correlation between nutritional status and the antibody response to the groups A and C meningococcal vaccine (36,68).

    Pneumococcal vaccine. Two observational studies have examined the interaction between malnutrition and response to pneumococcal vaccination in children in Ghana and The Gambia. In Ghana, 115 children between 6 mo and 7 y old were classified as severely malnourished, moderately malnourished, or well nourished according to their weight and serum albumin level (29). They were all vaccinated with pneumococcal polysaccharide vaccine and visited 2 wk thereafter for measurement of antibody response. All malnourished children received appropriate dietary therapy over this 2-wk period. Response rates (nonsignificant increase in specific antibody titers compared with baseline) and mean specific antibody titers did not differ among the 3 groups. In the Gambian study, the 23-valent pneumococcal capsular polysacharide vaccine was administered to 472 children aged 6.5–9.5 y old who were suffering from moderate to moderately severe PEM. Specific IgG antibody response was tested 30 d thereafter using ELISA. The response to serotypes 1, 5, 14, and 23 of the vaccine were not significantly related to any of the nutritional status indicators (3). The antibody response was not associated with plasma leptin level either, whereas leptin level was correlated to children's body fat (60).

Recombinant vaccines

    Hepatitis B. We found only 1 study investigating the relationship between malnutrition and hepatitis B vaccine response. This observational study was conducted in Egypt among 27 infants suffering from marasmic-kwashiorkor or marasmus malnutrition and 13 healthy children used as controls (69). All these children received 3 doses of hepatitis B recombinant vaccine at 0, 1, and 6 mo of age. At mo 8, the geometric mean titer (GMT) of antibodies to hepatitis B was significantly higher in the control group than in the undernourished group, but the 2 groups of malnourished children did not differ. However, the seroprotection rate did not differ among the 3 groups.

Summary of findings for protein-energy

As summarized in Table 4, cross-sectional association studies indicate that PEM seems to have a negative effect on cell-mediated immunity (tuberculin DTH) after BCG vaccination, but there have been no randomized trials. Malnutrition has no detectable effect on antibody responses to other vaccines, including measles, poliomyelitis, tetanus, and diphtheria. The only study that assessed antibody affinity to tetanus toxoid found lower affinity in malnourished children. It must be noted that this study was conducted by Chandra et al. and results must therefore be taken with caution.¹² Findings on other vaccines were not always consistent, especially regarding a possible negative effect of PEM on meningococcal vaccine. However, the data comes almost exclusively from observational studies (only 3 RCT identified for all vaccines). Cell-mediated and humoral immune responses to antigens therefore seem to have different sensitivity to PEM. Similar findings have been shown in animal studies, which have in addition demonstrated that when antibody response was depressed it was due to the impairment of T cells rather than a direct effect on B cells.

    VA and immunity. VA is one of the most widely studied nutrients in relation to immune functions and several very good reviews are available (70–72). Immune functions affected by VA deficiency are summarized in Table 3. There is now substantive evidence emerging from animal and in vitro studies showing that VA and its metabolites have a powerful role in the regulation of immune responses (73–75). In particular, it has been shown that specific subsets of intestinal dendritic cells and macrophages are able to convert VA into retinoic acid (RA) and that RA then enhances the induction of FoxP3+ T regulatory cells (Tregs). In addition, Tregs induced by RA have a unique and highly specific tropism to the small intestine (73–75). Several studies have demonstrated that RA promotes Th2 and inhibits Th1 immune response pathways (71). Accumulating evidence now shows that a high concentration of RA can also suppress the generation of Th17 cells [for reviews, see (72,73)]. Conversely, a recent study reported that RA at very low concentrations promotes Th17 (76). RA may also regulate B cell proliferation and differentiation (77).

VA and vaccine responses

Details of the studies reviewed can be found in Appendix 2.

    BCG. The effect of VA administered at birth along with BCG vaccine has been studied in a large (n = 1894 and 1426 in final analyses) randomized, placebo-controlled trial in Guinea-Bissau (78). Newborns were given either 15 mg retinol palmitate or a placebo and responses to vaccination were assessed at 2 and 6 mo of age by measuring the prevalence and size of the scar induced by BCG vaccination and DTH after intradermal injection of 0.1 mL tuberculin PPD. All analyses were performed with both sexes combined and separately for each sex. At 2 and 6 mo of age, the PPD response and scar generated by BCG did not differ between the VA and placebo groups. However, the authors reported that the percentage of PPD responders (defined as a skin reaction to PPD of >1 x 1 mm) was significantly lower in 2-mo-old boys who received VA than in placebo recipients. No such difference was observed in 6–mo-old boys or in girls at any age.

In Malawi, a cross-sectional study was performed in children <13 y of age reported to have received BCG vaccine within their first 3 d of life. The serum retinol concentration was measured at the time of follow-up and children were classified as VA severely deficient (serum retinol <0.35 µmol/L), VA deficient (serum retinol <0.70 µmol/L), or normal (serum retinol 0.70 µmol/L) (79). The authors reported that significantly more children with VA deficiency had a visible scar generated by BCG than children with normal VA status.

    Measles Vaccination at 6 mo of age. The effects of VA on antibody response to measles vaccination have been investigated in 9 studies. The only observational study identified did not show any association between serum VA levels at the time of vaccination or 3 mo postvaccination and specific antibody response to measles vaccination in 0- to 24-mo-old infants (80). Also identified were 8 clinical trials that assessed the effect of VAS at the time of measles immunization on the antibody response. An RCT in Indonesia in which infants received the Schwarz MV and either the supplement [30 mg retinol equivalent (RE)] or a placebo at 6 mo of age reported that supplemented infants had lower seroconversion rates at age 12 mo, but only in infants with high levels of specific antibodies before vaccination (titers >8) (81). It was suggested that the combined immune effects of VA and circulating maternal antibodies could neutralize the viral antigens contained in the vaccine, therefore leading to a reduced production of protective antibodies by the vaccinee. This hypothesis was also supported by the fact that significantly fewer supplemented infants developed a rash after vaccination than the other infants. In contrast, a similar RCT in Guinea-Bissau did not reveal such a negative effect of VA on seroconversion to MV vaccine, independently of maternal antibody concentrations at baseline (82). In this study, the Schwarz MV was administered at 6 mo of age and seronconversion was assessed 3 mo thereafter. The authors suggested that differences in findings between the 2 countries could be due to the prevalence of VA deficiency before supplementation, which was likely to be higher in Indonesia than in Guinea-Bissau. In the Guinea-Bissau study, infants received an additional dose of MV and VA supplement at the age of 9 mo. Seroconversion rates and mean antibody titers did not differ between the VA and placebo groups at 18 mo (83). Infants were then followed-up until they reached 6–8 y of age (84). This follow-up revealed that VAS at age 6 and 9 mo had no long-term effect on the percentage of children with protective titers.

Vaccination at 9 mo of age. Other studies have assessed the effect of VA administered with MV at the age of 9 mo only. The Guinea-Bissau study described above encompassed an arm in which infants received MV and VAS at 9 mo of age only. For those infants, seroconversion rates were not significantly different in the VA and placebo groups. However, the anti-measles GMT was significantly higher in children supplemented with VA than in children who received a placebo, particularly in boys (83). In 6–8-y olds, VAS at 9 mo of age was associated with higher GMT and with higher protection rates against measles (84). In India, the administration of VA to 9-mo-old infants when vaccinated with the Edmonston Zagreb strain MV resulted in significantly higher seroconversion rates 1 mo after vaccination compared with children who were not supplemented, independently of the initial serum retinol levels (85). Similar results were found in children having low maternal antibodies at baseline; the VA impact could not be assessed in children with high maternal antibodies, because the number of such children was too small (n = 5). However, 3 other RCT conducted in India and Indonesia did not show any effect of VA on seroconversion to measles vaccination given at 9 mo of age; these seroconversion rates were globally high in both the supplemented and control groups (86–88). Moreover, a clinical trial in Bangladesh assessed the effect of VA supplements on MV response when administered near to vaccination or not. VAS administered within 4 wk before or 2 wk after measles vaccination was not associated with altered vaccine failure rate compared with administration of VA outside these time limits (89).

As already noted, contradictory results from these studies may be explained by various factors, including preintervention VA status, but also age at vaccination and supplementation and hence presence/absence of maternal antibodies, vaccine strain, and serological assays used to detect measles antibodies [plaque reduction neutralization (PRN) assay, hemagglutination inhibition (HAI) assay, or ELISA]. Our meta-analysis of these studies (81,83,85–89) revealed that VAS did not affect MV response when given at 6 or 9 mo of age [odds ratio (OR) (95% CI) = 1.08 (0.81–1.45); P = 0.59] (Fig. 2). However, the tests for heterogeneity were borderline significant in this analysis (heterogeneity test I² = 45% and P-value for chi-square test = 0.09). We performed other meta-analyses after grouping studies that used similar serological assays to detect measles antibodies. Three studies used the PRN assay, considered as the gold standard, 2 studies used the HAI assay, 1 study used ELISA, and the last study did not provide the method they used. Both PRN and HAI assays test for specific inactivation of measles virus virulence, whereas the ELISA assay does not. When the HAI assay was used, VAS was associated with a higher seroconversion rate to MV [OR (95% CI) = 2.70 (1.28–5.68); P = 0.009], whereas no effect was observed when PRN or ELISA methods were used (Fig. 3).

View larger version (20K): [in this window] [in a new window]   FIGURE 2  Meta-analysis of odds ratio of seroconversion to measles in children receiving VA supplementation or a placebo.

View larger version (29K): [in this window] [in a new window]   FIGURE 3  Meta-analysis of odds ratio of seroconversion to measles assessed by different assays in children receiving VA supplementation or a placebo.

View larger version (23K): [in this window] [in a new window]   FIGURE 4  Meta-analysis of odds ratio of seroconversion to measles in children with high or low baseline antibodies receiving VA supplementation or a placebo.

View larger version (32K): [in this window] [in a new window]   FIGURE 5  Meta-analysis of odds ratio of seroconversion to each of the 3 poliovirus types in children receiving VA supplementation or a placebo.

View larger version (27K): [in this window] [in a new window]   FIGURE 6  Meta-analysis of odds ratio of seroprotection to each of the 3 poliovirus types in children receiving V A supplementation or a placebo.

    Diphtheria, tetanus, and pertussis. Two small observational studies (n = 20 and 22) in Indian children found no association between VA deficiency (serum retinol level < 0.70 µmol/L) and antibody response to diphtheria and tetanus toxoids assessed 2 or 4 wk after vaccination, respectively (95,96). A larger observational study conducted in 1055 Ecuadorian children also found no association (57).

The effect of VAS on antibody response to diphtheria and/or tetanus toxoids has been investigated in 8 clinical trials (91,95,97–102). Doses and schedules of VA supplements, target population, and study designs (randomization, use of placebo, etc.) varied considerably across these studies.

In a clinical trial in India, 49 children between 1 and 6 y of age received either 30 or 60 mg RE VA at the time of diphtheria and tetanus immunization (95). The intervention was not placebo controlled and the randomization procedure was not clearly defined. Analyses were also limited because comparisons between the supplemented groups and the control group were not presented. However, although the administration of 30 compared with 60 mg RE VAS resulted in similar antibody levels to diphtheria and tetanus toxoids, these levels seemed to be higher than those observed in the control group.

In Bangladesh, a RCT conducted among 1- to 6-y-old children who received 60 mg RE VA with a first injection of tetanus vaccine did not show any effect of VA on antibody response assessed 4 wk after the first and second injections (97). Age and sex were balanced by the randomization between groups, but the control arm did not receive a placebo; the method used to detect antibodies was a mouse-protection assay.

In Indonesia, the antibody response to tetanus in preschool children with mild VA deficiency was assessed in a randomized, double-blind, placebo-controlled trial (100). In this study, 236 clinically normal and xerophthalmic children received either 60 mg RE VA or a placebo 2 wk before immunization with DTP and OPV vaccines. Specific IgG subclass responses were available in 139 children (103). In tetanus-naïve children, those who received VA had significantly higher total IgG and IgG1 titers against tetanus toxoid than children who received placebo, regardless of the initial VA status of participants. In contrast, VAS did not affect the other subclasses of Ig (IgG2, IgG3). In children who had been previously vaccinated with tetanus toxoid, VAS resulted in higher IgG1 and IgG3 memory responses to tetanus toxoid compared with the placebo group.

These results did not agree with those from 3 subsequent trials conducted in younger children (<18 mo old). In Bangladesh, the administration of 15 mg RE VA with each of the 3 doses of DTP vaccine in infants <6 mo old did not result in different anti-tetanus antibody levels compared with the placebo group (102). Similarly, a trial in Turkey where infants were given VA alone, vitamin E alone, vitamins A and E, or placebo did not find any difference in GMT of serum tetanus antitoxin or seroprotection rates between the 4 groups before and after the 3 doses of DTP vaccine (99). However, it seemed that the specific antibody response was higher in children who received VA (with or without vitamin E) compared with the other arms of the trial, but this comparison was not statistically tested. In addition, the sample sizes from these 2 studies were small (n = 33 and 24 children per arm, respectively). A more recent randomized, double-blind, placebo-controlled trial was carried out in Ghana with a much larger sample of 850 infants aged 0–6 mo (91). VA supplements were administered to either infants alone, breast-feeding mothers alone, or both. A 4th group of infants and their mothers received a placebo. Approximately 3 mo after tetanus vaccination, the authors observed no difference in tetanus antitoxin levels among the 4 groups, or in antibody affinity, measured in a subgroup of 259 infants. At 6 mo postimmunization, all infants from all groups had reached a seroprotective level against tetanus.

Contradictory results between the Indonesian trial and other trials may be explained by the difference in the age of the volunteers studied. The Turkish, Bangladeshi, and Ghanaian trials enrolled very young infants in whom the antibody response may have been affected by passive immunity acquired from their mothers. Another major difference was the VAS schedule. VA supplements were administered 2 wk before immunization in the Indonesian study but simultaneously with each of the 3 doses of the vaccine in the other studies. Those differences, in addition to other differences in study designs, outcome measurement, and/or outcome expression prevented pooling these datasets in a meta-analysis.

An additional clinical trial (randomization not specified) conducted in young American adults found that administration of VA over a 4-mo period did not have any effect on any characteristic of tetanus antitoxin response (antibody increase, time to peak, or antibody levels postimmunization) compared with the control group (no placebo) (101). A RCT investigating the association between VA and diphtheria antitoxin response has been conducted in HIV-infected drug users but will not be detailed in this paper (98).

The trial from Bangladesh described previously was the only one that also assessed immune responses to diphtheria and pertussis vaccination (102). VAS did not affect antibody titers to pertussis toxin. However, antibody response to diphtheria antigen was slightly, but significantly, more enhanced in individuals who received VA than in those who did not (P = 0.04; results were only presented graphically). This positive effect remained significant after adjustment for potential confounding factors.

    Cholera. The effect of VA on the killed, B-subunit whole-cell oral cholera vaccine response was investigated in a trial in Bangladeshi children with VA deficiency (serum retinol level < 0.70 µmol/L) (104). This trial had 4 arms in which children were given VA alone, zinc alone, VA and zinc, or placebo. Vibriocidal antibody titers, an indirect marker of protection against cholera, were measured in each group using V. cholerae O1, El Tor, Ogawa as the target organism. Vaccine response (defined as a 4-fold increase in vibriocidal antibody titers compared with baseline) did not differ in individuals who received VA and those who did not. In this study, the highest seroconversion rate was observed in the group that received both VA and zinc, suggesting a synergic effect of those nutrients. Further analyses confirmed that VAS had no effect on the level of cholera toxin specific IgA and IgG levels (105).

    Rabies. The effect of VAS on rabies specific antibody responses has been examined in a small RCT in which 20 Pakistani adults received 60 mg RE VA and 5 doses of rabies vaccine, whereas the 20 other adults, matched for serum VA level, BMI, age, sex, and socioeconomic status, received the vaccine only (106). Thirty days after the last injection, serum anti-rabies antibody titers were significantly higher in the VA group than in the control group, regardless of the VA status at baseline. A larger observational study in 472 Gambian children reported no correlation between serum retinol status and antibody response to 1 or 2 doses of rabies vaccine (3).

    Influenza. A RCT conducted in 59 HIV-infected children found no effect of VAS (60 mg RE) on antibody response to inactivated influenza vaccine (107). An observational study among 27 young American adults and 61 institutionalized elderly did not find any correlation between VA status and antibody response to trivalent influenza vaccine (108).

Recombinant vaccines

    Hepatitis B. In Ghana, a large randomized, nonplacebo–controlled trial investigated the effect of VAS on immune response to recombinant hepatitis B vaccine in 0- to 18-wk-old infants (109). The infants from the supplemented group received 15 mg RE VA along with the 3 doses of the vaccine (6, 10, and 14 wk of age) and mothers from the supplemented and control groups received 120 mg RE just after delivery. Four weeks after the 3rd vaccine and vitamin administration, slightly more infants in the VA group (93.9%) had seroprotective levels of antibodies than in the control group (90.2%) (P = 0.04).

Conjugate vaccines

    Hib. The Ghanaian study described above also assessed the effect of VAS on conjugate Hib vaccine response in infants (109). VA did not increase infant responses to Hib vaccine, which were high in both supplemented and control groups. The interactions between VA and Hib vaccine specific responses have also been addressed in a trial conducted in HIV-infected drug users; antibody titers did not differ between the group supplemented with VA and the group not supplemented with VA (98).

Polysaccharide vaccines

    Pneumococcus. An observational Gambian study did not show any association between response to the 23-valent pneumococcal vaccine (serotypes 1, 5, 14, and 23) and retinol status of children aged 7–9 y (3). The study in HIV-infected drug users mentioned above furthermore investigated the interactions between VA and pneumococcal vaccine (both polysaccharide and conjugate vaccines) (98). Antibody titers after vaccination did not differ between the group supplemented with VA and the group not supplemented with VA, except for the serotype 23F (with the conjugate vaccine), which was slightly, but significantly, higher in the group that received VA.

Summary of findings for VA

VAS might be associated with a small decrease in response to the tuberculin skin test after BCG vaccination (Table 5). Regarding MV, although mean antibody titers tended to be higher in participants who received VA, seronconversion rates were universally high and did not significantly differ between supplemented and unsupplemented participants. One study indicated that VAS might be associated with a reduction in seroconversion only in children with high maternal antibodies, suggesting that the effect of VAS might be highly dependent on age at vaccination. On balance, the data suggested that VA has no effect on seroconversion to any of the 3 serotypes of poliomyelitis vaccine, although 1 study has shown a positive effect on seroprotection to poliovirus type 1 only (and this may be a multiple testing artifact). VA may also have a positive effect on hepatitis B and rabies response, but additional studies would be required to confirm these findings. In contrast, VA has no discernable effect on other vaccines, including diphtheria and tetanus, cholera, influenza, Hib, and pneumococcal vaccines.

VAS in children >6 mo of age is recognized to be a cost-effective intervention that is associated with a significant decrease (23–30%) in all-cause childhood mortality in low-income countries (110,111). For reasons of convenience, WHO therefore recommends that VAS should be given at the time of contact with children for immunization after 6 mo of age. Postnatal VAS at younger ages has provided more varied and complex results. Clinical trials from India, Indonesia, and Bangladesh have reported significantly reduced mortality in the first months of life when neonates were supplemented with VA (112–114). However, this beneficial effect has not been shown in Africa. Studies from Guinea-Bissau and Zimbabwe have found no overall effect of VAS given at or just after birth on infant survival (115,116). In addition, other studies where children were supplemented between 1 and 5 mo of age did not show any positive effect on overall mortality (117–119) or even showed an increased mortality risk in infants who received VA compared with those who did not (8). These age-pattern variations have been ascribed to breast-feeding, which would cover the VA needs in younger infants, whereas older children with little or no breast-feeding would benefit more from VAS (117). Alternatively, Aaby et al. (8) team working in Guinea-Bissau have suggested that VAS could interact with the vaccines by amplifying their nonspecific effects. According to this thesis, VAS would have beneficial effects when administered with live vaccines such as BCG at birth or measles after 6 of age but no effect or even negative effects when administered with inactivated DTP vaccine between the age of 1 and 5 mo (5,6,8). The section below reviews the existing evidence for or against this hypothesis, as well as other possible adverse events related to the coadministration of VA and a vaccine. The studies reviewed are presented in Appendix 3.

    Mortality. Only 1 randomized placebo-controlled trial has specifically investigated the impact of high-dose VA given with BCG vaccine at birth on infant mortality (115). In this study in Guinea-Bissau, 4345 infants were randomized to receive either 15 mg RE VA or a placebo when immunized with BCG vaccine and were then followed until age 12 mo. VAS was not associated with survival during the first year of life, the overall mortality ratio being 1.07 (95% CI = 0.79–1.44; P = 0.66). The authors observed that there was a tendency for a sex differential effect of VAS (P for interaction = 0.10), with mortality rate ratios tending to be reduced in boys [0.84 (95% CI = 0.55–1.27)] and increased in girls [1.39 (95% CI = 0.90–2.14)]. Actual P-values were not provided by the authors but the relative risk included 1 for both sexes. The other studies mentioned above assessing the effect of VAS at the time window of BCG vaccination could not be considered here, because information on vaccination status was not provided (112–114,116).

    Other adverse events. Possible adverse events caused by the coadministration of VA supplements and BCG vaccine at birth were also addressed in 2000 infants from the Guinea-Bissau study (120). During the first month after supplementation, hospitalization rate, bulging fontanelle, irritability, fever, and vomiting did not differ between the supplemented and the placebo groups. The only difference found was a slightly lower risk of diarrhea in the VA group than in the placebo group, observed within 3 d postsupplementation [18.7 vs. 24.4%; relative risk (95% CI) = 0.77 (0.61–0.96)].

VA and DTP/OPV vaccines

    Mortality. In routine immunization, OPV and DTP vaccines are generally administered together, which makes it difficult to separate potential effects of these vaccines. However, we identified 1 RCT that examined the effects of different doses of VA administered with OPV on childhood morbidity and mortality (6). This large study enrolled 4983 children aged 6 mo to 5 y living in Guinea-Bissau. They were given either the dose of VA recommended by WHO (30 mg RE at age 6–11 mo and 60 mg RE at age 12 mo every 3–6 mo) or one-half this dose along with OPV, and mortality and morbidity were assessed after 6 and 9 mo. Although overall mortality rate did not differ between the different dosing of VA, the authors reported that the lower dose was associated with significantly reduced mortality in girls only (mortality rate ratio, 95% CI = 0.19, 0.06–0.66). WHO coordinated a multicenter randomized trial in Ghana, India, and Peru to assess the safety of combined maternal postpartum and young infants' VAS administered with DTP/OPV immunizations. In the VA group, mothers received 60 mg RE VA and their infants received 7.5 mg RE VA with each of the 3 doses of DTP/OPV vaccine, whereas both mothers and infants received placebos in the other group (119). The results indicated mortality did not differ between the 2 groups during the first 9 mo of life; the VA:placebo group mortality rate ratio was 0.96 (95% CI = 0.73–1.27) for this period. In Bangladesh, 200 infants aged 6–17 wk received randomly either 7.5 mg RE VA or a placebo with each of the 3 doses of DTP/OPV vaccine (118). The mortality rate was similar in the 2 groups and all deaths were attributable to diarrhea. Another study conducted in Bangladesh enrolled 200 children aged 1–6 mo who had diarrhea and who sought treatment. These children received either 15 mg RE VA or a placebo with each of the 3 doses of DTP vaccine (121). Mortality rates did not differ between the 2 groups. Finally, a recent study conducted in Ghana randomly assigned 1077 children to receive or not 15 mg RE VA with the pentavalant vaccine DTP-Hib-HepB at 6, 10, and 14 wk of age (122). The authors reported a total of 6 deaths, all in the intervention group except one, resulting in a risk ratio of 4.6 (95% CI = 0.5–40.0; P = 0.12). Combining results from these studies, the effect of receiving VAS with DTP vaccine on subsequent mortality was 1.05 (95% CI = 0.82–1.36) (Fig. 7).

View larger version (20K): [in this window] [in a new window]   FIGURE 7  Meta-analysis of odds ratio of mortality in children receiving VA supplementation or a placebo with DTP vaccine.

    Other adverse events. The RCT in Guinea-Bissau that examined different doses of VA with OPV also reported results on children's morbidity (6). The lower dose of VA was associated with a lower hospital case fatality in girls only but with a slightly higher morbidity (diarrhea, fever) in children aged 6–18 mo in either sex. The WHO multicenter study described previously did not show any significant differences in morbidity, including acute lower respiratory infections and diarrhea, between the group that received VA and the group that did not (119). Regarding side-effects, an increase in bulging fontanelle was observed after each dose of VAS in the WHO study; however, <1% in each group presented with a bulging fontanelle and the condition resolved spontaneously. The Bangladeshi RCT, also described above, where children received 7.5 mg RE VA or a placebo with each of the 3 doses of DTP/OPV vaccine at 6, 10, and 14 wk of age, investigated the effect of VAS on adverse events other than mortality (118). This study showed that decreased feeding, irritability, diarrhea, and vomiting events were comparable in the 2 groups. However, significantly more children receiving VA developed bulging fontanelle than those who received placebo, although this generally disappeared spontaneously within 48 h. In the study from Mahalanabis et al. (121) conducted in Bangladesh, vomiting and diarrhea did not differ between the treatment groups, but more infants in the VA group had a bulging fontanelle than in the placebo group [relative risk (95% CI) = 4.78 (1.06–21.5); P < 0.02], mostly after the second and 3rd doses of VA. In Ghana, Newton et al. (122) did not report higher bulging fontanelle in children supplemented with 3 x 15 mg RE VA along with a DTP-Hib-HepB vaccine than in nonsupplemented children. However, VAS was associated with a significantly lower occurrence of any illnesses and fever. Three other studies conducted in Bangladesh and Tanzania evaluated adverse events, but not mortality, after VA and DTP/OPV administrations. They all reported a significantly higher risk of bulging fontanelle, but no other serious side-effects, in children who received VA compared with those who did not (126–128). In one of the Bangladeshi studies, some children were revisited when they reached the age of 3 y for assessment of their motor and neurological development. This analysis revealed that bulging fontanelle was not associated with long-term physical and developmental abnormalities (129). In contrast, an Indonesian study that administered VA or placebo with the 3 doses of DTP/OPV vaccine and with MV found only 1 case of bulging fontanelle from the placebo group (130). This study also reported significantly higher prevalence of vomiting in the placebo group than in the VA group. Three RCT in which the primary outcome was the effect of VAS on antibody response to OPV and/or DTP vaccines also assessed possible adverse events (93,94,99). They did not observe any difference in adverse events between the supplemented and control recipients. Finally, an observational study reported that 12- to 48-mo-old children who had been supplemented with 60 mg RE VA with OPV as part of the immunization campaign were significantly less likely to have diarrhea than children who did not receive VA (131).

An interesting RCT conduted in 197 Gambian mother-infant pairs investigated the effect of an early, high-dose regimen of VAS on various health outcomes, including Helicobater pylori infection, nasopharyngeal pneumococcal carriage, gut epithelial integrity, bulging fontanelle, and general morbidity (diarrhea, vomiting, fever, and cough) (132). In 1 group, mothers were supplemented with 120 mg RE VA within 24 h of delivery and their infants received 15 mg with the 3 doses of DTP at 6, 10, and 14 wk of age and 30 mg at 9 mo of age. Mothers in the other group received 60 mg RE VA within 24 h of delivery and their infants received placebo with the 3 doses of DTP at 6, 10, and 14 wk of age and 30 mg RE VA at 9 mo of age. The study showed no differential beneficial effect of the high-dose VA on health outcomes. If anything, it rather indicated a detrimental effect of increased clinic attendances and poor gut integrity. None of the VA regimen was associated with bulging fontanelle, nausea, or irritability.

VA and MV

    Mortality. VAS given with MV is thought to have no adverse effect, or even possibly a slight beneficial effect, based on observations that measles vaccination given after 6 mo of age was associated with a greater-than–expected reduction in all-cause mortality (8,133,134). However, no study has prospectively assessed the impact of coadministration of VA and MV on child mortality. Benn et al. (8) have reanalyzed data from one of their previous trials, in which children were randomized to measles or inactivated polio vaccine and to VA or placebo to investigate the effect of VAS on the immune response to MV. By the age of 18 mo, the 227 children who had received VA and MV had a mortality ratio of 0.46 (95% CI = 0.14–1.47) compared with the 235 children who had received placebo and MV. The authors also claimed that the effect of VA was different between the MV group and polio vaccine groups. However, the number of deaths involved was very small. Their analysis was based on the observation that in the MV group, 3 deaths occurred in the placebo group, whereas there were no deaths in the VA group. In the polio vaccine group, all 3 deaths occurred in the VA group, whereas there were no deaths in the placebo group.

    Other adverse events. Some studies have specifically evaluated the adverse events that might be caused by the simultaneous administration of VA and MV. In Guinea-Bissau, the administration of 30 mg RE VA at the age of 6 and 9 mo with MV did not cause any bulging fontanelle, as in the placebo group (135). An Indian study reported similar results after administration of 30 mg RE VA with MV in 9- to 12-mo-old children (136). In addition, this study found no difference in other adverse events, including vomiting, irritability, fever, and loose motions, between the supplemented and placebo groups. Two studies, initially designed to assess the effect of VAS on MV efficacy, reported that possible adverse events did not differ between intervention groups (86,88).

Summary of findings

VAS along with vaccination is associated with a higher risk of bulging fontanelle, which disappears rapidly and does not have short- or long-term health consequences (Table 6). It is also associated with a lower risk of diarrhea, but does not seem to have any effect on other adverse events. There is only weak evidence that VAS with BCG, DTP/OPV, or measles vaccination has an effect on subsequent mortality. The controversy regarding a potential detrimental effect of VA administered with DTP vaccines has led to considerable discussion in the literature and in specially convened meetings (7,125,137–139). The hypothesized negative interaction is biologically plausible, including the male-female difference. However, the evidence produced so far is not abundant and mainly produced by a single research group. Statistical methods adopted by this group to support their hypotheses were sometimes not adequate. The group also places excessive weight on nonsignificant trends. In addition, it is well known that survival analyses may lead to substantial bias depending on the method used to determine vaccination status (140). A meeting was recently held in London to discuss the nonspecific effects of vaccines, including the possible role of VA on these effects, and has led to a first publication on methodological aspects (141). Datasets from several countries have also been gathered and should be reanalyzed to further address this issue.

    Iron and immunity. The interactions between iron and the immune system have been the subject of a number of reviews over the recent years (142–145). Adverse effects of iron deficiency that may potentially affect vaccine responses include a decrease in circulating T cell number and in vitro proliferative responses to mitogens and impairement of IL-2 production (146–148). In humans, iron supplementation generally improves hematological status, but effects on cell-mediated immunity are less clear. Some studies showed that iron supplementation had a beneficial effect on T lymphocyte percentage (147,149–151); however, some of them were not placebo controlled, whereas other studies showed no effects (152). Discrepancies in findings may be due to initial iron status and/or to possible infections that alter immune parameters. It is well known that pathogens such as infectious microorganisms and viruses require iron for replication and survival (153,154).

    Iron and vaccine responses. There is limited data on vaccine response in relation to either iron deficiency or iron supplementation in humans; studies that were identified (Appendix 4) are therefore reviewed together rather than by vaccine type. We identified only 1 observational study (155) and 1 trial (156) on iron and influenza vaccine. The trial was conducted among an extremely small number of institutionalized patients (<5 participants/group) vaccinated against influenza, for which the results cannot be interpreted. Several observational studies have compared responses to diphtheria and tetanus toxoids between individuals with differing iron status. Two of them did not find significant differences in antibody titers to either diphtheria or tetanus toxoids between iron-deficient anemic children and nonanemic children (151,157). However, the sample sizes were small in both studies (n = 38 and 40), which makes the results difficult to interpret. In contrast, in a study involving 1554 Ecuadorian children aged 0–59 mo, anemic children >1 y had significantly lower antibody titers to tetanus and diphtheria toxoids after vaccination with DTP vaccine than did control children (57). In addition, a higher proportion of anemic children had antibody levels below the protective level for diphtheria antitoxin. This result was not found when data were stratified into 1-y age group intervals, which may be explained by the relatively small number of children within each stratum (n 30). All children < 1 y old were excluded from the analyses to avoid the influence of maternal antibodies. Other observational studies have investigated this issue but were excluded because of very small sample size (<10 participants) (158) or inappropriate measurement of immune response (159).

Summary of findings for iron

Data available on iron and vaccine response are limited and no clinical trial with acceptable quality has been identified (Table 7). From these data, children with iron deficiency anemia seem to have intact antibody responses to vaccination. Although iron deficiency is known to affect T lymphocytes, the antibody response is preserved, even when it requires help from Th cells. Animal studies differ from human studies in that they show a clear impairment of antibody-mediated immunity in iron-deficient animals (160–162).

    Zinc and immunity. Zinc is known to play a central role in multiple aspects of the immune system. This role has been extensively studied and is well reported in a number of reviews (163–167). The bulk of the evidence comes from animal experiments, although substantial human studies are now confirming some of the work from animals. Zinc deficiency has been associated with thymic atrophy and a decrease in thymulin activity, which have direct adverse consequences on the development, differentiation, and function of T cells (168,169). Zinc deficiency has also been shown to modify the Th1/Th2 balance by decreasing the Th1 pathway and its associated cytokines (e.g. IL-2 and interferon-). B cell development and antibody production also seem to be affected, albeit to a lesser extent than T lymphocytes (165). Conversely, zinc supplementation has been linked with an increase in T cell development and functions, an increase in antibody production, and a boost in delayed-typed hypersensitivity reactions in mice and pigs (170,171). Excess of zinc intake may compromise Th1 function (172). The effect of zinc deficiency and/or supplementation on Th17 or Treg cells is unknown.

    Zinc and vaccine responses. The literature provides much more data regarding the effect of zinc deficiency or supplementation on vaccine responses (Appendix 5).

One RCT assessed the effect of zinc supplementation in women during pregnancy on the immune response to BCG vaccines in their infants. In this study conducted in Bangladesh, 559 pregnant women were randomized to receive either 30 mg/d zinc acetate or a placebo from 12–16 wk of gestation and until delivery (173). The newborns were vaccinated with BCG within 72 h of birth and followed to 24 wk of age when the skin test was performed (using 0.1 mL PPD). The authors reported no differences in the proportion of infants with a negative response (size of induration <5 mm) or in the mean size of induration to the skin test between the zinc-supplemented and placebo groups. Similar results were observed when normal-weight and low-birth weight babies were considered separately. Although it was stated that serum zinc was assessed in infants at 4 and 24 wk of age, zinc levels in infants were not presented in the article.

Inactivated vaccines

    DTP. The effect of zinc status on the immune response to diphtheria and tetanus toxoid vaccines was studied in 1554 0- to 59-mo-old Ecuadorian children. In this observational study, serum antibody titers and the proportion of protective antibodies did not differ between zinc-deficient and zinc-sufficient children for either toxoids (57). In healthy elderly (>70 y) people, however, supplementation with 220 mg zinc sulfate twice per day for 1 mo was associated with a higher IgG antibody response to tetanus toxoid vaccine, as indicated by a higher proportion of responders and elevated antibody titers in the supplemented compared with the placebo group (174).

Other studies have investigated the association between zinc and diphtheria vaccine response in a very small sample of chronic hemodialysis patients (n = 16) (175) and in HIV-positive drug users (98). Zinc levels or supplementation were found to have no impact on antibody response to diphtheria toxoid.

    Cholera. Two RCT have investigated the effect of zinc supplementation on the immune response to killed, B-subunit whole-cell cholera vaccine. One study was conducted in Bangladesh and enrolled 256 children aged 2–5 y old who were all suffering from VA deficiency (serum retinol level < 0.70 µmol/L) (104). They were randomly assigned to receive VA and zinc (AZ group), VA and a placebo (A group), zinc and a placebo (Z group), or both placebo (P group). The Z and AZ group received 20 mg of zinc daily for 42 d and all groups received 2 doses of oral cholera vaccine administered during a 2-wk interval. Vibriocidal antibody titers were measured at baseline and 1 wk after the first and second doses, while zinc status was measured at baseline and 1 wk after the second dose of vaccine. The proportion of children who seroconverted (4-fold rise in antibody titers from baseline) was higher in the zinc-supplemented groups (Z and AZ groups) than in the groups that did not receive zinc (A and P groups) (55 vs. 39% after the first dose of vaccine, P = 0.013 and 64 vs. 51%, P = 0.048 after the second dose). However, individually, only the AZ group had a significantly higher seroconversion rate than the other groups. This suggests that zinc and VA supplements have a synergistic positive effect on the immune response. Additional analyses were published 1 y later (105). The sera of children were assayed for antibodies to cholera toxin. Surprisingly, children who received zinc had reduced levels of serum antibodies to cholera toxin compared with children who received placebo, in particular cholera toxin specific IgA. Seroconversion rates also tended to be lower in the zinc-supplemented group, although the difference was significant only for cholera toxin IgA in group Z compared with group A (66.7 vs. 85.3%; P = 0.009). The authors suggested that zinc supplementation may induce the epithelial transfer of IgA, thus increasing the cholera toxin IgA response in the mucosal surface where protection is needed. The second RCT that recruited healthy students in Norway supports the results from Bangladesh, despite a very small sample size. In this study, 30 volunteers randomly received either 45 mg of elemental zinc thrice daily for 9 d or nothing with each of the 2 doses of the cholera vaccine, administered during a 17-d interval (176). The increase in serum vibriocidal antibodies from pre- to postvaccination was higher in the zinc-supplemented vaccinees than in the nonsupplemented vaccinees. The increase in serum anti-cholera toxin IgA and IgG was 13-fold lower in the zinc-supplemented group than in the control group, whereas the fecal anti-cholera toxin IgA response tended to be higher in the zinc-supplemented group (4-fold higher increase, P = 0.084). Zinc may therefore have different modulating effects on vibriocidal and cholera toxin antibody responses.

Recently, the research group working in Bangladesh has published results from a trial that was not placebo controlled (177). In this study, 340 Bangladeshi children aged 6–18 mo received 2 doses of inactivated vibrio cholerae O1 bacteria associated with recombinantly produced cholera toxin B-subunit in a 2-wk interval. Among them, 70 were supplemented with zinc acetate every day for 42 d starting 3 wk before administration of the first dose of vaccine. The other group received no supplementation. Each vaccinee served as his own control based on the prevaccination immune status. Results were presented for the younger children (6–9 mo) and older children (10–18 mo) separately. The authors reported a significantly higher vibriocidal response in the zinc group (older children only) compared with the group that received the standard vaccination. Responses to cholera toxin were comparable in the groups with or without zinc supplementation, whatever the age group considered.

    Rabies. In The Gambia, baseline zinc levels had no appreciable influence on response to 1 or 2 doses of rabies vaccine in children between 6.5 and 9.5 y of age (3).

    Influenza. Two observational studies conducted in the US assessed the association between zinc status in young adults and/or the elderly and their immune response to influenza vaccine (108,178). Neither of these found a significant association between serum zinc levels and the antibody response to this vaccine. Similarly, in RCT conducted in Italy, where institutionalized elderly were administered zinc sulfate, with or without arginine, or no supplements for a 60-d period, the mean influenza antibody titers did not differ between the 3 groups (179). In The Netherlands, another a RCT conducted in elderly participants who received 440 mg/d zinc sulfate for 28 d or a placebo with the trivalent influenza vaccine did not show any difference in antibody titers against the 3 strains of the vaccine between the 2 groups (180). In contrast, another RCT conducted in France among long-term institutionalized patients demonstrated that zinc supplementation administered with selenium for 2 y did result in a better antibody response to the 3 serotypes of the influenza vaccine than in the placebo group (seroprotection rate: 44.1 vs. 27.7%; P < 0.05) and in the group who received vitamins C, A, and E (44.1 vs. 12.1%; P < 0.05) (181). Another trial conducted in hemodialysed patients reported that zinc supplementation did not affect antibody response to influenza vaccine (182).

Conjugate vaccines

    Hib. In Bangladesh, zinc supplementation of women during their last 2 trimesters of pregnancy had no influence on antibody response to Hib vaccine in their infants (173). Almost all infants in the zinc and placebo groups achieved protective antibody titers (0.15 mg/L) and a large majority in both groups even achieved long-term protection as defined as titers 1 mg/L. The other clinical trial identified targeted HIV-infected patients and reported no significant difference in antibody titers between individuals supplemented or not with zinc (98).

Pneumococcus

The effect of zinc supplementation on immune response to the heptavalent pneumococcal protein conjugate vaccine (PCV) was evaluated in an RCT in Bangladesh (183). In this study, 241 infants were assigned to either 5 mg of elemental zinc daily from 4 to 33 wk of age or a placebo. All children received 3 doses of PCV. Antibody titers did not differ between the zinc and placebo groups after each dose of the vaccine, except for the serotype 9V, which was higher in the zinc group after the 3rd dose of vaccine (4.09 vs. 3.33 mg/L; P < 0.05; possible multiple testing artifact). However, for all serotypes, seroprotection rates were high and did not differ significantly between the groups. The clinical trial enrolling HIV-infected patients also examined the response to conjugate pneumococcal vaccine but found no significant differences in antibody titers with or without zinc supplementation (98).

Polysaccharide vaccines

An observational study analyzed the correlation between zinc status and pneumococcal vaccine specific responses in Gambian children. Children were given a single injection of the 23-valent pneumococcal capsular polysaccharide vaccine and their zinc status was monitored (3). The children had varying zinc levels, which were divided into quartiles for the analysis. One month after immunization, the specific IgG responses to serotypes 1, 5, 14, and 23F of the pneumococcal vaccine were not significantly related to the children's zinc status.

Recombinant vaccines

    Hepatitis B. Only 1 clinical trial conducted in hemodialysed patients has been identified and reported no significant effect of zinc supplementation on specific antibody titers after vaccination (184).

Summary of findings for zinc

A larger number of studies were available for zinc (Table 8). However, most of them were conducted in particular target groups, such as hemodialysed patients, institutionalized elderly, or HIV-positive drug users, and very few of them concerned children. Globally, zinc deficiency and zinc supplementation have no discernible effect on vaccine responses to BCG, diphtheria and tetanus toxoids, rabies, influenza, and pneumococcal vaccines. However, zinc supplementation may have an effect on vibriocidal antibody responses after vaccination against cholera. Animal studies showed that the T-dependent antibody response was more affected by zinc deficiency than T-independent responses. However, it has also been shown that both responses were impaired by zinc deficiency following subsequent vaccination (e.g. booster), suggesting that immunologic memory is affected by zinc status. This finding is important for the maintenance of vaccine efficacy and has not been explored in humans.

    Vitamin D and immunity. Evidence showing the effects of vitamin D on immune functions has already been described by others (185,186). Recent evidence, essentially from in vitro and animal studies, suggest that the active metabolite of vitamin D, calcitriol or 1,25-dihydroxycholecalciferol, has a strong immunomodulatory role (75). Certain dendritic cells and macrophages have the ability to convert vitamin D into its active metabolite, whereas activated T cells (and possibly activated B cells as well) can perform the final step of converting calcidiol or 25-hydroxycholecalciferol into 1,25-dihydroxycholecalciferol. Vitamin D has well-known inhibitory effects on certain immune cells, including T cell proliferation and CD8 cell cytotoxicity. Both Th1 and Th2 cells are also direct targets of vitamin D. Vitamin D and vitamin D-receptor deficiencies are associated with elevated Th1 responses, whereas vitamin D supplementation results in a shift of the Th1/Th2 balance toward Th2-cell responses (187,188). The active form of vitamin D seems to promote the induction of Foxp3 Treg and T regulatory 1 cells and is therefore implicated in the control of autoimmunity (189). Conversely, vitamin D decreases Th17-cell responses, probably through the inhibition of IL-6 and IL-23 production (190). Vitamin D may also inhibit B-cell proliferation, plasma cell differentiation, and IgG secretion through its effects on antigen-presenting cells and/or Th cells (191). A direct effect of vitamin D on B cells is not confirmed, because results indicating the presence of vitamin D-receptor on B cells are variable [reviewed in (75)].

Vitamin D and vaccine responses

Very few studies have examined the influence of vitamin D or its derivatives on vaccine responses in humans (Appendix 6). Most of these studies have been conducted in immunocompromised patients, in particular hemodialysis patients, because vitamin D deficiency is common in this group. Because this target group is not relevant for the objectives of this article, the findings will not be discussed but are presented in Appendix 6 (192–194).

Summary of findings for vitamin D

The small number of available studies that investigated the associations between vitamin D (and its derivatives) and vaccine response makes it extremely difficult to draw conclusions for this nutrient (Table 9). From this data, there is no evidence that vitamin D has any effect on vaccine response. This conflicts with information available from animal studies that indicates that vitamin D and its derivatives can enhance both mucosal and systemic immune responses to vaccination (196–200).

	Discussion

TOP ABSTRACT Introduction Methods Results Discussion LITERATURE CITED

The basic limitations include the following. Basic knowledge on the effects of macro- and micronutrient deficiencies and/or supplementation on human immune function remains sparse, with many conficting findings. Methods for assessing vaccine efficacy are frequently rudimentary and rely on proxy outcomes that may not correlate with functional protection. Very few studies have adequately examined possible distal and nonspecific effects of vaccination. There is a paucity of studies that adequately examine interactions between nutritional status and/or supplementation and vaccine responses. Apart from protein-energy and VA nutrition, and OPV, BCG, measles, diphtheria, and tetanus vaccines, there are virtually no data and it is not currently possible to reach a consensus conclusion. Even where there are sufficient data to allow meta-analysis, differences in age group, background nutritional and immunological status, and study design prevent secure interpretation. Interpretation is further hampered by the likelihood of positive reporting bias in the published literature and a failure to adjust for multiple testing.

Quantity of available data

The Expert Advisory Board was first surprised by the paucity of published data on the effects of nutritional deficiencies on immune function and specifically on the potential interactions between nutrition and vaccine responses. We found very few studies that examined the effects of iron, zinc, and vitamin D on vaccine responses, whereas data on VA and PEM were more numerous but still not substantial. Regarding vaccines, most of the research has been done on BCG, measles, diphtheria, and tetanus, while studies on other vaccines are sparse. As an example, we found only 1 study on pertussis vaccine, probably because the serological test for this vaccine is complex. Quality of existing studies was often questionable and heterogeneity between these studies hampered comparisons. They often led to contradictory findings, which thus made it difficult to draw strong conclusions and recommendations. Contradictory results were also observed between human and animal studies. The lack of understanding of the mechanisms involved in potential interactions between nutrition and vaccine responses clearly hampers progress in this area.

Quality of available data

We identified a mix of observational studies and randomized trials. Observational studies generally compared vaccine responses in individuals who were nutritionally deficient to individuals who were considered normal. The distinction must be made between longitudinal studies in which the vaccine was administered within the study and cross-sectional studies in which vaccination status was assessed retrospectively (e.g. from vaccination cards) and therefore not always reliably and without the potential of bias. Possible confounding factors, including age, sex, breast-feeding, infections, or antibody levels at baseline were not always well controlled in observational studies. In clinical trials, randomization procedures were not always clearly expressed or were nonexistent. In addition, some trials were not blinded and placebo controlled, mostly due to logistical reasons.

A major limitation in both types of studies was the sample size, which was frequently too small. This leads to a lack of power to detect differences between treatment groups (type 2 errors). In some studies, the number of volunteers recruited was quite large, but the number lost to follow-up was considerable (due to moving from the study area, deaths, and refusals from the population to undergo complete examinations, including several blood samples) and may have introduced bias.

Vaccine outcomes were mainly restricted to measurements of cell-mediated immunity and antibody production following vaccination. These measurements, which represent the magnitude of a vaccine response, are the easiest way to assess the immune response. However, serological assays used to detect antibody levels were not always appropriate to the vaccine considered. Moreover, as methods have obviously changed over time, some assays used in older studies are now out of date. For example, in the 1950s, skin tests were sometimes used to detect diphtheria toxoid, but results from these studies are nowadays not exploitable. The quality of vaccine responses, including antibody affinity, is a crucial factor of vaccine efficacy, but was rarely evaluated in the studies we reviewed. Even studies with measures of affinity may not correlate with functional protection. The technical and logistical complexity of measuring such outcomes likely explains this gap.

Heterogeneity of available data

Another constraint was the heterogeneity existing between studies. Definitions of VA, vitamin D, iron, and zinc deficiency were generally similar between studies. However, the inflammation state affects the assessment of nutrient status, e.g. iron and VA. Thus, studies conducted in populations with different rates of infection, and whether markers of inflammation were measured to control for this or not, might have generated some heterogeneity between studies in the definitions of deficient compared with nondeficient children. Definitions and growth references used for PEM were very diverse. In clinical trials, doses and times of administration of nutritional supplements also varied greatly between studies, thus making the comparisons difficult.

Although vaccine responses were mainly assessed by antibody production before and after vaccination, the outcomes were expressed in various ways between studies. Some studies used seroconversion rates, whereas others used seroprotection rates, with different definitions and cutoffs. Continuous outcomes were expressed either as GMT or variations in titers before and after vaccination, with possible variations in units. This variety of vaccine outcomes was a substantial limitation in performing meta-analyses. Subgroup-comparisons, e.g., between participants with low compared with high antibody at baseline, further increased this heterogeneity, because cutoffs used to determine the subgroups were not always similar. Moreover, postvaccination outcomes were measured at different times between studies, with variations ranging from a few days to several months. This may partly explain the contradictory results found in some studies regarding the effect of nutrition on vaccine responses. The serological assay used to detect antibody levels also differed across studies. Methods such as ELISA, HAI, or PRN assays have different sensitivity and specificity to detect different antibodies. Such variations in studies that examined responses of the same vaccine may also lead to variations in conclusions regarding the effect of nutrition and vaccine response.

Whereas it is important to note the effects of these methodological differences in limiting cross-study comparisons, they would usually not compromise the validity of between-group comparisons within individual studies.

Publication bias

It is recognized that studies that demonstrate an effect (either positive or negative) are more likely to be published than studies that show no effect. When published, it is also more likely that outcomes for which there were null findings received less prominence. Assessment of publication bias can be conducted by performing funnel plots with Review Manager. However, the number of studies in each meta-analysis was too small to assess this bias in our context of interactions between nutrition and vaccine responses.

An additional factor likely to have been present even at the study design stage is the almost universal pre hoc expectation that undernutrition is likely to negatively affect immunity and hence vaccine responses. This may bias interpretations.

Failure to adjust for multiple testing

Many of the published studies have conducted multiple subgroup analyses (e.g. by age group, sex, baseline antibody levels, serotypes, etc.) and have failed to adjust for this by setting more stringent tests of significance. A number (perhaps most) of the marginally significant results reported may be artifacts resulting from multiple testing.

On the whole, our analysis indicates that malnutrition has little or no effect on immune responses to vaccination, at least as measured using currently employed outcome measures. This statement conflicts with the known effects of malnutrition on increasing mortality from a wide variety of infection-related outcomes [e.g. (201)]. There may be several possible explanations for this apparent paradox. Malnutrition may affect morbidity and mortality through mechanisms unrelated to cognate immunity (e.g. through a breakdown in barrier and innate defense mechanisms). Because the focus is mainly on antibody titers, we may also miss some other crucial effects. Examples could include alteration of a particular Ig subclass, affinity of antibody produced, delay in antibody response, whether the T-cell dependent or non-T cell dependent pathway is involved, or changes in mucosal immunity or collateral effects on the overall setting of immune bias (e.g. Th1/Th2 bias, Tregs, etc.). The evidence for any impact of nutritional status on B lymphocytes is weak in humans.

Widely adopted vaccines have, by definition, been selected because they employ highly immunogenic antigens that induce strong immune responses. Vaccine schedules (age group, booster immunization) have been further designed to maximize efficacy. Thus, in many of the studies reviewed, seroprotective levels were reached in almost all individuals irrespective of nutritional status, resulting in little variance. The previous statement is particularly true for live vaccines, such as measles, which are known to induce very strong immune responses. Furthermore, it is possible that live attenuated vaccines may persist for longer in a malnourished child and hence create a longer antigenic exposure than in a well-nourished child, potentially canceling out any immune deficiency. This might explain some of the complex, and apparently contradictory, results in some zinc studies.

The high antibody response to certain vaccines, whatever the nutritional status, may also be ascribed to environmental conditions, in particular background exposure to certain antigens, and these influences might be stronger in malnourished children living in less clean environments. As an example, S. typhi is very cross-reactive with other bacteria such as Escherichia coli or Shigella, which are highly present in environments with poor hygiene. Individuals who have frequent contacts with these pathogens may therefore already have high antibody levels before any vaccination. Besides, malnourished children may be suffering more intercurrent infections or indirect immunostimulation (e.g. by translocation across the gut of bacteria or bacterial toxins) and these may create an adjuvant effect that enhances responses to vaccines. Once again, this might cancel out any tendency toward an inherent immunodeficiency. Evidence in favor of this hypothesis comes from the surprising observations that some vaccine responses are most potent in seasons of poor nutritional status and heavy infectious load (202).

In conclusion, despite decades of research, the effects of nutritional status on vaccine responses in humans remains unclear. Reasons for this include the difficulty in separating effects of malnutrition from those of infection, because the 2 almost always coexist, the use of outdated measures of vaccine efficacy and difficulties of assessing cellular responses in field settings, some concerns over the validity of some of the published data, and a lack of research investment. With the exception of protein-energy, VA, and zinc, the evidence base on nutrient-vaccine interactions is almost nonexistent. Surprisingly, we found little convincing evidence to indicate that current nutritional status or coadministration of nutrient supplements has clinically important effects on vaccine efficacy. However, recent in vitro and animal research has confirmed that VA, vitamin D, and zinc can have potent immunomodulatory effects, suggesting that important interactions that could affect vaccine efficacy may have been overlooked. We found occasional evidence of marginal effects of micronutrients, but the evidence base is very limited and frequently contradictory. An effect of moderate to severe PEM on cell-mediated responses to BCG is the strongest likelihood but remains unproven. Previous studies on the interactions between nutrients and vaccines have employed simple endpoints (most usually antibody titers or DTH tests) that may not correlate strongly with protection and give no indication of other potentially beneficial or detrimental immunological effects (e.g. alterations in T cell bias or resetting of inflammatory responses). Yet, such effects are now amenable to study using modern immunological techniques and investment in such research could yield major advances in understanding the immunomodulatory effects of nutrients on the specific and nonspecific effects of vaccines and hence may have translational impact with respect to current and future vaccine protocols.

Although some vaccines are known to have nonspecific effects that can both positively and negatively affect later mortality, these effects are generally under-researched. Claims of a negative interaction between VAS and DTP vaccines leading to excess mortality in girls therefore warrant serious attention. Our view is that the proposed effects are biologically plausible but unproven. Although there are acknowledged limitations in the analyses that have generated the controversy, we support the view that the concerns warrant further investigation using appropriate and carefully designed studies. The Expanded Program on Immunization strategy was implemented 30 y ago. Now that both disease patterns and available vaccines have changed, the immunization schedule might need to be revised. Randomized controlled trials investigating child survival according to several vaccine formulations and/or immunization schedules, with or without nutrient supplementation (in particular VA), are now justified.

Recommendations for future research

Given the poor state of existing knowledge on interactions between nutrition and vaccine responses, there is a strong need to invest in this area. Modern immunological methods offer fresh opportunities to explore newly emerging facets of immune function and should be applied in studies carefully designed to account for possible confounding factors associated with malnutrition. For instance, recent findings from in vitro and experimental animal research have considerably advanced our understanding of retinoid metabolism and effects on Tregs and Th1/Th2/Th17 balance. Likewise, the mechanisms by which vitamin D augments defenses against intracellular organisms have recently been discovered. Such information can guide new in vivo research designs.

With regard specifically to vaccine responses, studies should ideally investigate actual protective efficacy, but it is recognized that this is generally not feasible except in the very largest of studies. Relying therefore on intermediate proxies for protection, there is a need to focus not only on antibody levels but also on antibody affinity/avidity, timing of immune response, mucosal and cellular immunity, T cell bias, etc. There is a need for standardization in experimental designs and statistical analyses. Although our analysis indicates weak effects of nutrition on current vaccines, this may be because these vaccines all have a very high efficacy. Additional research may prove fruitful on vaccines with marginal efficacy that are currently under development (e.g. HIV and malaria vaccines).

With regard to the VA/DTP vaccine controversy, we support the on-going initiatives to collect available demographic datasets to extend the post hoc analyses so far conducted. Until this has been completed, and depending upon the outcome, we do not endorse the large investment that would be required to conduct a fresh RCT investigating mortality with and without VA without first investigating the mechanisms by which VA affect immune responses. Besides, the past analysis revealing potential adverse effects was conducted in an area of unacceptably high mortality in Guinea Bissau and it would be unethical to conduct such a trial without putting in place better health care and hence greatly reducing mortality which in turn would increase the sample size required. A possible approach to address this would be to conduct intensive small-scale studies using modern methods to better understand the immunological sequelae of coadministration of micronutrients and vaccines. Such studies would be ethically appropriate and highly feasible in a number of third world settings where advanced immunological laboratories already exist. Creation of a network of such laboratories to share expertise in advanced methodologies and research protocols would be advantageous.

Given the fact that the vast majority of the annual toll of 10 million child deaths worldwide is caused by infections, that malnutrition is a contributory factor in at least 30% of these deaths, and that many are vaccine and/or micronutrient preventable, there is a strong case for investment in a major new discovery program to delineate the nature, strengths, and mechanisms of interactions between nutrients (especially VA) and the specific and nonspecific responses to vaccination. At the very least, such a program would provide an evidence base for the revision and optimization of vaccine and micronutrient supplementation schedules and could potentially open up new avenues of intervention, including through the enhancement of novel vaccine development.

	ACKNOWLEDGMENTS

	FOOTNOTES

 
¹ Published in a supplement to The Journal of Nutrition. From a report submitted to the Bill and Melinda Gates Foundation (BMGF), August 25, 2008. The supplement coordinator for this supplement is Andrew M. Prentice, London School of Hygiene and Tropical Medicine. Guest editors for this supplement are Catherine J. Field and A. Catharine Ross. Publication costs for this supplement were defrayed in part by the payment of page charges. This publication must therefore be hereby marked "advertisement" in accordance with 18 USC section 1734 solely to indicate this fact. The work was commissioned by the BMGF. Supplement Coordinator disclosure: Andrew Prentice is employed by UKMRC and was paid for travel expenses to present the results of the analysis to BMGF. Supplement Guest Editor disclosures: No conflicts to disclose. The opinions expressed in this publication are those of the authors and are not attributable to the sponsors or the publisher, editor, or editorial board of The Journal of Nutrition.

² Author disclosures: BMGF supported M. Savy's salary. K. Edmond, P. E. M. Fine, A. Hall, B. J. Hennig, S. E. Moore, K. Mulholland, U. Schaible, and A. M. Prentice were paid a small consultancy fee payable to their research funds. U. Schaible received a separate and unconnected grant from BMGF. M. Savy and A. Prentice were paid travel expenses to travel to Seattle to present the results of the analysis to BMGF.

³ Supplemental Files 1 and 2 and Tables 1–3 are available with the online posting of this paper at jn.nutrition.org.

⁴ Current address: MRC International Nutrition Group, Nutrition and Public Health Intervention Research Unit, London School of Hygiene and Tropical Medicine, Keppel street, London WC1E 7HT, UK.

⁹ Abbreviations used: BCG, bacille Calmette-Guerin; DTH, delayed-type hypersensitivity; DTP, diphtheria-pertussis-tetanus; GMT, geometric mean titer; HAI assay, hemagglutination inhibition assay; Hib, haemophilus influenzae type b; IL, interleukin; MUAC, mid-upper arm circumference, MV, measles vaccine; OPV, oral poliovirus vaccine; OR, odds ratio; PCV, pneumococcal conjugate vaccine; PEM, protein-energy malnutrition; PPD, purified protein derivative; PRN, plaque reduction neutralization; RA, retinoic acid; RCT, randomized controlled trial; RE, retinol equivalent; Th, T helper; Tregs, T regulatory cell; TU, tuberculin unit; VA, vitamin A; VAS, vitamin A supplementation.

¹⁰ See cautionary note regarding Chandra's studies in the Methods section.

¹¹ See cautionary note regarding Chandra's studies in the Methods section.

¹² See cautionary note regarding Chandra's studies in the Methods section.

	LITERATURE CITED

TOP ABSTRACT Introduction Methods Results Discussion LITERATURE CITED

 

1. UNICEF. Countdown to 2015: maternal, newborn and child survival. Tracking progress in maternal, newborn and child survival, the 2008 report. New York: UNICEF; 2008.

2. Bhutta ZA, Ahmed T, Black RE, Cousens S, Dewey K, Giugliani E, Haider BA, Kirkwood B, Morris SS, et al. What works? Interventions for maternal and child undernutrition and survival. Lancet. 2008;371:417–40.[CrossRef][Medline]

3. Moore SE, Goldblatt D, Bates CJ, Prentice AM. Impact of nutritional status on antibody responses to different vaccines in undernourished Gambian children. Acta Paediatr. 2003;92:170–6.[Medline]

4. Van Loveren H, Van Amsterdam JG, Vandebriel RJ, Kimman TG, Rumke HC, Steerenberg PS, Vos JG. Vaccine-induced antibody responses as parameters of the influence of endogenous and environmental factors. Environ Health Perspect. 2001;109:757–64.[Medline]

5. Benn CS, Fisker AB, Jorgensen MJ, Aaby P. Why worry: vitamin A with DTP vaccine? Vaccine. 2007;25:777–9.[CrossRef][Medline]

6. Benn CS, Martins C, Rodrigues A, Jensen H, Lisse IM, Aaby P. Randomised study of effect of different doses of vitamin A on childhood morbidity and mortality. BMJ. 2005;331:1428–32.[Abstract/Free Full Text]

7. Newton S, Owusu-Agyei S. Need to worry: wrong vitamin A doses reported! Vaccine. 2007;25:4514.[CrossRef][Medline]

8. Benn CS, Bale C, Sommerfelt H, Friis H, Aaby P. Hypothesis: vitamin A supplementation and childhood mortality: amplification of the non-specific effects of vaccines? Int J Epidemiol. 2003;32:822–8.[Abstract/Free Full Text]

9. White C. Three journals raise doubts on validity of Canadian studies. BMJ. 2004;328:67.[Free Full Text]

10. Smith R. Investigating the previous studies of a fraudulent author. BMJ. 2005;331:288–291. doi:10.1136/bmj.331.7511.288. Available from: http://www.bmj.com/cgi/content/full/331/7511/288.[Free Full Text]

11. Plotkin SA. Immunologic correlates of protection induced by vaccination. Pediatr Infect Dis J. 2001;20:63–75.[CrossRef][Medline]

12. Ada G. Vaccines and vaccination. N Engl J Med. 2001;345:1042–53.[Free Full Text]

13. Beverley PC. Immunology of vaccination. Br Med Bull. 2002;62:15–28.[Abstract/Free Full Text]

14. Murphy KM, Travers P, Walport M. Janeway's immunobiology. 7th ed. New York: Garland Science; 2007.

15. Plotkin SA, Oreinsten WA, Offit PA. Vaccines. 4th ed. Philadelphia: Elsevier Inc.; 2004.

16. Plotkin SA. Vaccines: correlates of vaccine-induced immunity. Clin Infect Dis. 2008;47:401–9.[CrossRef][Medline]

17. Orenstein WA, Bernier RH, Hinman AR. Assessing vaccine efficacy in the field. Further observations. Epidemiol Rev. 1988;10:212–41.[Free Full Text]

18. Muhlemann K, Weiss NS. Can herd immunity influence the assessment of vaccine efficacy in nonrandomized studies? Am J Public Health. 1997;87:113–4.[Free Full Text]

19. The Nordic Cochrane Centre. Review Manager (RevMan) version 5.0. Copenhagen: The Cochrane Collaboration; 2008.

20. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557–60.[Free Full Text]

21. Faulk WP, Mata LJ, Edsall G. Effects of malnutrition on the immune response in humans: a review. Trop Dis Bull. 1975;72:89–103.[Medline]

22. Scrimshaw NS, SanGiovanni JP. Synergism of nutrition, infection, and immunity: an overview. Am J Clin Nutr. 1997;66:S464–77.[Medline]

23. Chandra RK. 1990 McCollum Award lecture. Nutrition and immunity: lessons from the past and new insights into the future. Am J Clin Nutr. 1991;53:1087–101.[Free Full Text]

24. Nassar MF, Younis NT, Tohamy AG, Dalam DM, El Badawy MA. T-lymphocyte subsets and thymic size in malnourished infants in Egypt: a hospital-based study. East Mediterr Health J. 2007;13:1031–42.[Medline]

25. Salimonu LS, Johnson AO, Williams AI, Adeleye GI, Osunkoya BO. Lymphocyte subpopulations and antibody levels in immunized malnourished children. Br J Nutr. 1982;48:7–14.[CrossRef][Medline]

26. Doherty JF, Golden MH, Remick DG, Griffin GE. Production of interleukin-6 and tumour necrosis factor-alpha in vitro is reduced in whole blood of severely malnourished children. Clin Sci (Lond). 1994;86:347–51.[Medline]

27. Munoz C, Arevalo MT, Lopez M, Impaired LS. Interleukin-1 and tumor necrosis factor production in protein-calorie malnutrition. Nutr Res. 1994;14:347–52.[CrossRef]

28. Kielmann AA, Uberoi IS, Chandra RK, Mehra VL. The effect of nutritional status on immune capacity and immune responses in preschool children in a rural community in India. Bull World Health Organ. 1976;54:477–83.[Medline]

29. Neumann CG, Lawlor GJ Jr, Stiehm ER, Swenseid ME, Newton C, Herbert J, Ammann AJ, Jacob M. Immunologic responses in malnourished children. Am J Clin Nutr. 1975;28:89–104.[Free Full Text]

30. Sinha DP, Bang FB. Protein and calorie malnutrition, cell-mediated immunity, and B.C.G. vaccination in children from rural West Bengal. Lancet. 1976;2:531–4.[CrossRef][Medline]

31. Harland PS, Brown RE. Tuberculin sensitivity following B.C.G. vaccination in undernourished children. East Afr Med J. 1965;42:233–8.[Medline]

32. Heresi GP, Saitua MT, Schlesinger L. Leukocyte migration inhibition factor production in marasmic infants. Am J Clin Nutr. 1981;34:909–13.[Abstract/Free Full Text]

33. Heyworth B. Delayed hypersensitivity to PPD-S following BCG vaccination in African children: an 18-month field study. Trans R Soc Trop Med Hyg. 1977;71:251–3.[CrossRef][Medline]

34. McMurray DN, Loomis SA, Casazza LJ, Rey H, Miranda R. Development of impaired cell-mediated immunity in mild and moderate malnutrition. Am J Clin Nutr. 1981;34:68–77.[Abstract/Free Full Text]

35. Schlesinger L, Stekel A. Impaired cellular immunity in marasmic infants. Am J Clin Nutr. 1974;27:615–20.[Abstract]

36. Greenwood BM, Bradley-Moore AM, Bradley AK, Kirkwood BR, Gilles HM. The immune response to vaccination in undernourished and well-nourished Nigerian children. Ann Trop Med Parasitol. 1986;80:537–44.[Medline]

37. Narain R, Lakshminarayana P, Vallishayee RS, Narmada R, Diwakara AM. Influence of nutrition on delayed skin hypersensitivity. Indian J Med Res. 1979;69:6–17.[Medline]

38. Rafii M, Hashemi S, Nahani J, Mohagheghpour N. Immune response in malnourished children. Clin Immunol Immunopathol. 1977;8:1–6.[CrossRef][Medline]

39. Satyanarayana K, Bhaskaram P, Seshu VC, Reddy V. Influence of nutrition on postvaccinal tuberculin sensitivity. Am J Clin Nutr. 1980;33:2334–7.[Abstract/Free Full Text]

40. Chopra K, Kundu S, Chowdhury DS. Antibody response of infants in tropics to five doses of oral polio vaccine. J Trop Pediatr. 1989;35:19–23.[Abstract/Free Full Text]

41. Chandra RK. Reduced secretory antibody response to live attenuated measles and poliovirus vaccines in malnourished children. BMJ. 1975;2:583–5.[Abstract/Free Full Text]

42. Hafez M, Aref GH, Mehareb SW, Kassem AS, El-Tahhan H, Rizk Z, Mahfouz R, Saad K. Antibody production and complement system in protein energy malnutrition. J Trop Med Hyg. 1977;80:36–9.[Medline]

43. Marshall R, Habicht JP, Landrigan PJ, Foege WH, Delgado H. Effectiveness of measles vaccine given simultaneously with DTP. J Trop Pediatr Environ Child Health. 1974;20:126–9.[Medline]

44. Baer CL, Bratt DE, Edwards R, McFarlane H, Utermohlen V. Response of mildly to moderately malnourished children to measles vaccination. West Indian Med J. 1986;35:106–11.[Medline]

45. Dao H, Delisle H, Fournier P. Anthropometric status, serum prealbumin level and immune response to measles vaccination in Mali children. J Trop Pediatr. 1992;38:179–84.[Abstract/Free Full Text]

46. Ekunwe EO. Malnutrition and seroconversion following measles immunization. J Trop Pediatr. 1985;31:290–1.[Free Full Text]

47. Halsey NA, Boulos R, Mode F, Andre J, Bowman L, Yaeger RG, Toureau S, Rohde J, Boulos C. Response to measles vaccine in Haitian infants 6 to 12 months old. Influence of maternal antibodies, malnutrition, and concurrent illnesses. N Engl J Med. 1985;313:544–9.[Abstract]

48. Idris S, El Seed AM. Measles vaccination in severely malnourished Sudanese children. Ann Trop Paediatr. 1983;3:63–7.[Medline]

49. Ifekwunigwe AE, Grasset N, Glass R, Foster S. Immune responses to measles and smallpox vaccinations in malnourished children. Am J Clin Nutr. 1980;33:621–4.[Abstract/Free Full Text]

50. McMurray DN, Loomis SA, Casazza LJ, Rey H. Influence of moderate malnutrition on morbidity and antibody response following vaccination with live, attenuated measles virus vaccine. Bull Pan Am Health Organ. 1979;13:52–7.[Medline]

51. Monjour L, Bourdillon F, Froment A, Claudio-Ribeiro D, Fabre M, Chastang C, Gentilini M. [Efficacy of measles vaccination in young malnourished African children]. Bull Soc Pathol Exot Filiales. 1984;77:271–7.[Medline]

52. Powell GM. Response to live attenuated measles vaccine in children with severe kwashiorkor. Ann Trop Paediatr. 1982;2:143–5.[Medline]

53. Wesley A, Coovadia HM, Watson AR. Immunization against measles in children at risk for severe disease. Trans R Soc Trop Med Hyg. 1979;73:710–5.[CrossRef][Medline]

54. Brown RE, Katz M. Smallpox vaccination in malnourished children. Trop Geogr Med. 1966;18:129–32.[Medline]

55. Brown RE, Katz M. Failure of antibody production to yellow fever vaccine in children with kwashiorkor. Trop Geogr Med. 1966;18:125–8.[Medline]

56. Chandra RK, Chandra S, Gupta S. Antibody affinity and immune complexes after immunization with tetanus toxoid in protein-energy malnutrition. Am J Clin Nutr. 1984;40:131–4.[Abstract/Free Full Text]

57. Brussow H, Sidoti J, Dirren H, Freire WB. Effect of malnutrition in Ecuadorian children on titers of serum antibodies to various microbial antigens. Clin Diagn Lab Immunol. 1995;2:62–8.[Medline]

58. el-Molla A, el-Ghoroury A, Hussein M, Badr-el-Din MK, Hassen AH, Aref GH, el-Araby I. Antibody production in protein calorie malnutrition. J Trop Med Hyg. 1973;76:248–50.[Medline]

59. Glass RI, Svennerholm AM, Stoll BJ, Khan MR, Huda S, Huq MI, Holmgren J. Effects of undernutrition on infection with Vibrio cholerae O1 and on response to oral cholera vaccine. Pediatr Infect Dis J. 1989;8:105–9.[Medline]

60. Moore SE, Morgan G, Collinson AC, Swain JA, O'Connell MA, Prentice AM. Leptin, malnutrition, and immune response in rural Gambian children. Arch Dis Child. 2002;87:192–7.[Abstract/Free Full Text]

61. Dossetor J, Whittle HC, Greenwood BM. Persistent measles infection in malnourished children. BMJ. 1977;1:1633–5.[Abstract/Free Full Text]

62. Coovadia HM, Parent MA, Loening WE, Wesley A, Burgess B, Hallett F, Brain P, Grace J, Naidoo J, et al. An evaluation of factors associated with the depression of immunity in malnutrition and in measles. Am J Clin Nutr. 1974;27:665–9.[Medline]

63. Suskind R, Sirishinha S, Vithayasai V, Edelman R, Damrongsak D, Charupatana C, Olson RE. Immunoglobulins and antibody response in children with protein-calorie malnutrition. Am J Clin Nutr. 1976;29:836–41.[Abstract/Free Full Text]

64. Pretorius PJ, De Villiers LS. Antibody response in children with protein malnutrition. Am J Clin Nutr. 1962;10:379–83.[Abstract]

65. Reddy V, Srikantia SG. Antibody response in Kwashiorkor. Indian J Med Res. 1964;52:1154–8.[Medline]

66. Mathews JD, Mackay IR, Whittingham S, Malcolm LA. Protein supplementation and enhanced antibody-producing capacity in New Guinean school-children. Lancet. 1972;2:675–7.[Medline]

67. Mohammed I, Damisah MM. The immunological response to polyvalent meningococcal vaccine in Bauchi State, Nigeria. Trans R Soc Trop Med Hyg. 1982;76:351–3.[CrossRef][Medline]

68. Greenwood BM, Bradley AK, Blakebrough IS, Whittle HC, Marshall TF, Gilles HM. The immune response to a meningococcal polysaccharide vaccine in an African village. Trans R Soc Trop Med Hyg. 1980;74:340–6.[CrossRef][Medline]

69. el-Gamal Y, Aly RH, Hossny E, Afify E, el-Taliawy D. Response of Egyptian infants with protein calorie malnutrition to hepatitis B vaccination. J Trop Pediatr. 1996;42:144–5.[Abstract/Free Full Text]

70. Villamor E, Fawzi WW. Effects of vitamin A supplementation on immune responses and correlation with clinical outcomes. Clin Microbiol Rev. 2005;18:446–64.[Abstract/Free Full Text]

71. Stephensen CB. Vitamin A, infection, and immune function. Annu Rev Nutr. 2001;21:167–92.[CrossRef][Medline]

72. Ross AC. Vitamin A supplementation and retinoic acid treatment in the regulation of antibody responses in vivo. Vitam Horm. 2007;75:197–222.[CrossRef][Medline]

73. Kim CH. Regulation of FoxP3 regulatory T cells and Th17 cells by retinoids. Clin Dev Immunol. 2008;2008:416910.[Medline]

74. Manicassamy S, Pulendran B. Retinoic acid-dependent regulation of immune responses by dendritic cells and macrophages. Semin Immunol. 2009;21:22–7.[CrossRef][Medline]

75. Mora JR, Iwata M, von Andrian UH. Vitamin effects on the immune system: vitamins A and D take centre stage. Nat Rev Immunol. 2008;8:685–98.[CrossRef][Medline]

76. Uematsu S, Fujimoto K, Jang MH, Yang BG, Jung YJ, Nishiyama M, Sato S, Tsujimura T, Yamamoto M, et al. Regulation of humoral and cellular gut immunity by lamina propria dendritic cells expressing Toll-like receptor 5. Nat Immunol. 2008;9:769–76.[CrossRef][Medline]

77. Ertesvag A, Naderi S, Blomhoff HK. Regulation of B cell proliferation and differentiation by retinoic acid. Semin Immunol. 2009;21:36–41.[CrossRef][Medline]

78. Diness BR, Fisker AB, Roth A, Yazdanbakhsh M, Sartono E, Whittle H, Nante JE, Lisse IM, Ravn H, et al. Effect of high-dose vitamin A supplementation on the immune response to Bacille Calmette-Guerin vaccine. Am J Clin Nutr. 2007;86:1152–9.[Abstract/Free Full Text]

79. Jason J, Archibald LK, Nwanyanwu OC, Kazembe PN, Chatt JA, Norton E, Dobbie H, Jarvis WR. Clinical and immune impact of Mycobacterium bovis BCG vaccination scarring. Infect Immun. 2002;70:6188–95.[Abstract/Free Full Text]

80. Willy ME, Hoover DR, Halsey NA, Maloney EM, Pate EJ, Wiktor SZ, Blattner WA, Manns A. Relationship between unsupplemented vitamin A serum concentrations and measles vaccine response in Jamaican children. Pediatr Infect Dis J. 1998;17:526–8.[CrossRef][Medline]

81. Semba RD, Munasir Z, Beeler J, Akib A, Muhilal, Audet S, Sommer A. Reduced seroconversion to measles in infants given vitamin A with measles vaccination. Lancet. 1995;345:1330–2.[CrossRef][Medline]

82. Benn CS, Whittle H, Aaby P, Bale C, Michaelsen KF, Olsen J. Vitamin A and measles vaccination. Lancet. 1995;346:503–4.[CrossRef][Medline]

83. Benn CS, Aaby P, Bale C, Olsen J, Michaelsen KF, George E, Whittle H. Randomised trial of effect of vitamin A supplementation on antibody response to measles vaccine in Guinea-Bissau, west Africa. Lancet. 1997;350:101–5.[CrossRef][Medline]

84. Benn CS, Balde A, George E, Kidd M, Whittle H, Lisse IM, Aaby P. Effect of vitamin A supplementation on measles-specific antibody levels in Guinea-Bissau. Lancet. 2002;359:1313–4.[CrossRef][Medline]

85. Bhaskaram P, Rao KV. Enhancement in seroconversion to measles vaccine with simultaneous administration of vitamin A in 9-months-old Indian infants. Indian J Pediatr. 1997;64:503–9.[Medline]

86. Cherian T, Varkki S, Raghupathy P, Ratnam S, Chandra RK. Effect of Vitamin A supplementation on the immune response to measles vaccination. Vaccine. 2003;21:2418–20.[CrossRef][Medline]

87. Semba RD, Akib A, Beeler J, Munasir Z, Permaesih D, Muherdiyantiningsih, Komala, Martuti S, Muhilal. Effect of vitamin A supplementation on measles vaccination in nine-month-old infants. Public Health. 1997;111:245–7.[CrossRef][Medline]

88. Bahl R, Kumar R, Bhandari N, Kant S, Srivastava R, Bhan MK. Vitamin A administered with measles vaccine to nine-month-old infants does not reduce vaccine immunogenicity. J Nutr. 1999;129:1569–73.[Abstract/Free Full Text]

89. de Francisco A, Ahmed F. Measles vaccine failure not associated with vitamin A supplementation. Trans R Soc Trop Med Hyg. 1996;90:441.[CrossRef][Medline]

90. Rahman MM, Alvarez JO, Mahalanabis D, Wahed MA, Islam MA, Habte D, Fuchs GJ. Effect of vitamin A administration on response to oral polio vaccination. Nutr Res. 1998;18:1125–33.[CrossRef]

91. Newton S, Cousens S, Owusu-Agyei S, Filteau S, Stanley C, Linsell L, Kirkwood B. Vitamin A supplementation does not affect infants' immune responses to polio and tetanus vaccines. J Nutr. 2005;135:2669–73.[Abstract/Free Full Text]

92. Semba RD, Muhilal, Mohgaddam NE, Munasir Z, Akib A, Permaesih D, Muherdiyantiningsih, Osterhaus A. Integration of vitamin A supplementation with the expanded program on immunization does not affect seroconversion to oral poliovirus vaccine in infants. J Nutr. 1999;129:2203–5.[Abstract/Free Full Text]

93. Bhaskaram P, Balakrishna N. Effect of administration of 200,000 IU of vitamin A to women within 24 hrs after delivery on response to OPV administered to the newborn. Indian Pediatr. 1998;35:217–21.[Medline]

94. Bahl R, Bhandari N, Kant S, Molbak K, Ostergaard E, Bhan MK. Effect of vitamin A administered at Expanded Program on Immunization contacts on antibody response to oral polio vaccine. Eur J Clin Nutr. 2002;56:321–5.[CrossRef][Medline]

95. Bhaskaram P, Arun Jyothi S, Visweswara Rao K, Narasinga Rao BS. Effects of subclinical vitamin A deficiency and administration of vitamin A as a single dose on immune function in children. Nutr Res. 1989;9:1017–25.[CrossRef]

96. Kutty PM, Mohanram M, Vinodini R. Humoral immune response in vitamin A deficient children. Acta Vitaminol Enzymol. 1981;3:231–5.[Medline]

97. Brown KH, Rajan MM, Chakraborty J, Aziz KM. Failure of a large dose of vitamin A to enhance the antibody response to tetanus toxoid in children. Am J Clin Nutr. 1980;33:212–7.[Abstract/Free Full Text]

98. Deloria-Knoll M, Steinhoff M, Semba RD, Nelson K, Vlahov D, Meinert CL. Effect of zinc and vitamin A supplementation on antibody responses to a pneumococcal conjugate vaccine in HIV-positive injection drug users: a randomized trial. Vaccine. 2006;24:1670–9.[CrossRef][Medline]

99. Kutukculer N, Akil T, Egemen A, Kurugol Z, Aksit S, Ozmen D, Turgan N, Bayindir O, Caglayan S. Adequate immune response to tetanus toxoid and failure of vitamin A and E supplementation to enhance antibody response in healthy children. Vaccine. 2000;18:2979–84.[CrossRef][Medline]

100. Semba RD, Muhilal, Scott AL, Natadisastra G, Wirasasmita S, Mele L, Ridwan E, West KP Jr, Sommer A. Depressed immune response to tetanus in children with vitamin A deficiency. J Nutr. 1992;122:101–7.[Abstract/Free Full Text]

101. Sidell N, Connor MJ, Chang B, Lowe NJ, Borok M. Effects of 13-cis retinoic acid therapy on human antibody responses to defined protein antigens. J Invest Dermatol. 1990;95:597–602.[CrossRef][Medline]

102. Rahman MM, Mahalanabis D, Hossain S, Wahed MA, Alvarez JO, Siber GR, Thompson C, Santosham M, Fuchs GJ. Simultaneous vitamin A administration at routine immunization contact enhances antibody response to diphtheria vaccine in infants younger than six months. J Nutr. 1999;129:2192–5.[Abstract/Free Full Text]

103. Semba RD, Muhilal, Scott AL, Natadisastra G, West KP Jr, Sommer A. Effect of vitamin A supplementation on immunoglobulin G subclass responses to tetanus toxoid in children. Clin Diagn Lab Immunol. 1994;1:172–5.[Medline]

104. Albert MJ, Qadri F, Wahed MA, Ahmed T, Rahman AS, Ahmed F, Bhuiyan NA, Zaman K, Baqui AH, et al. Supplementation with zinc, but not vitamin A, improves seroconversion to vibriocidal antibody in children given an oral cholera vaccine. J Infect Dis. 2003;187:909–13.[CrossRef][Medline]

105. Qadri F, Ahmed T, Wahed MA, Ahmed F, Bhuiyan NA, Rahman AS, Clemens JD, Black RE, Albert MJ. Suppressive effect of zinc on antibody response to cholera toxin in children given the killed, B subunit-whole cell, oral cholera vaccine. Vaccine. 2004;22:416–21.[CrossRef][Medline]

106. Siddiqui FQ, Ahmad MM, Kakar F, Akhtar S, Dil AS. The role of vitamin A in enhancing humoral immunity produced by antirabies vaccine. East Mediterr Health J. 2001;7:799–804.[Medline]

107. Hanekom WA, Yogev R, Heald LM, Edwards KM, Hussey GD, Chadwick EG. Effect of vitamin A therapy on serologic responses and viral load changes after influenza vaccination in children infected with the human immunodeficiency virus. J Pediatr. 2000;136:550–2.[CrossRef][Medline]

108. Gardner EM, Bernstein ED, Popoff KA, Abrutyn E, Gross P, Murasko DM. Immune response to influenza vaccine in healthy elderly: lack of association with plasma beta-carotene, retinol, alpha-tocopherol, or zinc. Mech Ageing Dev. 2000;117:29–45.[CrossRef][Medline]

109. Newton S, Owusu-Agyei S, Ampofo W, Zandoh C, Adjuik M, Adjei G, Tchum S, Filteau S, Kirkwood BR. Vitamin A supplementation enhances infants' immune responses to hepatitis B vaccine but does not affect responses to Haemophilus influenzae type b vaccine. J Nutr. 2007;137:1272–7.[Abstract/Free Full Text]

110. Glasziou PP, Mackerras DE. Vitamin A supplementation in infectious diseases: a meta-analysis. BMJ. 1993;306:366–70.[Abstract/Free Full Text]

111. Fawzi WW, Chalmers TC, Herrera MG, Mosteller F. Vitamin A supplementation and child mortality. A meta-analysis. JAMA. 1993;269:898–903.[CrossRef]

112. Humphrey JH, Agoestina T, Wu L, Usman A, Nurachim M, Subardja D, Hidayat S, Tielsch J, West KP Jr, et al. Impact of neonatal vitamin A supplementation on infant morbidity and mortality. J Pediatr. 1996;128:489–96.[CrossRef][Medline]

113. Rahmathullah L, Tielsch JM, Thulasiraj RD, Katz J, Coles C, Devi S, John R, Prakash K, Sadanand AV, et al. Impact of supplementing newborn infants with vitamin A on early infant mortality: community based randomised trial in southern India. BMJ. 2003;327:254.[Abstract/Free Full Text]

114. Klemm RD, Labrique AB, Christian P, Rashid M, Shamim AA, Katz J, Sommer A, West KP Jr. Newborn vitamin A supplementation reduced infant mortality in rural Bangladesh. Pediatrics. 2008;122:e242–50.[Abstract/Free Full Text]

115. Benn CS, Diness BR, Roth A, Nante E, Fisker AB, Lisse IM, et al. Effect of 50 000 IU vitamin A given with BCG vaccine on mortality in infants in Guinea-Bissau: randomised placebo controlled trial. BMJ. 2008;336:1416–20.

116. Malaba LC, Iliff PJ, Nathoo KJ, Marinda E, Moulton LH, Zijenah LS, Zvandasara P, Ward BJ, Humphrey JH. Effect of postpartum maternal or neonatal vitamin A supplementation on infant mortality among infants born to HIV-negative mothers in Zimbabwe. Am J Clin Nutr. 2005;81:454–60.[Abstract/Free Full Text]

117. West KP Jr, Katz J, Shrestha SR, LeClerq SC, Khatry SK, Pradhan EK, Adhikari R, Wu LS, Pokhrel RP, et al. Mortality of infants < 6 mo of age supplemented with vitamin A: a randomized, double-masked trial in Nepal. Am J Clin Nutr. 1995;62:143–8.[Abstract/Free Full Text]

118. Rahman MM, Mahalanabis D, Wahed MA, Islam MA, Habte D. Administration of 25,000 IU vitamin A doses at routine immunisation in young infants. Eur J Clin Nutr. 1995;49:439–45.[Medline]

119. WHO. Randomised trial to assess benefits and safety of vitamin A supplementation linked to immunisation in early infancy. WHO/CHD Immunisation-Linked Vitamin A Supplementation Study Group. Lancet. 1998;352:1257–63.[CrossRef][Medline]

120. Nante JE, Diness BR, Ravn H, Roth A, Aaby P, Benn CS. No adverse events after simultaneous administration of 50 000 IU vitamin A and Bacille Calmette-Guerin vaccination to normal-birth-weight newborns in Guinea-Bissau. Eur J Clin Nutr. 2008;62:842–8.[CrossRef][Medline]

121. Mahalanabis D, Rahman MM, Wahed MA, Islam MA, Habte D. Vitamin A megadoses during early infancy on serum retinol concentration and acute side effects and residual effects on 6 month follow-up. Nutr Res. 1997;17:649–59.[CrossRef]

122. Newton S, Owusu-Agyei S, Filteau S, Gyan T, Kirkwood BR. Vitamin A supplements are well tolerated with the pentavalent vaccine. Vaccine. 2008;26:6608–13.[CrossRef][Medline]

123. WHO. Report of an informal consultation on vitamin A supplementation. Geneva: WHO; 2000.

124. Benn CS, Aaby P, Nielsen J, Binka FN, Ross DA. Does vitamin A supplementation interact with routine vaccinations? An analysis of the Ghana Vitamin A Supplementation Trial. Am J Clin Nutr. 2009;90:629–39.[Abstract/Free Full Text]

125. Stephensen BC and Livingston KA. Vitamin supplements and vaccines: maximize benefits, evaluate potential risks (Editorial). Am J Clin Nutr 2009;90:457–8.[Free Full Text]

126. Baqui AH, de Francisco A, Arifeen SE, Siddique AK, Sack RB. Bulging fontanelle after supplementation with 25,000 IU of vitamin A in infancy using immunization contacts. Acta Paediatr. 1995;84:863–6.[CrossRef][Medline]

127. de Francisco A, Chakraborty J, Chowdhury HR, Yunus M, Baqui AH, Siddiqui AK, Sack RB. Acute toxicity of vitamin A given with vaccines in infancy. Lancet. 1993;342:526–7.[CrossRef][Medline]

128. Idindili B, Masanja H, Urassa H, Bunini W, van Jaarsveld P, Aponte JJ, Kahigwa E, Mshinda H, Ross D, et al. Randomized controlled safety and efficacy trial of 2 vitamin A supplementation schedules in Tanzanian infants. Am J Clin Nutr. 2007;85:1312–9.[Abstract/Free Full Text]

129. van Dillen J, De Francisco A, Overweg-Plandsoen WCG. Long-term effect of vitamin A with vaccines. Lancet. 1996;347:1705.[Medline]

130. Semba RD, Munasir Z, Akib A, Melikian G, Permaesih D, Muherdiyantiningsih, Marituti S, Muhilal. Integration of vitamin A supplementation with the Expanded Programme on Immunization: lack of impact on morbidity or infant growth. Acta Paediatr. 2001;90:1107–11.[CrossRef][Medline]

131. Gorstein J, Bhaskaram P, Khanum S, Hossaini R, Balakrishna N, Goodman TS, deBenoist B, Krishnaswamy K. Safety and impact of vitamin A supplementation delivered with oral polio vaccine as part of the immunization campaign in Orissa, India. Food Nutr Bull. 2003;24:319–31.[Medline]

132. Darboe MK, Thurnham DI, Morgan G, Adegbola RA, Secka O, Solon JA, Jackson SJ, Northrop-Clewes C, Fulford TJ, et al. Effectiveness of an early supplementation scheme of high-dose vitamin A versus standard WHO protocol in Gambian mothers and infants: a randomised controlled trial. Lancet. 2007;369:2088–96.[CrossRef][Medline]

133. Kristensen I, Aaby P, Jensen H. Routine vaccinations and child survival: follow up study in Guinea-Bissau, West Africa. BMJ. 2000;321:1435–8.[Abstract/Free Full Text]

134. Aaby P, Samb B, Simondon F, Seck AM, Knudsen K, Whittle H. Non-specific beneficial effect of measles immunisation: analysis of mortality studies from developing countries. BMJ. 1995;311:481–5.[Abstract/Free Full Text]

135. Stabell C, Bale C, Pedro da Silva A, Olsen J, Aaby P. No evidence of fontanelle-bulging episodes after vitamin A supplementation of 6- and 9-month-old infants in Guinea Bissau. Eur J Clin Nutr. 1995;49:73–4.[Medline]

136. Arya S, Chellani H, Pandey J. Evaluation of safety of oral vitamin ‘A’ megadose co-administered with measles vaccination. Indian Pediatr. 2000;37:1341–7.[Medline]

137. Stephensen CB. Commentary: a hypothesis concerning vitamin A supplementation, vaccines, and childhood mortality. Int J Epidemiol. 2003;32:828–9.[Free Full Text]

138. Tielsch JM. Vitamin A supplements in newborns and child survival. BMJ. 2008;336:1385–6.[Free Full Text]

139. Prentice AM, Savy M, Darboe MK, Moore SE. Commentary: Challenging public health orthodoxies - prophesy or heresy? Int J Epidemiol 2009;38:591–3.[Free Full Text]

140. Jensen H, Benn CS, Lisse IM, Rodrigues A, Andersen PK, Aaby P. Survival bias in observational studies of the impact of routine immunizations on childhood survival. Trop Med Int Health. 2007;12:5–14.[Medline]

141. Fine PE, Williams TN, Aaby P, Kallander K, Moulton LH, Flanagan KL, Smith PG, Benn CS. Epidemiological studies of the ‘non-specific effects’ of vaccines. I. Data collection in observational studies. Trop Med Int Health. 2009;14:969–76.[CrossRef][Medline]

142. Brock JH, Mulero V. Cellular and molecular aspects of iron and immune function. Proc Nutr Soc. 2000;59:537–40.[Medline]

143. Munoz C, Rios E, Olivos J, Brunser O, Olivares M. Iron, copper and immunocompetence. Br J Nutr. 2007;98 Suppl 1:S24–8.[Medline]

144. Oppenheimer SJ. Iron and its relation to immunity and infectious disease. J Nutr. 2001;131:S616–33; discussion S33–5.[Abstract/Free Full Text]

145. Weiss G. Iron and immunity: a double-edged sword. Eur J Clin Invest. 2002;32 Suppl 1:70–8.[CrossRef][Medline]

146. Kuvibidila S, Dardenne M, Savino W, Lepault F. Influence of iron-deficiency anemia on selected thymus functions in mice: thymulin biological activity, T-cell subsets, and thymocyte proliferation. Am J Clin Nutr. 1990;51:228–32.[Abstract/Free Full Text]

147. Mullick S, Rusia U, Sikka M, Faridi MA. Impact of iron deficiency anaemia on T lymphocytes and their subsets in children. Indian J Med Res. 2006;124:647–54.[Medline]

148. Galan P, Thibault H, Preziosi P, Hercberg S. Interleukin 2 production in iron-deficient children. Biol Trace Elem Res. 1992;32:421–6.[CrossRef][Medline]

149. Berger J, Dyck JL, Galan P, Aplogan A, Schneider D, Traissac P, Hercberg S. Effect of daily iron supplementation on iron status, cell-mediated immunity, and incidence of infections in 6–36 month old Togolese children. Eur J Clin Nutr. 2000;54:29–35.[CrossRef][Medline]

150. Krantman HJ, Young SR, Ank BJ, O'Donnell CM, Rachelefsky GS, Stiehm ER. Immune function in pure iron deficiency. Am J Dis Child. 1982;136:840–4.[Abstract/Free Full Text]

151. Chandra RK, Saraya AK. Impaired immunocompetence associated with iron deficiency. J Pediatr. 1975;86:899–902.[CrossRef][Medline]

152. Thibault H, Galan P, Selz F, Preziosi P, Olivier C, Badoual J, Hercberg S. The immune response in iron-deficient young children: effect of iron supplementation on cell-mediated immunity. Eur J Pediatr. 1993;152:120–4.[CrossRef][Medline]

153. Drakesmith H, Prentice A. Viral infection and iron metabolism. Nat Rev Microbiol. 2008;6:541–52.[CrossRef][Medline]

154. Schaible UE, Kaufmann SH. Iron and microbial infection. Nat Rev Microbiol. 2004;2:946–53.[CrossRef][Medline]

155. Fulop T Jr, Wagner JR, Khalil A, Weber J, Trottier L, Payette H. Relationship between the response to influenza vaccination and the nutritional status in institutionalized elderly subjects. J Gerontol A Biol Sci Med Sci. 1999;54:M59–64.[Medline]

156. Crogan NL, Velasquez D, Gagan MJ. Testing the feasibility and initial effects of iron and vitamin C to enhance nursing home residents' immune status following an influenza vaccine. Geriatr Nurs. 2005;26:188–94.[CrossRef][Medline]

157. Bagchi K, Mohanram M, Reddy V. Humoral immune response in children with iron-deficiency anaemia. BMJ. 1980;280:1249–51.[Abstract/Free Full Text]

158. MacDougall LG, Jacobs MR. The immune response in iron-deficient children. Isohaemagglutinin titres and antibody response to immunization. S Afr Med J. 1978;53:405–7.[Medline]

159. Macdougall LG, Anderson R, McNab GM, Katz J. The immune response in iron-deficient children: Impaired cellular defense mechanisms with altered humoral components. J Pediatr. 1975;86:833–43.[CrossRef][Medline]

160. Nalder BN, Mahoney AW, Ramakrishnan R, Hendricks DG. Sensitivity of the immunological response to the nutritional status of rats. J Nutr. 1972;102:535–41.[Abstract/Free Full Text]

161. Dhur A, Galan P, Hercberg S. Iron status, immune capacity and resistance to infections. Comp Biochem Physiol A. 1989;94:11–9.

162. Kuvibidila S. Iron deficiency, cell-mediated immunity and resistance against infections: present knowledge and controversies. Nutr Res. 1987;7:989–1003.[CrossRef]

163. Fraker PJ, King LE, Laakko T, Vollmer TL. The dynamic link between the integrity of the immune system and zinc status. J Nutr. 2000;130:S1399–406.[Abstract/Free Full Text]

164. Keen CL, Gershwin ME. Zinc deficiency and immune function. Annu Rev Nutr. 1990;10:415–31.[CrossRef][Medline]

165. Shankar AH, Prasad AS. Zinc and immune function: the biological basis of altered resistance to infection. Am J Clin Nutr. 1998;68:S447–63.[Abstract]

166. Ibs KH, Rink L. Zinc-altered immune function. J Nutr. 2003;133:S1452–6.[Abstract/Free Full Text]

167. Prasad AS. Zinc in human health: effect of zinc on immune cells. Mol Med. 2008;14:353–7.[CrossRef][Medline]

168. Fraker PJ, King LE. Reprogramming of the immune system during zinc deficiency. Annu Rev Nutr. 2004;24:277–98.[CrossRef][Medline]

169. Prasad AS. Zinc: an overview. Nutrition. 1995;11:93–9.[Medline]

170. Lastra MD, Pastelin R, Herrera MA, Orihuela VD, Aguilar AE. Increment of immune responses in mice perinatal stages after zinc supplementation. Arch Med Res. 1997;28:67–72.[Medline]

171. Shinde P, Dass RS, Garg AK, Chaturvedi VK, Kumar R. Effect of zinc supplementation from different sources on growth, nutrient digestibility, blood metabolic profile, and immune response of male Guinea pigs. Biol Trace Elem Res. 2006;112:247–62.[CrossRef][Medline]

172. Bao B, Prasad A, Beck FW, Suneja A, Sarkar F. Toxic effect of zinc on NF-kappaB, IL-2, IL-2 receptor alpha, and TNF-alpha in HUT-78 (Th(0)) cells. Toxicol Lett. 2006;166:222–8.[CrossRef][Medline]

173. Osendarp SJ, Fuchs GJ, van Raaij JM, Mahmud H, Tofail F, Black RE, Prabhakar H, Santosham M. The effect of zinc supplementation during pregnancy on immune response to Hib and BCG vaccines in Bangladesh. J Trop Pediatr. 2006;52:316–23.[Abstract/Free Full Text]

174. Duchateau J, Delepesse G, Vrijens R, Collet H. Beneficial effects of oral zinc supplementation on the immune response of old people. Am J Med. 1981;70:1001–4.[CrossRef][Medline]

175. Kreft B, Fischer A, Kruger S, Sack K, Kirchner H, Rink L. The impaired immune response to diphtheria vaccination in elderly chronic hemodialysis patients is related to zinc deficiency. Biogerontology. 2000;1:61–6.[CrossRef][Medline]

176. Karlsen TH, Sommerfelt H, Klomstad S, Andersen PK, Strand TA, Ulvik RJ, Ahren C, Grewal HM. Intestinal and systemic immune responses to an oral cholera toxoid B subunit whole-cell vaccine administered during zinc supplementation. Infect Immun. 2003;71:3909–13.[Abstract/Free Full Text]

177. Ahmed T, Svennerholm AM, Al Tarique A, Sultana GN, Qadri F. Enhanced immunogenicity of an oral inactivated cholera vaccine in infants in Bangladesh obtained by zinc supplementation and by temporary withholding breast-feeding. Vaccine. 2009;27:1433–9.[CrossRef][Medline]

178. Stiedemann M, Harrill I. Relation of immunocompetence to selected nutrients in elderly women. Nutr Rep Int. 1980;21:931–42.

179. Provinciali M, Montenovo A, Di Stefano G, Colombo M, Daghetta L, Cairati M, Veroni C, Cassino R, Della Torre F, et al. Effect of zinc or zinc plus arginine supplementation on antibody titre and lymphocyte subsets after influenza vaccination in elderly subjects: a randomized controlled trial. Age Ageing. 1998;27:715–22.[Abstract/Free Full Text]

180. Remarque EJ, Witkamp L, Masurel N, Ligthart GJ. Zinc supplementation does not enhance antibody formation to influenza virus vaccine in the elderly. Aging Immunol Infect Dis. 1993;4:17–23.

181. Girodon F, Galan P, Monget AL, Boutron-Ruault MC, Brunet-Lecomte P, Preziosi P, Arnaud J, Manuguerra JC, Herchberg S. Impact of trace elements and vitamin supplementation on immunity and infections in institutionalized elderly patients: a randomized controlled trial. MIN. VIT. AOX. geriatric network. Arch Intern Med. 1999;159:748–54.[Abstract/Free Full Text]

182. Grekas D, Alivanis P, Kotzadamis N, Kiriazopoulou V, Pyrpasopoulos M, Tourkantonis A. Influenza vaccination in chronic hemodialysis patients. The effect of zinc supplementation. Ren Fail. 1992;14:575–8.[CrossRef][Medline]

183. Osendarp SJ, Prabhakar H, Fuchs GJ, van Raaij JM, Mahmud H, Tofail F, Santosham M, Black RE. Immunization with the heptavalent pneumococcal conjugate vaccine in Bangladeshi infants and effects of zinc supplementation. Vaccine. 2007;25:3347–54.[CrossRef][Medline]

184. Brodersen HP, Holtkamp W, Larbig D, Beckers B, Thiery J, Lautenschlager J, Probst HJ, Ropertz S, Yavari A. Zinc supplementation and hepatitis B vaccination in chronic haemodialysis patients: a multicentre study. Nephrol Dial Transplant. 1995;10:1780.[Free Full Text]

185. Cantorna MT, Zhu Y, Froicu M, Wittke A. Vitamin D status, 1,25-dihydroxyvitamin D3, and the immune system. Am J Clin Nutr. 2004;80:S1717–20.[Abstract/Free Full Text]

186. Cantorna MT, Yu S, Bruce D. The paradoxical effects of vitamin D on type 1 mediated immunity. Mol Aspects Med. 2008;29:369–75.[CrossRef][Medline]

187. Mahon BD, Wittke A, Weaver V, Cantorna MT. The targets of vitamin D depend on the differentiation and activation status of CD4 positive T cells. J Cell Biochem. 2003;89:922–32.[CrossRef][Medline]

188. Pichler J, Gerstmayr M, Szepfalusi Z, Urbanek R, Peterlik M, Willhelm M. 1,25(OH)2D3 inhibits not only Th1 but also Th2 differentiation in human cord blood T cells. Pediatr Res. 2002;52:12–8.[CrossRef][Medline]

189. Penna G, Roncari A, Amuchastegui S, Daniel KC, Berti E, Colonna M, Adorini L. Expression of the inhibitory receptor ILT3 on dendritic cells is dispensable for induction of CD4+Foxp3+ regulatory T cells by 1,25-dihydroxyvitamin D3. Blood. 2005;106:3490–7.[Abstract/Free Full Text]

190. Daniel C, Sartory NA, Zahn N, Radeke HH, Stein JM. Immune modulatory treatment of trinitrobenzene sulfonic acid colitis with calcitriol is associated with a change of a T helper (Th) 1/Th17 to a Th2 and regulatory T cell profile. J Pharmacol Exp Ther. 2008;324:23–33.[Abstract/Free Full Text]

191. Chen S, Sims GP, Chen XX, Gu YY, Chen S, Lipsky PE. Modulatory effects of 1,25-dihydroxyvitamin D3 on human B cell differentiation. J Immunol. 2007;179:1634–47.[Abstract/Free Full Text]

192. Moe SM, Zekonis M, Harezlak J, Ambrosius WT, Gassensmith CM, Murphy CL, Russell RR, Batiuk TD. A placebo-controlled trial to evaluate immunomodulatory effects of paricalcitol. Am J Kidney Dis. 2001;38:792–802.[Medline]

193. Peces R, de la Torre M, Alcazar R, Urra JM. Prospective analysis of the factors influencing the antibody response to hepatitis B vaccine in hemodialysis patients. Am J Kidney Dis. 1997;29:239–45.[CrossRef][Medline]

194. Antonen JA, Hannula PM, Pyhala R, Saha HH, Ala-Houhala IO, Pasternack AI. Adequate seroresponse to influenza vaccination in dialysis patients. Nephron. 2000;86:56–61.[CrossRef][Medline]

195. Kriesel JD, Spruance J. Calcitriol (1,25-dihydroxy-vitamin D3) coadministered with influenza vaccine does not enhance humoral immunity in human volunteers. Vaccine. 1999;17:1883–8.[CrossRef][Medline]

196. Daynes RA, Enioutina EY, Butler S, Mu HH, McGee ZA, Araneo BA. Induction of common mucosal immunity by hormonally immunomodulated peripheral immunization. Infect Immun. 1996;64:1100–9.[Abstract]

197. Enioutina EY, Visic D, McGee ZA, Daynes RA. The induction of systemic and mucosal immune responses following the subcutaneous immunization of mature adult mice: characterization of the antibodies in mucosal secretions of animals immunized with antigen formulations containing a vitamin D3 adjuvant. Vaccine. 1999;17:3050–64.[CrossRef][Medline]

198. Verminnen K, Loock MV, Cox E, Goddeeris BM, Vanrompay D. Protection of turkeys against Chlamydophila psittaci challenge by DNA and rMOMP vaccination and evaluation of the immunomodulating effect of 1 alpha,25-dihydroxyvitamin D(3). Vaccine. 2005;23:4509–16.[CrossRef][Medline]

199. Ivanov AP, Dragunsky EM, Chumakov KM. 1,25-dihydroxyvitamin D3 enhances systemic and mucosal immune responses to inactivated poliovirus vaccine in mice. J Infect Dis. 2006;193:598–600.[CrossRef][Medline]

200. Van Der Stede Y, Verfaillie T, Cox E, Verdonck F, Goddeeris BM. 1alpha,25-dihydroxyvitamin D3 increases IgA serum antibody responses and IgA antibody-secreting cell numbers in the Peyer's patches of pigs after intramuscular immunization. Clin Exp Immunol. 2004;135:380–90.[CrossRef][Medline]

201. Man WD, Weber M, Palmer A, Schneider G, Wadda R, Jaffar S, Mulholland EK, Greenwood BM. Nutritional status of children admitted to hospital with different diseases and its relationship to outcome in The Gambia, West Africa. Trop Med Int Health. 1998;3:678–86.[Medline]

202. Moore SE, Collinson AC, Fulford AJ, Jalil F, Siegrist CA, Goldblatt D, Hanson LA, Prentice AM. Effect of month of vaccine administration on antibody responses in The Gambia and Pakistan. Trop Med Int Health. 2006;11:1529–41.[CrossRef][Medline]

This Article Abstract Full Text (PDF) Online Supplemental Material Free Access Article All Versions of this Article: 139/11/2154S    most recent jn.109.105312v1 Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Savy, M. Articles by Prentice, A. M. PubMed PubMed Citation Articles by Savy, M. Articles by Prentice, A. M.