 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
a Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts
b Department of Neurology, Children's Hospital, Boston, Massachusetts
c Department of Biostatistics, Harvard School of Public Health, Boston, Massachusetts
d Department of Nutrition, Harvard School of Public Health, Boston, Massachusetts
e Department of Community Health, Muhimbili University College of Health Sciences, Dar es Salaam, Tanzania
f Maternal and Child Health, Harvard School of Public Health, Boston, Massachusetts
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS. The Bayley Scales of Infant Development, 2nd Edition, were administered at 6, 12, and 18 months of age to a subset of children (N = 327). We assessed the effect of vitamin A and multivitamin (vitamins B, C, and E) supplementation using linear regression models and Cox proportional hazard models for the Mental Development Index, the Psychomotor Development Index, and raw scores separately.
RESULTS. Multivitamin supplementation was associated significantly with a mean increase in Psychomotor Development Index score of 2.6 (95% confidence interval: 0.1–5.1). Multivitamins were also significantly protective against the risk for developmental delay on the motor scale (relative risk: 0.4; 95% confidence interval: 0.2–0.7) but not on the Mental Development Index. Vitamin A supplementation had no significant effect on these outcomes.
CONCLUSIONS. Maternal multivitamin supplements provide a low-cost intervention to reduce the risk for developmental delays among infants who are born to HIV-positive mothers in developing countries.
Key Words: multivitamins • mental development • psychomotor development • supplementation • children • Tanzania
Abbreviations: LBW—low birth weight • BSID-II—Bayley Scales of Infant Development, 2nd Edition • MDI—Mental Development Index • PDI—Psychomotor Development Index • ESR—erythrocyte sedimentation rate • CI—confidence interval • SES—socioeconomic status
ALTHOUGH STUDIES HAVE sought to improve mental and psychomotor development in children through protein energy1 and multivitamin child supplements,2–4 few have examined the effect of maternal supplements on infant development. Nutritional deficiencies are common in children in developing countries.5 In 1996, 43% of preschool Tanzanian children were stunted (low height for age).6 Vitamin A, iodine, and iron deficiencies are also prevalent.7–9 Pepping et al8 found that vitamin A deficiency was a significant problem in Tanzania, with 1.5% of the preschool age children whom they surveyed having xerophthalmia and 1.6% having Bitot's spots. In addition, children are have HIV-1 infection; antenatal HIV-1 seroprevalence has been 
12% since 1995 in Dar es Salaam.10 Because multivitamin deficiencies are associated with faster HIV disease progression in adults11 and higher risk for infant morbidity,12 they may have a negative impact on the mental and psychomotor development of the child.
Daily consumption of multivitamin supplements by HIV-1–infected pregnant women has been shown to delay significantly the progression of HIV disease.13 Multivitamin supplements also resulted in significant reductions in the risk for low birth weight (LBW), severe preterm birth, and fetal loss and improved the immunologic profile of the mothers and increased their hemoglobin concentrations.14 Vitamin A supplementation during pregnancy has also been shown to reduce the risk for preterm delivery among HIV-infected women.15 Baylin et al16 demonstrated that vitamin A supplements that are given to HIV-infected mothers during pregnancy and lactation increases serum vitamin A in their infants during the first 6 months of life. Similarly, multivitamin supplements (B, C, and E) increase serum vitamins B12 and E.
Improved maternal multivitamin status after supplementation to the mother during the prenatal and postnatal periods could result in improved mental and motor development in the breastfed child through 3 possible mechanisms: (1) by reducing the risk for LBW and prematurity in the prenatal supplement phase, both of which have been shown to be associated with poor neurodevelopment17–19; (2) through postnatal supplements to the mothers, which would improve her own health, allowing her to be more interactive with the child; or (3) by improving the child's vitamin status and directly reducing child morbidity.20 We conducted a randomized trial among HIV-1–infected women and examined the effect of maternal multivitamin or vitamin A supplementation on the mental and psychomotor development of children in Tanzania.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
During enrollment for each pregnant woman, a number of characteristics were captured in response to questions asked by the counselor as part of a structured questionnaire interview. Enrolled women were followed up at the study clinic at Muhimbili National Hospital monthly during pregnancy and with their infants after delivery for a minimum of 18 months. It was hypothesized that pharmacologic doses were needed, ie, that HIV-infected individuals needed intakes in multiples of the RDA to achieve adequate plasma nutrient values. Women received a daily dose during pregnancy and continued after delivery, throughout follow-up, 1 of the following regimens: vitamin A (30 mg of ß-carotene plus 5000 IU preformed vitamin A); multivitamins excluding vitamin A (20 mg of B1, 20 mg of B2, 25 mg of B6, 100 mg of niacin, 50 µg of B12, 500 mg of C, 30 mg of E, and 0.8 mg of folic acid); multivitamins including vitamin A, all formulated in 2 tablets; or 2 tablets of placebo. At delivery, women who were taking vitamin A supplementation received an additional oral dose of vitamin A (200000 IU), and the others received an extra dose of placebo. At each monthly visit to the study clinic, used bottles were exchanged for new bottles, and the remaining pills were counted. All children, regardless of maternal regimen assignment, were given vitamin A 100000 IU at 6 months and 200000 IU every 6 months thereafter as per national guidelines in Tanzania. Children had postnatal exposure to the trial supplements through their mothers' breast milk. Data used in this article were collected within this framework.
All women gave informed consent to participate in the study according to a strict protocol that was approved by the institutional review boards for Muhimbili University College of Health Sciences, the Tanzanian National AIDS Control Program, and Harvard School of Public Health. These boards also approved the study protocol.
Measures
The Bayley Scales of Infant Development, 2nd Edition22 (BSID-II), was used to assess the current developmental functioning of children aged 6 to 18 months in this study. The BSID-II consists of 3 scales: mental, motor, and behavior rating scale. Only the Mental Development Index (MDI) and the Psychomotor Development Index (PDI) were considered in this study. The BSID-II was administered at 6, 12, and 18 months to a subset of the infants who were born from first-study pregnancies, attended clinic on Mondays or Fridays, and had an identification number that ended in an uneven digit. A Kiswahili translation of all of the instructions to the child was used during each assessment. Tests were administered by 1 of 2 trained Tanzanian nurses, who were masked to the HIV-1 status of the children. Test administration occurred at the beginning of the clinic visit for each child. Children who were acutely ill were not tested, but their mothers were asked to return with them once they had recovered.
Statistical Analysis
BSID-II raw scores signify the number of individual items that a child passes on either the MDI or the PDI. The MDI and the PDI are conversions of these raw scores, compared with a standardized and representative sample of US children. Any assessment for which the achieved raw score "fell off" the development index range (ie, <50) was represented by a value of 49 in analyses. The MDI and PDI and their associated raw scores each were considered separately as outcomes for analyses. The analyses of raw mental and motor scores were considered in an effort to differentiate between a delay in achieving new developmental capacities and losing previously acquired capacities, as well as to avoid making representation of children who had raw scores that did not convert to a score within the scaled score range. Only singleton births were included in the analyses because twins are at greater risk and have different mental and psychomotor development than singletons.
For longitudinal analyses, linear regression models were fitted for each of the outcomes separately using all available data. In each case, additional confounders were considered. Given the factorial design, we examined the effect of multivitamin versus no multivitamin supplementation and of vitamin A versus no vitamin A supplementation on child mental and psychomotor development. Each supplement was also considered a possible modifier of the effect of the other supplement by adding an interaction term. LBW (<2500 g), the child's HIV-1 status, maternal CD4 count at baseline (CD4 <200 cells/mm3), maternal total lymphocyte count at baseline (1340 cells/µL, representing the lowest quartile), maternal erythrocyte sedimentation rate (ESR; 81 mm/hour, representing the highest quartile) at baseline, duration of supplementation during pregnancy (the difference between gestational age at enrollment and gestational age at delivery), and timing of the initiation of treatment during pregnancy (as a continuous covariate) all were considered as potential modifiers of the effect of maternal supplementation as well.
The average absolute effects of supplementation on mental and motor outcomes at 6, 12, and 18 months were calculated separately. The timing of assessments was not always according to schedule, so the data were grouped for these analyses around the scheduled time points. Repeat assessments were not used in these analyses. Instead, the assessment that was performed closest to the scheduled assessment time was used.
Any observed difference in mental and psychomotor development scores over 6 to 18 months between groups defined by vitamin A supplementation is more likely attributable to vitamin A supplementation in the first 6 months because all children received vitamin A supplements at 6, 12, and 18 months. For this reason, particular attention was paid to the effect of vitamin A supplementation on mental and motor outcomes at 6 months. The empirical variance was used to estimate 95% confidence intervals (CIs) for all results from longitudinal analyses to reflect that the data are repeated measures within individuals.
Cox proportional hazards regression models were used to investigate the effect of maternal supplementation on time to an index score <70, where an index score of <70 represents "developmental delay." All analyses were determined a priori and were performed using SAS version 6.12 (SAS, Inc, Cary, NC).
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Breastfeeding was almost universally adopted in this population. The percentage of children who were alive and breastfeeding was 95% at 6 months, 92% at 12 months, and 79% at 18 months.23 The median adherence to maternal supplementation between randomization and delivery was 96% (mean 91%) and 92% (mean 88%) at 30 weeks postpartum and did not differ significantly between groups.14 Median compliance was 89% at 12 months, 84% by 18 months, and 84% by 24 months.24
Table 1 shows the distributions of maternal and child characteristics between groups that were defined by the type of supplementation received and that these distributions were similar. There were also no significant differences, at the 5% level, between groups in terms of biochemical markers, specifically maternal vitamin A, vitamin E, and selenium status at baseline (not presented in Table 1). As expected on the basis of our previous report,14 comparing the 2 groups defined by supplementation, children whose mothers received multivitamin supplements were significantly less likely to have LBW (<2500 g; P = .02). The only significant difference between those who received vitamin A supplementation and those who did not was that mothers in the latter group tended to be slightly older (P = .03).
|
|
|
Table 4 reports the results of Cox proportional hazards models that were used to investigate the effects of the supplements on time to developmental delay. The results of multivariate analyses indicate that multivitamin supplements were significantly protective against the risk for developmental delay on the motor scale (relative risk: 0.4; 95% CI: 0.2–0.7; P = .004) but not on the mental scale. Vitamin A supplementation was not protective against the risk for developmental delay on either scale.
|
81 mm/hour), their raw motor scores were 3.4 items (95% CI: 1.5–5.2) higher for children whose mother received multivitamin supplementation compared with those whose mother did not receive multivitamin supplements. The corresponding comparison between the 2 multivitamin supplementation groups among children who were born to mothers with low ESR (<81 mm/hour), raw motor scores differed by 0.3 items (95% CI: –0.7 to 1.2; P = .004 for interaction). Considering the PDI, among children who were born to mothers with high ESR, multivitamin supplementation was associated with a 7.1-point (95% CI: 1.1–13.0) increase in score, whereas among children who were born to mothers with low ESR, scores differed by 1.6 points (95% CI: –1.4 to 4.6; P = .11 for interaction). ESR group did not modify the effect of multivitamin supplementation on mental scores. Maternal CD4 count at baseline, maternal total lymphocyte count at baseline, duration of multivitamin supplementation, the timing of initiation of supplementation, the child's HIV-1 status, and the child's birth weight were not significant modifiers of the effect of multivitamin supplementation on mental or motor function. The interaction between the 2 supplements was significant for raw motor scores but not for motor index scores. Among those who did not receive vitamin A supplementation, multivitamin supplements were associated with 2.0-item (95% CI: 0.5–3.4) increase in raw motor score, compared with those who did not receive multivitamin supplements. Meanwhile, among children whose mothers received vitamin A supplementation, multivitamin supplements were associated with a 0.01-item (95% CI: –1.0 to 0.9; P = .03 for interaction) decrease in raw motor score compared with those who did not receive multivitamin supplements. For the PDI, P = .23 for interaction, although the same trend of a suggested effect of multivitamin supplementation only among those who did not take vitamin A supplementation simultaneously was observed. Vitamin A supplementation did not modify the effect of multivitamin supplementation on mental scores.
Maternal CD4 count at baseline was an important modifier of the effect of vitamin A supplementation on infant mental and psychomotor development. Among children whose mother's CD4 count at baseline was <200 cells/mm3, vitamin A supplementation was associated with a 10.2-point (95% CI: 4.8–15.5) increase in MDI and a 12.1-point (95% CI: 4.2–20.0) increase in PDI compared with those who did not receive vitamin A supplements. Among children whose mother's CD4 count at baseline was 200 cells/mm3, vitamin A supplementation was associated with a 0.5-point (95% CI: –1.6 to 2.5; P = .0009 for interaction) increase in MDI and a 1.0-point (95% CI: –1.6 to 3.6; P = .009 for interaction) increase in PDI compared with the group that did not receive vitamin A supplements. The same pattern of effect modification by maternal CD4 count at baseline was observed when the raw mental (P = .004 for interaction) and motor scores (P = .05 for interaction) were considered. Maternal total lymphocyte count at baseline, multivitamin supplementation, the child's HIV-1 status, the child's birth weight, timing of initiation, and duration of supplementation during pregnancy were not significant modifiers of the effect of vitamin A supplementation on mental or motor function. Considering modifiers of vitamin A supplementation at 6 months, maternal CD4 count was a significant modifier of the effect of vitamin A supplements on motor scores (P = .009), with the effect of vitamin A supplements apparent only among children whose mothers had CD4 <200 cells/mm3 at baseline, as seen in the overall analyses.
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The design of this study does not permit the effects of prenatal and postnatal multivitamin supplementation to be distinguished because all of the women received the supplements throughout. Neither is it possible to know which nutrient produced the beneficial effect observed because they all were in the same tablet. B-complex vitamin supplements,26 as well as riboflavin supplementation specifically,27, 28 have previously been associated with improvements in arm-hand steadiness tests, a measure of fine motor skills. This improvement may be mediated through riboflavin's effect on pyridoxine metabolism. Vitamin C, another component of our multivitamin supplementation, has been shown to improve absorption of dietary iron29, 30 and may help to avoid iron deficiency anemia in children. Such anemia is associated with developmental delays.31, 32
Psychomotor development improvements may have been through reducing the risks of LBW and being small for gestational age during the prenatal supplement phase. As Table 1 demonstrated, a significantly smaller proportion of children whose mothers received multivitamin supplements had LBW (<2500 g). However, LBW was not a significant effect modifier of the effect of maternal multivitamin supplementation on motor scores in this study. Another potential mechanism during the prenatal supplement phase is through the reduction of mother-to-child transmission of HIV-1, which has been associated with developmental delays, particularly psychomotor delays, as well as neurologic complications and immune dysfunction. However, multivitamin supplements were associated only with mother-to-child transmission in this population among children who were born to women who were immunologically or nutritionally compromised.24
In addition, improvements in the mother's immune status13 may result in less morbidity, more energy, and ability to stimulate her child, resulting in improvements in developmental scores. In a similar way, improvement of child multivitamin status throughout the supplement period, including observed increases in serum vitamins B12 and E in these children,16 may also be associated with reduced morbidity, particularly among initially nutritionally deficient children,33 and as a result affect their mental and psychomotor development.34 Both vitamins E and C are antioxidants that are believed to enhance immune responses and phagocytosis. Children who have increased morbidity may lack energy and withdraw from contact with their peers and environment.35, 36 Also, mothers who coddle less mobile infants may further hinder their growth and independence.35, 36 Finally, the mother of a sick child may engage in less responsive communication with the child and thereby deprive him or her of social interaction.35–37
The effect of multivitamin supplements on motor scores was variable across subgroups. Children who were born to women with a proxy for advanced disease, ie, high ESR, benefited significantly from multivitamin supplements with respect to psychomotor function. Another marker of maternal advanced disease, CD4 count, was not a significant modifier of the multivitamin supplementation effect, although a trend of an effect primarily in the women with advanced disease, CD4 <200, was observed. These were secondary analyses, as were all analyses of effect modifiers, and could be attributable to chance or the result of small sample sizes. However, the observation that multivitamin supplements are protective for psychomotor development in the subgroup of children who are born to women with advanced disease is consistent with the associations reported for multivitamin supplementation with other outcomes in this same population.24 These secondary analyses do suggest that additional investigation of these modifiers is needed.
Vitamin A supplementation was associated with beneficial effects on motor scores at 6 months. Vitamin A supplements may improve the child's immune function and reduce oxidative stress. These effects may lead to reduced infections and improved integrity of the epithelial lining of the child's gastrointestinal tract.38 In turn, absorption of nutrients by the infant may be improved, and the child may be more interactive, able to learn, and achieve higher mental and psychomotor development scores.
Within the original study, every attempt was made to minimize loss to follow-up, through scheduled appointments at the study clinic; reimbursement for transport; free access to physicians, nurses, and counselors; and free medicine for those who could not afford it. Even so, loss to follow-up might be associated with higher morbidity/mortality, thus preferentially excluding children of lower nutritional and socioeconomic status. The population of children who provided at least 1 BSID-II assessment have previously been compared with the larger population of singletons who were born to HIV-1–infected mothers and were found to be at lower risk for LBW and had a smaller hazard of dying in the first 18 months.39 At the very least, this loss to follow-up reduces the study's power to detect a significant treatment effect. In particular, this study had low power to consider effect modification. It is possible that other factors considered in this article are indeed effect modifiers but were unable to reach statistical significance in this study.
The BSID-II provides us only with an overall cognitive score (MDI) and motor score (PDI) and does not allow us to investigate the effect of supplementation on components of mental and psychomotor development. The BSID-II test was developed and validated in the United States, and there are no normative data for the BSID-II in Kiswahili-speaking societies; thus, there could be significant cultural biases that at least partly explain the overall decrease in neurodevelopmental scores among children in this study over time. Despite this, the tasks that are required by the test battery for a child up to 2 years of age can be more readily applied globally than later items might be, so the issue of language may be less relevant to this particular study given the age range of the children being assessed. The design of the BSID-II also does not fully capture the significant interactions between child and caregiver that would contribute to the total score. Such data were not collected in this study.
To our knowledge, this is the first study to examine the effect of maternal multivitamin supplementation on infant mental and psychomotor development. This study strengthens the current recommendation to give multivitamin supplements (including vitamins B, C, and E) to HIV-1–infected pregnant women in developing country settings, not only to delay HIV disease progression and to reduce the risk for LBW, prematurity, and small for gestational age but also to support infant psychomotor development. However, it is important to underscore that the multivitamin supplements should not be considered as an alternative to antiretroviral therapy but as a complementary intervention that is part of a comprehensive care package. Individuals who are advanced enough in their disease to warrant antiretroviral therapy as per national guidelines should be provided with antiretroviral drugs.
|
ACKNOWLEDGMENTS |
|---|
We thank the mothers and their children; the field teams including the nurses, midwives, supervisors, and laboratory staff; and the administrative staff who made the study possible. In particular, we appreciate the dedication of Lightness Massam and Evaustina Nyanda, who performed all of the BSID-II assessments. We also thank Muhimbili Medical Centre, Muhimbili University College of Health Sciences, and the National AIDS Control Program in Dar es Salaam for their institutional support.
Hoffmann-La Roche donated the raw material used for preparing the vitamin and placebo tablets. None of the sponsors of the study had any role in study design, data collection, data analysis, data interpretation, or writing of the report.
|
FOOTNOTES |
|---|
Address correspondence to Nuala McGrath, ScD, c/o Wafaie Fawzi, Department of Nutrition, Harvard School of Public Health, 667 Huntington Ave, Boston, MA 02115. E-mail: mina{at}hsph.harvard.edu
The authors have indicated they have no financial relationships relevant to this article to disclose.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2006 American Academy of Pediatrics. All rights reserved. |
||
|
||
