| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH COMMUNICATION |
1 From the International Centre for Diarrhoeal Disease Research, Bangladesh, Dhaka, Bangladesh (FT, SEA, JDH, and FM); International Maternal and Child Health, Uppsala University, Uppsala, Sweden (LAP and E-CE); the Institute of Nutrition and Food Science, University of Dhaka, Dhaka, Bangladesh (SNH); and the Centres for International Child Health and the Centre for Paediatric Epidemiology and Biostatistics (DR), Institute of Child Health, University College London, London, United Kingdom (SMG-M)
2 Supported by The International Centre for Diarrheal Disease Research, Bangladesh; the UK Medical Research Council; the Swedish Research Council; the UK Department for International Development; the Global Health Research Fund, Japan; the Child Health and Nutrition Research Initiative; Uppsala University; the US Agency for International Development, under the Cooperative Agreement #388-G-00-02-00125-00; the Australian International Development Agency; the Government of Bangladesh; the Canadian International Development Agency; The Kingdom of Saudi Arabia; the Government of the Netherlands; the Government of Sri Lanka; the Swedish International Development Cooperative Agency; and the Swiss Agency for Development and Cooperation. 3 Reprints not available. Address correspondence to F Tofail, Child Development Unit, Clinical Sciences Division, ICDDR,B, 68 Shaheed Tajuddin Ahmed Sharani Mohakhali, Dhaka 1212, Bangladesh. E-mail: ftofail{at}icddrb.org.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objective: We aimed to compare the effects on infant development of early (8–10 wk gestation) or usual (
17 wk gestation) supplementation with food and MM, 30 mg Fe + 400 µg folate, or 60 mg Fe + 400 µg folate.
Design: A large, randomized, controlled trial of pregnancy supplementation was conducted in Bangladesh. A subsample of infants (n = 2853) were assessed on 2 problem-solving tests (support and cover tests), the motor index of the Bayley Scales of Infant Development, and Wolke's behavior ratings at 7 mo of age.
Results: There were no significant effects of any intervention in the group as a whole. However, infants of undernourished mothers [body mass index (BMI; in kg/m2) < 18.5] who received early food supplementation performed slightly but significantly (P = 0.035) better on the support test than did infants of mothers who received usual food supplementation (z score: 0.17; 95% CI: 0.01, 0.33). There were no benefits in infants of higher-BMI mothers (P = 0.024 for BMI x food interaction). Children of low-BMI mothers who received MMs had slightly better motor scores (z score: 0.28; 95% CI: 0.08, 0.48) and activity ratings (z score: 0.24; 95% CI: 0.037, 0.45) than did those who received 30 mg Fe + 400 µg folate, whereas other children did not benefit (P = 0.05 for both motor scores and BMI x micronutrients and for activity and BMI x micronutrients).
Conclusions: Small benefits from early food and MM supplementation were found in infants of low-BMI but not of high-BMI mothers. However, the benefits were of doubtful functional importance, and longer follow-up is required to determine programmatic implications.
Key Words: Maternal nutrition • pregnancy • food and micronutrient supplementation • problem-solving test • psychomotor development • Bangladesh • infants • body mass index
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
There is concern that deficiencies of macronutrients (12-14) and micronutrients (15-17) during pregnancy also may detrimentally affect the neurodevelopment of offspring. Animal research has shown that protein-energy deficiency during gestation may cause permanent alteration of brain structures and impaired cognitive function (18). In humans in developed countries, food supplementation in apparently healthy pregnant women had no effect on children's development (19-21), whereas studies from developing countries showed mixed results. In Taiwan, supplementation of marginally undernourished women throughout pregnancy and lactation benefited their infants’ motor but not mental development (22) or later intelligence (23). In Colombia, supplementation during the last trimester of pregnancy did not benefit the children's mental development (24).
Information on the effect of MM deficiencies in pregnancy on children's development is limited (25). Supplementing women with either 16 or 4 micronutrients during the periconceptional period made no difference to the offsprings’ cognitive development (26); however, the children whose mothers had received 16 micronutrients had lower anxiety at age 6 y (15).
A multicenter trial of the effect of giving 15 MMs during pregnancy on birth outcomes was launched in 6 countries (4). The trial in Bangladesh was integrated into the Bangladesh Integrated Nutrition Program (BINP), which provided food supplementation during pregnancy through community nutrition centers. This trial aimed to compare the effect on birth outcomes of MM supplementation during pregnancy with that of different doses of iron and folate supplementation (30 or 60 mg Fe in addition to 400 µg folate) and the effect of early (ie, periconceptual) food supplementation with that of usual food supplementation during pregnancy.
We took the opportunity to examine the effects of the supplements on children's cognitive and motor development and behavior at age 7 mo in a subsample. We hypothesized that the above treatment would benefit the infants of malnourished mothers more than those of better-nourished mothers.
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
85% of the population is Muslim. The International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR,B) has a Health and Demographic Surveillance System that has recorded vital demographic information in the area since 1966.
Intervention
With the use of the Health and Demographic Surveillance System, all pregnant women were identified within 6–8 wk of conception through regular home visits, and their pregnancies were confirmed by urine tests and ultrasonography. After enrollment, women were randomly assigned to 6 intervention groups—2 food groups, each of which was subdivided into 3 micronutrient groups.
Food
In the early-food group, women were urged to take supplementation as soon as pregnancy was identified, whereas, in the usual-food group, women were left to choose when to start. Food packets containing roasted rice powder (80 g), roasted lentil powder (40 g), molasses (20 g), and soy oil (12 mL) provided 600 kcal (2512 kJ) of energy. All enrolled pregnant women were encouraged to consume the supplements under supervision 6 d/wk at the local community nutrition center. The number of packets taken by the mothers was recorded during monthly home visits.
Micronutrients
Three types of micronutrients were provided for daily supplementation from week 14 of gestation until delivery; they included MMs, 60 mg Fe (fumarate) + 400 µg folate (the Fe60 group), and 30 mg Fe (fumarate) + 400 µg folate (the Fe30 group). The MM group received 15 different vitamins and minerals, developed by the United Nations Children's Fund (UNICEF); 1 tablet contained 150 µg I (potassium iodide), 15 mg Zn (sulfate), 65 µg Se (sodium selenite), 2 mg Cu (sulfate), 800 µg retinyl acetate (RE) vitamin A, 1.4 mg thiamine mononitrate, 1.4 mg vitamin riboflavin, 18 mg vitamin B-3 (niacin), 1.9 mg vitamin B-6 (pyridoxine hydrochloride), 2.6 µg vitamin B-12 (cyanocobalmin), 70 mg vitamin C, 200 IU vitamin D (vitamin D3), and 10 µg vitamin E (
-tocopherol acetate) in the recommended dietary allowance dose in addition to 30 mg Fe (fumarate) and 400 µg folate. The pills were identical in appearance, and monthly supplies were provided in identical bottles with electronic counting devices (eDEM; AARDEX Inc, Zurich, Switzerland), which had small microprocessors embedded in the caps. The microprocessors recorded the number of times the bottles were opened (27).
Sample
All live singleton babies born from 20 May 2002 to 20 December 2003 (n = 2853) constituted our subsample; this number was sufficient to allow detection of a difference of 4 developmental quotient points among the groups with 90% power at the 5% level.
Written informed consent was obtained from the parents or guardians of all of the children. The study was approved by the Ethics Review Committee of ICDDR,B.
Measurements
Socioeconomic status
On enrollment of the mothers, information was collected during a home visit. This information concerned their family wealth (number of possessions—eg, television, radio, domestic animals, chairs, tables, beds, bicycle, or rickshaw), deficits between income and expenditure (occasional or constant—yes or no), family structure and parental characteristics (mother's age, hemoglobin, and parity and parents’ education and employment), and housing quality (floor, walls, or roof made of mud—yes or no).
Anthropometry
On enrollment, maternal weight was measured with electronic scales (UNISCALE; Seca GmbH, Hamburg, Germany) that were accurate to 100 g, and height was measured to the nearest 0.1 cm by using a stadiometer. Infant birth weight was measured within 72 h after birth by using beam scales (Seca GmbH), which were precise to 10 g. Research assistants were trained to carry out the anthropometric measurements according to standard procedures (28).
Developmental assessments
When the children were 7 mo old, psychologists who were unaware of the children's groups assessed the children's development in the presence of their mothers at 1 of 4 local clinics. Two one-step means-end problem-solving tests (PSTs), the support and cover tests, were used to assess cognitive development, and the Psychomotor Developmental Index (PDI) of the Bayley Scales of Infant Development-II was used to assess motor development (29). The original procedures of the PSTs were described by Piaget (30), but the conduct and scoring of the tests were designed by Willatts (31, 32). In these tests, infants manipulate an intermediary to retrieve an object (a toy).
The support test involves placing a long cloth on a table in front of the child and then placing a toy out of the child's reach at the farthest end of the cloth. The infant has to pull the cloth to retrieve the toy. In the cover test, a toy is covered with a cloth while the infant is watching. The infant is then required to remove the cloth to retrieve the toy. These procedures were videotaped and scored later. Four trials were given in both PSTs. In each trial, 3 behaviors—cloth behavior (the way the child handled the cloth), fixation behavior (the way the child fixed his or her vision on the toy), and toy behavior (the way the child grasped the toy)—were scored on a 3-point scale: 0 for no evidence of intention, 1 for possible or ambiguous intention, and 2 for clear evidence of intention. The scores for each behavior were summed to give an intention score for each trial that ranged from 0 to 6, and the scores of 4 trials were then summed to give a total score ranging from 0 to 24. These tests were chosen because they are relatively easy and quick to perform, and because they can be scored from videotape, which facilitates ongoing quality control. The tests have been used in Jamaica (33), where they differentiated LBW from normal-birth-weight infants, were sensitive to an 8-wk intervention of stimulation, and predicted the developmental quotients at the age of 15 mo.
The Bayley Scales of Infant Development-II has been used in Bangladesh by the same group of researchers (34-36); it showed good interobserver and test-retest reliability (36). We also used a modified version of Wolke's behavior rating scale (37) to assess infant behavior during the assessments. This instrument has 5 ratings on 9-point scales that included the following items: the infant's activity level (very still = 1 to overactive = 9), emotional tone (unhappy = 1 to radiates happiness = 9), response to examiner in first 5 min (avoiding = 1 to friendly and inviting = 9), cooperation with the test procedure (resists all suggestions = 1 to always complies = 9), and the amount of vocalization (very quiet = 1 to constant vocalization = 9). These scales also were used previously in Bangladesh (34, 36).
At the beginning of the study, interobserver reliabilities among the 5 testers were assessed on 20 infants. Intraclass correlations for scores of both support and cover PSTs and Bayley PDI were >0.95, and those for all behavior ratings were >0.89. Test-retest reliabilities were assessed before the start of the study. Test-retest intraclass correlations over 24 h for both support and cover PST scores of 15 infants were >0.70 and those for the Bayley PDI in 10 infants over a 7-d interval were 0.9. To maintain the quality of testing, ongoing interobserver reliability was assessed weekly on 7% of all tests (PSTs and Bayley PDI: >0.90; all behavior ratings: >0.85).
Statistical analysis
We examined differences in enrollment characteristics between the lost and tested samples by using analyses of variance or chi-square tests. We conducted intention-to-treat analyses to assess the intervention effect on all child development outcomes. A 2-factor analysis of covariance (ANCOVA) was conducted by using micronutrients (3 levels) and food (2 levels) as factors and age as the covariate (where scores differed by age). Scores on the support test, activity, emotion, and cooperation ratings varied by sex, and therefore sex was used as an additional factor for all analyses of these outcomes.
To examine whether thin mothers responded differently to supplements than did better-nourished women, we examined the BMI x treatment interaction. We divided the sample into infants of mothers with high BMIs and infants of mothers with low BMIs. We took the World Health Organization's definition of BMI [ie, <18.5 for mild undernutrition (38)] as the cutoff for low BMI, and high BMI was 
18.5. This same value was used by the BINP program to determine which pregnant women to provide with food supplement.
We repeated all ANCOVAs by using maternal BMI (2 levels) as an additional factor. Post hoc Bonferroni tests were conducted to examine any significant difference among the 3 micronutrient groups. When BMI x treatment interactions were significant, appropriate subgroup analyses were performed. All statistical analyses were performed with the use of SPSS software (version 11.5; SPSS Inc, Chicago, IL).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). The main causes for loss to follow-up were being away from home on visits or because of floods (40%); refusal to be tested (27%), probably because of fear of venipuncture; death (16%); moving to a different residence (8%); and being sick at the time of testing (9%). The proportion of lost infants was similar in each of the 6 groups (NS, chi-square test). We first examined any differences between lost and tested children in enrollment characteristics by using a 2-factor analysis of variance (lost or tested by 6 supplement groups). There was no significant difference between lost and tested children in any of the 6 groups in the following characteristics: family income-expenditure deficits; structure of house; mothers’ and fathers’ educational level; mothers’ age, height, and BMI on enrollment; infants’ birth weight; and children's age at testing. Overall, mothers of children lost to follow-up had significantly (P = 0.01) fewer children (1.13 ± 1.11) than did mothers whose children were tested (1.26 ± 1.1) and also had significantly (P = 0.02) higher hemoglobin concentrations at enrollment (11.8 ± 1.3 g/L) than did mothers whose children were tested (11.7 ± 1.3 g/L). Among birth characteristics, infant gestational age was significantly lower in the lost children than in those tested (39.0 ± 2.1 and 39.2 ± 1.7 wk, respectively; P = 0.03). However, all differences were small, and none were significantly different across the 6 supplement groups (all 6 lost or tested x supplement group interactions were NS).
|
. Male infants constituted approximately one-half of the sample. The families were extremely poor; only one-third of the families had 1 adult in regular employment, and 19% of families had experienced occasional or continuous income-expenditure deficits in the previous year. There was no significant difference among the 6 groups in any measured enrollment or birth characteristics.
|
Intent-to-treat analyses
The means (±SDs) of the developmental outcomes are shown in Table 2. Age was significantly positively correlated with support (r = 0.08, P < 0.001) and cover (r = 0.07, P < 0.05) and significantly negatively correlated with the PDI (r = –0.21, P < 0.001). The male infants had significantly better total scores on the support test than did the females (11.5 ± 7.6 and 10.6 ± 7.6, respectively; P = 0.02).
|
Effect of supplementation stratified on mothers’ BMI in early pregnancy
We had initially hypothesized that supplementation would benefit undernourished pregnant women more than better-nourished pregnant women. Accordingly, women were divided into 2 groups according to their BMI on enrollment at 8–10 wk of pregnancy—the better-nourished (BMI 18.5) and the undernourished (BMI < 18.5) groups—and a second series of analyses of variance was conducted with the addition of BMI group as a factor (Table 2
).
Problem-solving tests
Maternal BMI had a significant effect on support (P = 0.018) and cover (P = 0.039) test scores. In both tests, the infants of undernourished mothers had slightly but significantly lower scores than did the infants of better-nourished mothers [support test: 10.25 ± 7.85 and 11.27 ± 7.49, respectively (P = 0.006); cover test: 12.31 ± 7.14 and 13.10 ± 6.97, respectively (P = 0.022)]. In the support test, the BMI x food group interaction was significant (P = 0.024), and it approached significance in the cover test (P = 0.058). The 3-way interaction of the 2 groups of treatment was not significant in either problem-solving test. The mean scores of the support test in children of low-BMI mothers in the early- and usual-food groups were 10.9 ± 7.9 and 9.6 ± 7.8, respectively, which gave an effect size (z score) of 0.17 (95% CI: 0.01, 0.33; P = 0.035). In contrast, in the children of the high-BMI mothers, the early- and usual-food group mean scores were 11.06 ± 7.53 and 11.49 ± 7.45, respectively, which gave an effect size (z score) of –0.06 (95% CI; –0.16, 0.04; P = 0.25) (Figure 2). There was no significant BMI x micronutrient interaction in any of the PSTs.
|
). In contrast, there was no effect in the high-BMI group. The interactions of food with BMI or with food and micronutrients were not significant.
|
To determine whether grouping the children according to the mother's BMI status (low or high) had introduced confounding, we compared the characteristics of the treatment groups in low-BMI mothers only. There were no significant differences among the groups in any of the enrollment characteristics shown in Table 1 or in the age and sex of the children.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The present trial was a large, randomized, controlled trial; the testers were unaware of children's groups; the mothers were unaware of their micronutrient supplement; and the tests were reliable. The loss of subjects was not biased, and we had initially hypothesized that low-BMI mothers were more likely than higher-BMI mothers to benefit from supplementation. It is therefore likely that the benefits can be attributed to the treatments. The population was a highly suitable one in which to test the effect of supplementation, because many of the mothers were undernourished and the children had a small average birth weight and a high incidence of LBW.
Food supplementation
The average time difference in commencing food supplements between the early- and usual-food groups was 31 d, and 10% of the early-food group overlapped with some of the usual-food group in starting time. The amount of food supplements reportedly taken by the early- and usual-food groups differed by a total of 18 600 kcal (77 875 kJ). However, it is possible that the amount actually consumed by the mothers was less and that the food was replaced or shared with other family members in these poor households (39). We cannot be certain whether differences in timing or total amount received (or in both) were the important factor. It is surprising that this relatively small difference in supplement had an effect on infants’ development. Had there been a group not receiving supplement, it is likely that the difference would have been considerably greater. To establish the full value of food supplementation, the need remains to examine its effects, when given to undernourished women throughout pregnancy, on the development of their children.
In Taiwan, food supplementation throughout pregnancy benefited infants’ motor but not mental development at the age of 8 mo more than did a placebo (23). However, the investigators did not analyze the benefits according to the mothers’ BMI. They used the Bayley Scales to assess mental development, and it is possible that the PSTs that we used are more sensitive to subtle changes than are the Bayley Scales.
Micronutrient supplementation
The motor development of infants of undernourished mothers also benefited from MMs, whereas that of infants of better-nourished mothers did not. Mothers with low BMIs also may be deficient in MMs. In a previous trial in the same community, infants supplemented with MMs showed greater improvements in motor scores than did infants receiving riboflavin (35). In the present study, in addition to micronutrients, all mothers received food supplementation, and, therefore, we should be cautious before extrapolating the findings to undernourished populations to whom food supplementation is not available.
It is possible that more subtle changes to the brain, such as improvement in neurotransmitters, dendritic branching, connections, and myelination, may have occurred (40, 41). However, the size of the benefit in development from both MMs and early-food supplementation is very small, and it is not certain whether there will be any functional advantage.
The consistent association of mothers’ BMI with both cognitive and motor function provides further evidence of the importance of maternal nutrition and child development. From a policy perspective, it appears that food supplementation should begin as early as possible in pregnancy and that it should be targeted toward undernourished women; moreover, efforts should be made to improve compliance. In addition, in countries where there are no food supplementation programs and where large numbers of women are undernourished, it is likely that the offspring are at high risk of poor development. In contrast to food supplementation, benefits of prenatal MM supplementation on birth outcomes are less clear; this is the first study that we know of that examined the effects on child development; and the benefits were limited to motor development and were small. Longer-term follow-up is required to determine whether any of the observed benefits have functional importance.
|
ACKNOWLEDGMENTS |
|---|
The authors’ responsibilities were as follows—LAP and SEA: the conduct of the main study; E-CE: designing and implementing the micronutrient component; SMG-M, FT, JDH, and SNH: writing the proposal, designing the child development portion of the study, supervision of the conduct of the study, and data analysis; FT, JDH, SNH, and FM: subject recruitment, data collection, and logistical organization of the child development portion of the study; DR: statistical analysis; and all authors: review and critique of the manuscript drafts.. This report is a part of FT's doctoral research work. None of the authors had a personal or financial conflict of interest.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  Q. Li, H. Yan, L. Zeng, Y. Cheng, W. Liang, S. Dang, Q. Wang, and I. Tsuji Effects of Maternal Multimicronutrient Supplementation on the Mental Development of Infants in Rural Western China: Follow-up Evaluation of a Double-Blind, Randomized, Controlled Trial Pediatrics, April 1, 2009; 123(4): e685 - e692. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or usual- (
) food group. Early-food group: low BMI, n = 278; high BMI, n = 777. Usual-food group: low BMI, n = 309; high BMI, n = 742. Support test: food x BMI interaction, P = 0.024; low BMI early food versus usual food, P = 0.035; high BMI early food versus usual food, NS (ANOVA). Cover test: food x BMI interaction, P = 0.058 (ANOVA).
, 60 mg Fe + folate (F);