American Journal of Clinical Nutrition, Vol. 83, No. 5, 1112-1117,
May 2006
© 2006 American Society for Nutrition
ORIGINAL RESEARCH COMMUNICATION |
Effect of iron supplementation during pregnancy on the intelligence quotient and behavior of children at 4 y of age: long-term follow-up of a randomized controlled trial1,2,3
Shao J Zhou,
Robert A Gibson,
Caroline A Crowther,
Peter Baghurst and
Maria Makrides
1 From the Child Nutrition Research Centre and the Department of Pediatrics (SJZ, RAG, and MM), the Department of Obstetrics & Gynaecology (CAC and MM), and the Departments of Paediatrics and Public Health (PB), Women's & Children's Hospital, University of Adelaide, North Adelaide, Australia.
2 Supported by the National Health & Medical Research Council (ID: 250431) and Channel 7 Children's Medical Research Foundation. The iron and placebo tablets used in the original trial were manufactured and donated by Soul Pattinson Manufacturing, Kingsgrove, NSW, Australia.
3 Reprints not available. Address correspondence to MM, Child Nutrition Research Centre, Level 1, Rieger Building, Women's & Children's Hospital, 72 King William Road, North Adelaide, SA 5006, Australia. E-mail: maria.makrides{at}cywhs.sa.gov.au.
 |
ABSTRACT
|
|---|
Background: Iron supplements are often prescribed during pregnancy despite the lack of intervention trials that have assessed the effects of supplementation in pregnancy on childhood development.
Objective: The objective was to determine whether iron supplementation during pregnancy influences childhood intelligence quotient (IQ) in an industrialized country.
Design: Pregnant women (n = 430) were randomly allocated to receive iron (20 mg/d) or placebo from 20 wk gestation until delivery, and the women and their children were followed up over the long term (4 y). Seventy percent of these families participated in the follow-up. The proportion of women with iron deficiency anemia at the end of pregnancy was 1% (2 of 146) in the iron group and 11% (15 of 141) in the placebo group. The primary outcome was the IQ of the children at 4 y of age, as assessed by the Stanford-Binet Intelligence Scale. Secondary outcomes included child behavior and the general health of the mothers.
Results: The mean IQ was not significantly different (P = 0.980) between the children of the iron-supplemented mothers (109 ± 11; n = 153) and the children of the mothers in the placebo group (109 ± 11; n = 149). However, the percentage of children with an abnormal behavior score was higher in the iron group (24 of 151, or 16%) than in the placebo group (12 of 149, or 8%); the relative risk was 1.97 (95% CI: 1.03, 3.80; P = 0.037). There was no significant difference in the health of the mothers between groups, as assessed by the SF-36 Health Survey.
Conclusions: Prenatal iron supplementation that reduces the incidence of iron deficiency anemia from 11% to 1% has no effect on the IQ of the offspring at 4 y of age.
Key Words: Iron supplementation • pregnancy • child development • intelligence quotient • IQ • behavior • industrialized country
|
INTRODUCTION
|
|---|
Routine iron supplementation is a common practice for preventing iron deficiency (ID) and iron deficiency anemia (IDA) in pregnancy, because the dietary iron intake of pregnant women often does not meet the recommended dietary intake. However, expert opinion on whether iron should be given routinely to pregnant women is divided. In countries such as the United States and France, pregnant women are often advised to routinely take iron supplements of 30 to 60 mg/d, whereas the policy in Australia and the United Kingdom is to screen and treat pregnant women with IDA. The Cochrane systematic review (1) and the review conducted by the US Preventive Task Force (2) both concluded that, although iron supplementation in pregnancy improves maternal iron status, there is a lack of evidence of benefit in terms of clinical outcome measures. In particular, the US Preventive Task Force has highlighted the need to assess the long-term effects of iron supplementation on child development as an area of priority for research (3).
Evidence from animal studies has consistently shown that inadequate iron nutrition during pregnancy leads to permanent structural and functional changes in the brain of offspring (4–6). Although the degree of ID induced in experimental animals is often more severe than in pregnant women, no human intervention trials have been specifically designed to investigate the effect of iron nutrition in pregnancy on child development. The aim of our study was to assess whether improved iron nutrition in pregnancy, through routine iron supplementation, influences the development of the children and the long-term health of mothers.
|
SUBJECTS AND METHODS
|
|---|
Participants
Mothers and children who participated in a double-blind randomized controlled trial of iron supplementation in pregnancy called the Australian Mothers' and Babies’ Iron Trial (7) were eligible for follow-up. In summary, 430 pregnant women receiving antenatal care at the Women's & Children's Hospital in Adelaide, Australia, were recruited to participate in the original trial during 1997–1999 (7). The participants were randomly allocated to receive either iron or placebo from 20 wk of gestation until birth. The dose of iron used in the trial was 20 mg/d, which was intended to increase the women's usual iron intake from 
12 mg/d (8) to the Recommended Dietary Intake for iron during pregnancy in Australia (22–36 mg/d) (9). The compliance rate was 86% for both groups based on tablet back-count and monthly phone calls (7). Women in the iron group had higher concentrations of hemoglobin and serum ferritin and lower incidences of ID and IDA at the end of pregnancy (Table 1
). The follow-up was conducted from May 2002 to January 2004, 4 y after birth. The 4-y follow up was chosen because developmental assessment at preschool age is considered a better predictor of school achievement than is the developmental assessment in the first 2 y of life—the time when most studies investigating the relation between ID or IDA and child development have been conducted. For the remainder of the report, the children are referred to as the iron group or the placebo group based on the group allocation of their mothers' in the original trial. This study was approved by the Human Research Ethics Committee at the Women's & Children's Hospital. Informed consent was obtained from all participants.
Assessments
The primary outcome was childhood intelligence quotient (IQ), assessed by using the Stanford-Binet Intelligence Scale (10), which was administered by qualified experienced psychologists. The Stanford-Binet Intelligence Scale is an internationally standardized tool for global developmental assessment, which has 4 subscales of IQ (verbal reasoning, visual reasoning, quantitative reasoning, and short-term memory) as well as a combined composite IQ (10). This scale has high internal consistency and test-retest reliability (10), and it has been used to assess the effect of IDA on child development for children aged >3 y (11) rather than the Bayley Scales (12), which are more often used in younger children. Most of the IQ assessments were conducted in a clinic room at the Women's & Children's Hospital. For a small number of families (n = 47) who were unable to attend the hospital clinic, the assessment took place in a private room of a child care center or preschool where the child was attending. Child behavior was assessed by using the Strength and Difficulties Questionnaire (SDQ) parent report form (13), which is a behavior-screening questionnaire used to assess behavioral problems such as emotion, conduct, hyperactivity, peer relation, and social behavior. The SDQ has been shown to have moderate-to-strong internal reliability, test-retest reliability, and external validity (14, 15). A total difficulties score of 
17 indicates abnormal behavior (16). The weight and height of the children were measured with an electronic scale and a stadiometer, respectively, by using standardized techniques. Information was also collected regarding sex, gestational age at birth, birth order, duration of breastfeeding, parental education, and quality of home environment with the Home Screening Questionnaire (HSQ) (17), which is a screening tool used to identify home environments likely to be suboptimal for the development of children. The HSQ has high internal reliability and test-retest reliability (18). These variables were assessed as potential confounders of the relation between iron supplementation and IQ and child behavior. Attendance at preschool was not considered to be a potential confounding variable because all Australian children attend preschool.
In addition, the general health of women was assessed by using the SF-36 (19), a self-administrated questionnaire that assesses 8 concepts of health and has been used in studies that examine the effect of ID and the general health of women (20, 21). Information on nonpregnancy-related health problems or hospital admissions since the completion of the original trial was also collected. If the women had any subsequent pregnancies, outcomes of the subsequent pregnancies were collected from medical records. Families and research staff involved in data collection were blinded to the group assignment until all primary analyses of the 4-y data had been completed.
Statistical analysis and sample size estimation
The data were analyzed by using SPSS software (version10.0; SPSS Inc, Chicago, IL). The primary analyses were based on intention to treat. Childhood IQ, behavior, and health of the women were compared between the iron and placebo groups by using independent-sample t tests. Differences between categorical variables were compared using chi-square tests. Multiple regression analyses were conducted for continuous outcomes, and logistic regression was conducted for dichotomous outcomes as secondary analyses to examine the effect of intervention on childhood IQ and behavior when potential IQ covariates were controlled. Statistical significance was set at a P value < 0.05 for all statistical tests. Our sample size estimation was based on IQ with an assumed SD of 16 (10). We initially estimated that 186 children in each group were needed to detect a minimum difference of 5 IQ points between the iron and placebo groups with 85% power. A 5-IQ point difference was considered clinically significant because it is of the order of magnitude associated with IQ differences in children who were fed breast milk rather than formula as infants (22) or were exposed to high lead concentrations (23). Differences of this magnitude have prompted public health authorities to promote human milk feeding and to guard against lead exposure. Halfway through the study, we reevaluated the sample size estimate using an SD of 12, which was typical of our study sample. One hundred six children in each group were required to achieve an equivalent difference with the same power.
|
RESULTS
|
|---|
Characteristics of the participants
Seventy percent of the mothers and children from the original trial were assessed 4 y after birth (Figure 1). At follow-up, there were no differences in the social and demographic characteristics between families from the iron and placebo groups (Table 1
). The key difference between the iron and placebo groups was the higher iron status of the iron-supplemented women than of the placebo group at the end of pregnancy. Although 30% of families did not participate in the follow-up, there were no significant differences in the sociodemographic characteristics between participants and nonparticipants, except that more women who did not attend the 4-y follow-up smoked during pregnancy [35 of 128 (27%) compared with 50 of 302 (17%); P = 0.012]. However, the proportion of mothers who smoked during pregnancy did not differ between the iron and placebo groups [25 of 153 (16%) compared with 25 of 149 (17%); Table 1].
IQ and behavior of the children
The mean IQ of the children at 4 y was 109 ± 11. There were no significant differences between the iron and placebo groups in either composite IQ or any subscales of IQ or in the proportion of children whose IQ fell 1 or 2 SD below the mean (Table 2). The mean scores for behavioral difficulties did not differ significantly between the groups. However, more children in the iron group than in the placebo group had an abnormal total behavior score [24 of 151 (16%) compared with 12 of 148 (8%); Table 2]. Adjustment for covariates (including maternal age, parental education levels, quality of home environment, duration of breastfeeding, gestational age at birth, birth order, and sex of the child) did not change the outcomes for either IQ or behavior.
In an exploratory analysis, the composite IQ of the children was negatively correlated with the total behavior score (r = –0.237, P < 0.0001), and children with abnormal behavior scores had a lower mean IQ than did children whose behavior scores were in the normal range (104 ± 11 compared with 110 ± 11; 95% CI: 2, 10; P = 0.001).
General health of the mothers and outcomes of subsequent pregnancies
There were no significant differences in any of the 8 health concepts assessed with the use of the SF-36 between the iron and placebo groups (Table 3). The proportion of women who had health problems requiring medical treatment or hospital admission since the completion of the original trial (6 mo postpartum) did not differ significantly between the iron and placebo groups (Table 3). The proportion of women who had at least one subsequent pregnancy was 76 of 152 (50%) in the iron group and was 81 of 148 (55%) in the placebo group (P = 0.696). There were no significant differences in the outcomes of subsequent pregnancies in terms of gestational age at birth, birth weight, birth length, birth head circumference, or proportion of women with IDA in subsequent pregnancies (data not shown).
|
DISCUSSION
|
|---|
To the best of our knowledge, this is the first human intervention trial to evaluate the effect of iron nutrition in pregnancy on early child development. The major finding in this sample taken from an industrialized country is that routine iron supplementation in pregnancy had no benefit on childhood IQ, despite the clear benefit of iron therapy on iron status in the mothers at birth.
Our finding is in contrast with the results from the cohort study of Tamura et al (24), which was designed to investigate the developmental outcomes of children born small for gestational age. They showed that children with cord ferritin concentrations in the lowest quartile scored lower on some mental and psychomotor tests at 5 y than did children with cord ferritin concentrations in the 2 middle quartiles (24). However, children with cord ferritin in the highest quartile also scored lower on some tests. Although the authors suggested that the children with cord ferritin in the highest quartile might have a falsely elevated ferritin concentration because of possible maternal infections (24), the possibility that a high cord ferritin concentration itself may adversely affect child development cannot be excluded. The differences in results between the 2 studies may be explained by the fact that the cohort study included a high proportion of small-for-gestational-age infants or by methodologic differences inherent in cohort studies and randomized trials. Our study was a randomized comparison that included women and children that were representative of the Australian population (7, 25). The dose of supplemental iron used in our original trial (7) was low compared with common iron supplements used in pregnancy (1). However, the 3-fold differences in iron intake between the iron and placebo mothers in our original trial resulted in changes in maternal hemoglobin and ferritin that were similar to the differences in trials in which women were supplemented with 100 mg Fe/d (1). Although we cannot be sure that low-dose iron supplements increased the supply of iron to the fetus or neonate, the maternal iron-status data indicate that a higher iron dose would be unlikely to result in increased benefit.
The composite IQ of the children in our study was higher than the standardized norm (10), but it is consistent with the validation study that showed a higher IQ in Australian children, particularly at younger ages (26). Our findings are applicable to those of other industrialized countries because the dietary iron intake of pregnant women and the prevalence of IDA in pregnancy in our study population are similar to those of pregnant women in other industrialized countries (27, 28). However, our results may not be generalized to other populations in whom ID and anemia are more prevalent and more severe or undernutrition is common.
The higher frequency of children with abnormal behavior scores in the iron group than in the placebo group was unexpected, especially because the mean scores for behavioral difficulties between groups were not significantly different and were comparable with normative values for Australian children (15). We are aware that the effect may have been due to chance because behavior was a secondary outcome and there were only a small number of children with abnormal scores. Therefore, this result needs to be interpreted with caution. Although there are no other data regarding the effect of prenatal iron supplementation on behavior of children, iron supplementation in infancy in a developing country has been shown to result in a more positive social interaction compared with placebo (29). The need for a balanced evaluation of safety and efficacy in iron-supplementation trials is clearly warranted.
Routine iron supplementation in pregnancy in otherwise well-nourished women in industrialized countries had no beneficial effects on the long-term general health of mothers consistent with the only other published report by Hemminki and Merilinen (30), although routine iron supplementation in pregnancy was linked with a higher risk of stillbirth and convulsion in infants than was selective iron supplementation in the Finish trial (30).
Currently, accidental iron overdose from iron tablets remains a common cause of childhood poisoning (3). From a public health perspective, the lack of apparent clinical benefit and the potential hazards associated with routine iron supplementation in pregnancy suggest that the risks may outweigh the benefits in well-nourished populations in whom the incidence of IDA at the end of pregnancy is 11%. Further research is required to clearly delineate the population subgroups in industrialized countries that may benefit from iron supplementation in pregnancy.
|
ACKNOWLEDGMENTS
|
|---|
We thank Jacinda Fisher for conducting the Stanford-Binet Intelligence tests and Mandy O'Grady, Heather Garreffa, and Jo Collins for administrative and clinical support. The funding organizations and Soul Pattison Manufacturing had no role in the design and conduct of the study, the analysis and interpretation of the data, or the preparation, review, and approval of the manuscript.
All authors contributed to the study design. SJZ collected and analyzed the data under the supervision of MM, RAG, and CAC. SJZ wrote the manuscript with contributions from all coauthors. PB provided statistical support for part of the data analysis. None of the authors had a conflict of interest.
|
REFERENCES
|
|---|
- Mahomed K. Iron supplementation in pregnancy. Cochrane Database Syst Rev 2003;2:CD000117.
- US Preventive Services Task Force. Routine iron supplementation during pregnancy. Review article. JAMA 1993;270:2848–54.[Medline]
- US Preventive Services Task Force. Routine iron supplementation during pregnancy. Policy statement. US Preventive Services Task Force. JAMA 1993;270:2846–8.[Medline]
- Kwik-Uribe CL, Golub MS, Keen CL. Chronic marginal iron intakes during early development in mice alter brain iron concentrations and behavior despite postnatal iron supplementation. J Nutr 2000;130:2040–8.[Abstract/Free Full Text]
- Taneja V, Mishra KP, Agarwal KN. Effect of maternal iron deficiency on GABA shunt pathway of developing rat brain. Indian J Exp Biol 1990;28:466–9.[Medline]
- de Deungria M, Rao R, Wobken JD, Luciana M, Nelson CA, Georgieff MK. Perinatal iron deficiency decreases cytochrome c oxidase (CytOx) activity in selected regions of neonatal rat brain. Pediatr Res 2000;48:169–76.[Medline]
- Makrides M, Crowther CA, Gibson RA, Gibson RS, Skeaff CM. Efficacy and tolerability of low-dose iron supplements during pregnancy: a randomized controlled trial. Am J Clin Nutr 2003;78:145–53.[Abstract/Free Full Text]
- McLennan W, Podger A. National Nutrition Survey Australia 1995. Canberra, Australia: Australian Government Publishing Service, 1998.
- National Health and Medical Research Council. Recommended Detary Intakes for use in Australia. Canberra, Australia: Australian Government Publishing Service, 1991.
- Thorndike RL, Hagen EP, Sattler JM. Stanford-Binet Intelligence Scale: Technical Manual. 4th ed. Itasca, IL: The Riverside Publishing Company, 1986.
- Pollitt E, Saco-Pollitt C, Leibel RL, Viteri FE. Iron deficiency and behavioral development in infants and preschool children. Am J Clin Nutr 1986;43:555–65.[Abstract/Free Full Text]
- Bayley N. Bayley scales of infant development. 2nd ed. San Antonio, TX: Harcourt Brace & Company, 1993.
- Goodman R. Strengths and difficulties questionnaire. 1999. Internet: http://www.sdqinfo.com/questionniares/english/c2.pdf (accessed February 2002).
- Goodman R. Psychometric properties of the strengths and difficulties questionnaire. J Am Acad Child Adolesc Psychiatry 2001;40:1337–45.[Medline]
- Hawes DJ, Dadds MR. Australian data and psychometric properties of the Strengths and Difficulties Questionnaire. Aust N Z J Psychiatry 2004;38:644–51.[Medline]
- Goodman R. Scoring the informant-rated strengths and difficulties questionnaire. 1999. Internet: http://www.sdqinfo.com/ScoreSheets/el.pdf (accessed February 2002).
- Coons C, Gay E, Fandal A, Frankenburg W. The home screening questionnaire reference manual. Denver, CO: JF Kennedy Child Development Centre, University of Colorado Health Science Centre, 1981.
- Frankenburg WK, Coons CE. Home screening questionnaire: its validity in assessing home environment. J Pediatr 1986;108:624–6.[Medline]
- Ware J. SF-36 health survey: a manual and interpretation guide. Boston, MA: The Health Institute, New England Medical Centre, 1993.
- Patterson AJ, Brown WJ, Powers JR, Roberts DC. Iron deficiency, general health and fatigue: results from the Australian Longitudinal Study on Women's Health. Qual Life Res 2000;9:491–7.[Medline]
- Patterson AJ, Brown WJ, Roberts DC. Dietary and supplement treatment of iron deficiency results in improvements in general health and fatigue in Australian women of childbearing age. J Am Coll Nutr 2001;20:337–42.[Abstract/Free Full Text]
- Anderson JW, Johnstone BM, Remley DT. Breast-feeding and cognitive development: a meta-analysis. Am J Clin Nutr 1999;70:525–35.[Abstract/Free Full Text]
- Carpenter DO. Effects of metals on the nervous system of humans and animals. Int J Occup Med Environ Health 2001;14:209–18.[Medline]
- Tamura T, Goldenberg RL, Hou J, et al. Cord serum ferritin concentrations and mental and psychomotor development of children at five years of age. J Pediatr 2002;140:165–70.[Medline]
- Chan A, Scott J, McCaul K, Keane R. Pregnancy outcome in South Australia 1997 Adelaide. Adelaide, Australia: South Australia Health Commission, 1998.
- de Lemos M. Results of the trial testing. The use of the Stanford - Binet Intelligence Scale in Australia. Melbourne, Victoria, Australia: The Australian Council for Education Research Ltd, 1994:4–7.
- Schofield C, Stewart J, Wheeler E. The diets of pregnant and post-pregnant women in different social groups in London and Edinburgh: calcium, iron, retinol, ascorbic acid and folic acid. Br J Nutr 1989;62:363–77.[Medline]
- WHO. Iron deficiency anaemia: assessment, prevention and control—a guide for programme managers. Geneva, Switzerland: WHO, 2001.
- Lozoff B, De Andraca I, Castillo M, Smith JB, Walter T, Pino P. Behavioral and developmental effects of preventing iron-deficiency anemia in healthy full-term infants. Pediatrics 2003;112:846–54.[Abstract/Free Full Text]
- Hemminki E, Merilinen J. Long-term follow-up of mothers and their infants in a randomized trial on iron prophylaxis during pregnancy. Am J Obstet Gynecol 1995;173:205–9.[Medline]
Received for publication October 26, 2005.
Accepted for publication January 17, 2006.
TOP
ABSTRACT
INTRODUCTION
