- Email: [email protected]
Cord serum ferritin concentrations and mental and psychomotor development of children at five years of age
ORIGINAL ARTICLES
Cord serum ferritin concentrations and mental and psychomotor development of children at five years of age Tsunenobu Tamura, MD, Robert L. Goldenberg, MD, Jinrong Hou, MD, Kelley E. Johnston, BS, Suzanne P. Cliver, BA, Sharon L. Ramey, PhD, and Kathleen G. Nelson, MD Objective: Our purpose was to evaluate the association between fetal iron status and mental and psychomotor development at 5 years of age. Study design: We evaluated the association of fetal iron status (umbilical cord serum ferritin concentrations) with test scores of mental and psychomotor development of 278 children. Six tests were given, including full-scale intelligence quotient (FSIQ), language ability, fine- and gross-motor skills, attention, and tractability. Results: Compared with children with cord ferritin in the 2 median quartiles, those in the lowest quartile scored lower on every test and had significantly worse language ability, fine-motor skills, and tractability. They were also 4.8-fold more likely to score poorly in fine-motor skills and 2.7-fold more likely to have poor tractability than children in the median quartiles. FSIQ in the highest quartile was slightly, but not significantly, lower than the median quartiles, but the odds ratio for having a FSIQ score of less than 70 for children in the highest quartile was 3.3 (95% CI 1.2-9.1). Conclusion: Poor iron status (low ferritin) in utero appears to be associated with diminished performance in certain mental and psychomotor tests. The reason for the association between high ferritin concentrations and low FSIQ scores is unknown. (J Pediatr 2002;140:165-70) Iron deficiency is considered to be a serious problem worldwide among infants and women of childbearing age.1,2 Iron is involved in a wide range of biochemical reactions and is essential for normal growth and development of fetuses and infants. It is well-established in animals
that the prenatal exposure to inadequate iron nutritional status results in abnormal brain development.2-6 Many investigators evaluated the association between iron nutritional status and mental and psychomotor development during early infancy and childhood.7-23
From the Departments of Nutrition Sciences, Obstetrics and Gynecology, and Pediatrics, and Civitan International Research Center, University of Alabama at Birmingham.
Supported by a grant and contract from National Institutes of Health (HD32901 and N01-HD4-2811) and by the Agency for Health Care Policy Research Contract (DHHS 282-92-0055). Submitted for publication Oct 9, 2000; revisions received Feb 1, 2001, and July 3, 2001; accepted Sept 27, 2001. Reprint requests: T. Tamura, MD, 218 Webb Bldg, Department of Nutrition Sciences, University of Alabama at Birmingham, Birmingham, AL 35294-3360. Copyright © 2002, Mosby, Inc. All rights reserved. 0022-3476/2002/$35.00 + 0 9/21/120688 doi:10.1067/mpd.2002.120688
The results of several observational trials suggest that iron deficiency early in life results in a long-lasting delay in mental and psychomotor development. However, trials to evaluate the effect of iron supplementation on development during the first 2 years of life have not
See editorial, p 145. provided unanimous results.7-20 In the 14 preventive and treatment intervention trials published to date, only some indicated that iron supplementation has a positive effect either by improving mental and psychomotor development or preventing its delay. In the majority of the studies, the patients were infants (aged <24 months) with iron deficiency, and Bayley Scales of Infant Development were used for the assessment of their development.7-20 A consensus on the relationship between iron nutritional status or iron supplementation and mental and psychomotor development during infancy and childhood has yet to be BMI FMS FSIQ GMS SGA TACL
Body mass index Peabody Fine Motor Scales Full-scale intelligence quotient Peabody Gross Motor Scales Small for gestational age Test for Auditory Comprehension of Language-Revised YAL-A Yale Children’s Inventory for testing attention YAL-T Yale Children’s Inventory for testing tractability
achieved. To our knowledge, there has been no human investigation presented in the literature to evaluate the association between fetal iron nutritional status and mental and psychomotor development of children. We, therefore, under165
TAMURA ET AL
THE JOURNAL OF PEDIATRICS FEBRUARY 2002
took this study to evaluate the iron status of fetuses as assessed by cord serum ferritin concentrations and correlated these values with the development of the same children at 5 years of age.
the test. The study was approved by the Institutional Review Board for Human Use of the University of Alabama at Birmingham, and consent was obtained from each child’s mother.
METHODS
Mental and Psychomotor Tests
Patients The mothers of the children evaluated in this investigation were selected from more than 1500 participants in a study to identify risk factors for pregnancies with small-for-gestational-age (SGA) infants and to evaluate mental and psychomotor development of their children at 5 years of age.24 The women were seeking care through the public health system in the Birmingham, Alabama area. In this population, <5% of the women were breast-feeding their infants and virtually all were eligible for the special supplemental program for Women, Infants, and Children. They had various risk factors for having an SGA infant, including smoking, short stature, low body mass index (BMI), and a history of having an infant with low birth weight. Birth weight was measured within 1 hour of delivery, and the gestational age was based on the information of the last menstrual period and confirmed by ultrasonography. Maternal blood samples were obtained at the mean of 25 and 36 weeks’ gestation, and umbilical cord blood samples were collected at delivery with plastic syringes. Serum was stored at –70°C until it was analyzed. Because of resource constraints, approximately 600 of more than 1500 children received mental and psychomotor tests at 5 years of age. Of these 600 cases, cord blood samples were available from only 278 children (130 females), and they were evaluated in this study. The tests were performed in a single 3-hour session at a mean of 5.5 (± 0.4, SD) years of age. None of these children had major malformations, known chromosomal anomalies, or known congenital viral infection, and all children appeared normal at the time of 166
Mothers took the Home Screening Questionnaire test, a measure of home environment for children, and the Peabody Picture Vocabulary Test-Revised, an evaluation of cognitive function given as a standardized score at the same time. The child’s tests included the Wechsler Intelligence Scale for Children-Revised, which yields a full-scale intelligence quotient (FSIQ), the Test for Auditory Comprehension of Language-1985 0Revised (TACL), the fineand gross-motor scales of the Peabody Developmental Motor Scales (FMS and GMS, respectively), and the Yale Children’s Inventory for attention and tractability (YAL-A and YAL-T, respectively) (ability to obey rules and follow general orders of decorum) of children.25,26 For FSIQ, TACL, GMS, and FMS, the higher the score, the better the performance, whereas for YAL-A and YAL-T, the lower the score, the better the performance. These tests were given by 2 evaluators with extensive training in the testing methodology, who had no knowledge of the ferritin concentrations or birth status of children, including gestational age, birth weight, Apgar score, and any neonatal complications.
Ferritin Assay Cord and maternal serum ferritin concentrations were determined using MAGIC Ferritin [125I] Radioimmunoassay kit (Ciba-Corning Diagnostic, Irvine, Calif), and the radioactivity was measured by a Gamma 400 counter (Beckman, Palo Alto, Calif). The interassay coefficient of variation was approximately 5.8% by repeated analyses of control samples supplied by the manufacturer.27
Statistical Analyses When appropriate, data are presented as mean ± SD. The association
between the test scores of child development and cord serum ferritin concentrations was evaluated by comparing the lowest and highest quartiles independently with the 2 median quartiles of ferritin concentrations. The scores of mental and psychomotor tests were compared after adjusting for the factors that could possibly affect scores, including birth weight, ethnicity, sex, maternal prepregnancy BMI (the body weight in kilograms divided by the square of the height in meters), education, age, smoking status, and scores of the Peabody Picture Vocabulary Test and the Home Screen Questionnaire. SAS statistical software (SAS Institute, Cary, NC) was used for χ2 test, Student t test, Pearson correlation coefficients, and multiple logistic regression. The comparisons of various characteristics between the ferritin quartiles were evaluated with analysis of variance with Tukey procedure as a post hoc test. A P value of <.05 was considered to be statistically significant.
RESULTS The characteristics of the mothers and the children (at birth and 5 years of age) are shown in Table I for groups divided by quartiles of cord ferritin concentrations. The mothers were African American (71.6%) and white (28.4%) and had a mean age of 29.9 (± 4.4) years at the time of mental and psychomotor tests, and mean maternal education was 12.0 (± 1.9) years. The mothers’ mean scores on the Peabody Picture Vocabulary Test and the Home Screening Questionnaire were 79 (± 9.8) and 36.5 (± 6.4), respectively, and both scores are typical of those in inner-city, resource-poor populations. The mothers delivered singleton infants at a mean gestational age of 38.7 (± 2.1) weeks with an average birth weight of 3135 (± 595) g. The prevalence of preterm birth (<37 weeks’ gestation) was 13.3%, and the prevalence of SGA infants was 22% as defined by <15th percentile birth weight for gesta-
TAMURA ET AL
THE JOURNAL OF PEDIATRICS
VOLUME 140, NUMBER 2 Table I. Characteristics of children and their mothers based on cord serum ferritin quartiles
Lowest quartile (<76 µg/L) (n = 66)
Characteristics Mothers Age (y) African American (%) Prepregnancy BMI Education (y) Peabody Picture Vocabulary Test (standardized score) Home screening questionnaire Smokers (%) Alcohol use (%) Illicit drug use (%) Children at birth Males (%) Gestational age (wk) Birth weight (g) Head circumference (cm) Children at age 5 Height (cm) Weight (kg) Head circumference (cm) Low birth weight (<2500 g) infants (%)
2 median quartiles (76–187 µg/L) (n = 142)
Highest quartile (≥188 µg/L) (n = 70)
30.1 ± 4.7 74.2 23.4 ± 5.4 12.2 ± 1.9
30.0 ± 4.3 71.8 24.1 ± 6.5 12.0 ± 1.8
29.5 ± 4.4 68.6 23.3 ± 5.7 12.0 ± 2.1
81 ± 11 35.6 ± 6.6 41.5 35.4 5.8
79 ± 10 36.3 ± 6.3 39.3 31.4 3.3
78 ± 9 37.8 ± 6.5 47.1 32.9 3.4
62.1* 38.2 ± 2.5 2981 ± 720† 33.3 ± 2.2
54.9 38.8 ± 1.9 3221 ± 517 33.7 ± 1.9
41.4* 39.0 ± 2.1 3105 ± 589 33.8 ± 1.8
112 ± 5 19.6 ± 2.8 51.6 ± 1.5 21.2
114 ± 6 20.6 ± 3.5 51.8 ± 1.5 9.2
112 ± 6 20.4 ± 5.0 51.5 ± 1.6 14.3
*Different from the 2 median groups at P < .05. †Different from the other 2 groups at P < .05.
Table II. Mean ± SD test scores in 3 groups on the basis of cord serum ferritin concentrations
Tests FSIQ TACL GMS FMS YAL-A YAL-T
Lowest quartile (<76 µg/L) (n = 66)
2 median quartiles (76–187 µg/L) (n = 142)
Highest quartile (≥188 µg/L) (n = 70)
80 ± 14 74 ± 21* 159 ± 12 76 ± 8* 70 ± 17† 75 ± 21*
84 ± 12 82 ± 17 162 ± 11 79 ± 6 65 ± 16 66 ± 20
83 ± 12 79 ± 21 160 ± 13 79 ± 7 68 ± 17 69 ± 20
All patients combined (n = 278) 83 ± 12 80 ± 19 161 ± 12 78 ± 7 67 ± 17 69 ± 20
*P < .01 for comparison with the 2 median quartiles. †P = .05 for comparison with the 2 median quartiles.
tional age on the basis of Alabama standards, which corresponds roughly to the 10th percentile established by Williams et al.24,28 There were no significant differences in the characteristics of patients in various quartiles, with the exception that the mean birth weight in the lowest ferritin quartile was significantly lower than the birth weights of the 2 median groups (P < .05). The percentage of male infants was highest in the lowest ferritin quar-
tile and lowest in the highest ferritin quartile (P < .05), and this finding most likely reflects the significantly higher cord serum ferritin concentrations in female infants than in male infants as previously reported.27 The percentage of smokers was slightly higher in the mothers with children in the highest ferritin quartile. Because of the differences (even if not significant) in birth weight, child sex, and smoking among results in the ferritin quartiles, we ad-
justed for these and other factors in subsequent analyses. The mean cord ferritin concentration was 142.8 (± 98.4) µg/L, which was 7to 10-fold higher than the mean maternal serum ferritin concentrations (22.3 [± 26.7] and 15.0 [± 18.9] µg/L at 25 and 36 weeks’ gestation, respectively). These values are comparable to values reported by other investigators.27,29 Cord ferritin concentrations correlated significantly but weakly with maternal 167
TAMURA ET AL
THE JOURNAL OF PEDIATRICS FEBRUARY 2002
Fig 1. Comparison of the percentage of children who had lower than the 15th percentile of mental and psychomotor test scores and 3 groups based on quartiles of cord serum ferritin concentrations.The star mark indicates the value is significantly different from that in the II + III group. I, lowest quartile; II and III, 2 median quartiles; IV, highest quartile. Table III. Adjusted odds ratios (95% CI) for test scores <15th percentile in the lowest and highest quartiles compared with the 2 median quartiles of cord serum ferritin
Tests (cutoff score for the 15th percentile) FSIQ (≤70) TACL (≤58) GMS (≤148) FMS (≤70) YAL-A (≥90) YAL-T (≥98)
Lowest quartile vs 2 median quartiles
Highest quartile vs 2 median quartiles
2.1 (0.8–5.3) 1.9 (0.7–4.8) 1.6 (0.6–3.9) 4.8 (1.9–12.0) 2.1 (1.0–4.7) 2.7 (1.2–5.9)
3.3 (1.2–9.1) 2.1 (0.8–5.4) 1.8 (0.7-4.4) 1.3 (0.4–3.9) 1.4 (0.6–3.3) 1.5 (0.6–3.4)
The odd ratios for poor scores in the lowest and highest groups were calculated independently by comparing with the 2 median quartiles. These were adjusted for maternal age, education, prepregnancy BMI, Peabody score, smoking, and race, as well as Home Screen Score, birth weight, and gender of children.
ferritin concentrations at 36 weeks’ gestation (r = 0.17, P < .02), whereas there was no correlation between cord and maternal serum ferritin concentrations at 25 weeks’ gestation. The children in the lowest cord ferritin quartile scored the worst on every test, had significantly lower scores in 168
TACL and FMS, and worse scores on YAL-A and YAL-T than children in the 2 median quartiles (Table II). Although lower, the scores for FSIQ and grossmotor function were not significantly different among these groups. There were also no significant differences in scores on any test between the children
in the highest ferritin quartile and children in the 2 median quartiles. To further evaluate the degree of association between low cord ferritin concentrations and low scores on these tests, we evaluated the relationship between cord ferritin concentrations and scores below the 15th percentile for each test. This percentile value was arbitrarily selected because no established cutoffs have been available to define abnormal test scores. The use of this value for the FSIQ test led to a cutoff of 70, which is commonly used to define mental retardation. The Figure shows percentages of children with test scores lower than the 15th percentile in the lowest, highest, and 2 median cord ferritin quartiles. For every test, the possibility of a U-shaped relationship with cord ferritin concentrations was observed, indicating that children in the 2 median quartiles uniformly performed better than children in either the lowest or highest quartile. For the FMS, YALA, and YAL-T, the most pronounced difference is between the low ferritin group and the group in the middle quartiles of ferritin. Adjusted odds ratios for scores below the 15th percentile on each test were calculated by comparing the lowest or highest quartile with the 2 median quartiles (Table III). For children in the lowest ferritin quartile for every test, the likelihood of having a low score was increased by at least 60%, and it was increased 2 to 5 times for FMS, YAL-A, and YAL-T. A 3.3-fold increase in the odds for having an FSIQ score lower than 70 (15th percentile cutoff score) was found in the highest ferritin quartile compared with the 2 median quartiles of cord serum ferritin concentrations.
DISCUSSION The growth and development of the human brain are extremely rapid during the last few months of pregnancy and the first several months of neonatal life.30 Iron deficiency in either fetal or
TAMURA ET AL
THE JOURNAL OF PEDIATRICS
VOLUME 140, NUMBER 2 neonatal life is believed to have a significant effect on mental development, including cognitive and behavioral performance, which is associated with long-lasting developmental differences in later life.2-4 To our knowledge, however, the investigation to evaluate direct associations between iron nutritional status in utero and multiple indicators of development in later life has not been conducted in humans. We found that the scores of various developmental tests were significantly compromised in children in the lowest cord ferritin quartile than children in the 2 median quartiles. Our observation of poorer performance in children in the lowest cord ferritin quartile supports the findings by many researchers that adequate iron nutritional status is important for normal brain development.2-4 Furthermore, we found that high ferritin, reflecting either excessive iron or an acute-phase reaction in intrauterine life, may be associated with poor FSIQ. Although the exact mechanistic effect of iron deficiency on brain development remains unclear, investigators have demonstrated that prenatal iron deficiency leads to alterations in neurotransmitter or neuronal metabolism and myelin formation in the brain of experimental animals.3,5,6 Our data indicate that poor prenatal iron status may be detrimental to language ability, motor functions, attention, and tractability. However, it is unknown whether iron supplementation early in neonatal life in our patients could have reversed poor development, and only early iron supplementation intervention trials can answer this question. If the results of such a trial are positive, the prevention of poor development by iron supplementation would then be a high priority in terms of public health policy. No report in the literature explains our observation of the association between the high risk for lower FSIQ among children in the highest cord ferritin quartile compared with children in the 2 median quartiles. The following 2 hypotheses may be offered. First, fer-
ritin is not only one of the best indicators of body iron stores,31 but also acts as an acute-phase reactant, such as in the case of infection or trauma.32 If elevated cord ferritin concentrations are indicative of an acute-phase reaction in fetal life, it is possible that an infection, perhaps of the maternal upper-genital tract or of the fetus itself, was the cause of low FSIQ. Maternal infection at delivery has been associated with cerebral palsy and mental retardation.33 Furthermore, elevated maternal ferritin concentrations have been associated with preterm delivery in a similar socioeconomic population.34,35 Second, our data may indicate that the increased cord ferritin concentrations may reflect excessive fetal iron stores that may be detrimental to the developing brain; it is known that iron-catalyzed oxidative damage of lipids occurs.36 This occurs in cultured human brain cells and in brain cells of experimental animals.37,38 Furthermore, neurobehavioral dysfunction, maze learning, and motor activity deficits induced by oral iron overload in early life were recently reported in laboratory rodents.39,40 Similar damage may occur during the late intrauterine and early neonatal life, which could lead to poor intellectual performance in humans. The possible mechanisms by which excessive iron stores reach the fetus may include (1) an excessive maternal iron intake during pregnancy, (2) a genetic disposition (eg, hemochromatosis),41 and (3) excess placental transfer of iron by an unidentified mechanism. Further studies are warranted to evaluate the nature of the relationship between high ferritin concentrations and low FSIQ scores. The U-shaped relationship between cord ferritin and mental and psychomotor developmental status (Figure) is similar to the association between maternal hemoglobin or hematocrit values and pregnancy outcome, including preterm delivery and fetal growth restriction.42,43 In each study, investigators demonstrated an increased incidence of unfavorable pregnancy outcome at both
ends of the hemoglobin or hematocrit distribution. This finding suggests that both high and low hemoglobin or hematocrit values may be associated with fetal and maternal morbidity and mortality. It is unknown whether the mechanisms in effect are similar, because a high hemoglobin or hematocrit value may reflect vasoconstriction, whereas ferritin concentration may be elevated for other reasons. Low socioeconomic status is known to affect cognitive and language development of children. However, the differences in developmental status in relation to cord ferritin concentrations observed in this study are independent from bias regarding socioeconomic status because they generally appeared to be uniform among our patients (Table I). Also, regression analyses adjusting for some of the most important maternal socioeconomic characteristics did not change any of the outcomes.
REFERENCES 1. Looker AC, Dallman PR, Carroll MD, Gunter EW, Johnson CL. Prevalence of iron deficiency in the United States. JAMA 1997;277:973-6. 2. Walker ARP. The remedying of iron deficiency: what priority should it have? Br J Nutr 1998;79:227-35. 3. Beard JL, Connor JR, Jones BC. Iron in the brain. Nutr Rev 1993;51:157-70. 4. Pollitt E. Iron deficiency and cognitive function. Annu Rev Nutr 1993;13:52137. 5. 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. 6. Kwik-Uribe C, Gietzen D, German JB, Golub MS, Keen CL. Chronic marginal iron intakes during early development in mice result in persistent changes in dopamine metabolism and myelin composition. J Nutr 2000;130:2821-30. 7. Oski FA, Honig AS. The effects of therapy on the developmental scores of iron-deficiency infants. J Pediatr 1978; 92:21-5. 8. Lozoff B, Brittenham GM, Viteri FE, Wolf AW, Urrutia JJ. The effects of 169
TAMURA ET AL
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
170
short-term oral iron therapy on developmental deficits in iron-deficient anemic infants. J Pediatr 1982;100:351-7. Oski FA, Honig AS, Helu B, Howanitz P. Effect of iron therapy on behavior performance in nonanemic, iron-deficient infants. Pediatrics 1983;71:877-80. Walter T, Kovalskys J, Stekel A. Effect of mild iron deficiency on infant mental development scores. J Pediatr 1983; 102:519-22. Aukett MA, Parks YA, Scott PH, Wharton BA. Treatment with iron increases weight gain and psychomotor development. Arch Dis Child 1986; 61:849-57. Deinard AS, List A, Lindgren B, Hunt JV, Chang P-N. Cognitive deficit in iron-deficient and iron-deficient anemic children. J Pediatr 1986;108:681-9. Lozoff B, Brittenham GM, Wolf AW, McClish DK, Kuhnert PM, Jimenez E, et al. Iron deficiency anemia and iron therapy effects on infant developmental test performance. Pediatrics 1987; 79:981-95. Walter T, De Andraca I, Chadud P, Perales CG. Iron deficiency anemia: adverse effects on infant psychomotor development. Pediatrics 1989;84:7-17. Idjradinata P, Pollitt E. Reversal of developmental delays in iron-deficient anaemic infants treated with iron. Lancet 1993;341:1-4. Moffatt MEK, Longstaffe S, Besant J, Dureski C. Prevention of iron deficiency and psychomotor decline in highrisk infants through use of iron-fortified infant formula: a randomized clinical trial. J Pediatr 1994;125:527-34. Lozoff B, Wolf AW, Jimenez E. Irondeficiency anemia and infant development: effects of extended oral iron therapy. J Pediatr 1996;129:382-9. Roncagliolo M, Garrido M, Walter T, Peirano P, Lozoff B. Evidence of altered central nervous system development in infants with iron deficiency anemia at 6 mo: delayed maturation of auditory brainstem responses. Am J Clin Nutr 1998;68:683-90. Williams J, Wolff A, Daly A, MacDonald A, Aukett A, Booth IW. Iron supplemented formula milk related to reduction in psychomotor decline in infants from inner city
THE JOURNAL OF PEDIATRICS FEBRUARY 2002
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
areas: randomized study. BMJ 1999;318:693-7. Morley R, Abbott R, Fairweather-Tait S, MacFadyen U, Stephenson T, Lucas A. Iron fortification follow on formula from 9 to 18 months improves iron status but not development or growth: a randomized trial. Arch Dis Child 1999; 81:247-52. Deinard A, Gilbert A, Dodds M, Egeland B. Iron deficiency and behavioral deficits. Pediatrics 1981;68:828-33. Lozoff B, Jemenez E, Wolf AW. Longterm developmental outcome of infants with iron deficiency. N Engl J Med 1991;325:687-94. Lozoff B, Jimenez E, Hagen J, Mollen E, Wolf AW. Poorer behavioral and developmental outcome more than 10 years after treatment for iron deficiency in infancy. Pediatrics 2000;105:e51. Goldenberg RL, DuBard MB, Cliver SP, Nelson KG, Blankson K, Ramey SL, et al. Pregnancy outcome and intelligence at age five years. Am J Obstet Gynecol 1996;175:1511-5. Test critiques. In: Keyser DJ, Sweetland RC, editors. Vol. I, V, and VI. Kansas City: Test Corporation of America; 1984, 1986, and 1987. Shaywitz SE, Schnell C, Shaywitz BA, Towle VR. Yale Children’s Inventory (YCI): an instrument to assess children with attention deficit and learning disabilities. I. Scale development and psychometric properties. J Abnorm Child Psychol 1986;14:347-64. Tamura T, Hou J, Goldenberg RL, Johnston KE, Cliver SP. Gender difference in cord serum ferritin concentrations. Biol Neonate 1999;75:343-9. Williams RL, Creasy RK, Cunningham GC, Hawes WE, Norris FD, Tashiro M. Fetal growth and perinatal viability in California. Obstet Gynecol 1982; 59:624-32. Rois E, Lipschitz DA, Cook JD, Smith NJ. Relationship of maternal and infant iron stores as assessed by determination of plasma ferritin. Pediatrics 1975;55: 694-9. Dobbing J, Sands J. Quantitative growth and development of human brain. Arch Dis Child 1973; 48:757-67. Cook JD, Finch CA, Smith NJ. Evaluation of iron status of a population. Blood 1976;48:449-55. Gabay C, Kushner I. Acute-phase proteins and other systemic responses to
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
inflammation. N Engl J Med 1999; 340:448-54. Leviton A, Paneth N, Reuss ML, Susser M, Allred EN, Dammann O, et al. Maternal infection, fetal inflammatory response, and brain damage in very low birth weight infants. Pediatr Res 1999;46:566-75. Tamura T, Goldenberg RL, Johnston KE, Cliver SP, Hickey CA. Serum ferritin: a predictor of early spontaneous preterm birth. Obstet Gynecol 1996; 87:360-5. Goldenberg RL, Tamura T, DuBard M, Johnston KE, Copper RL, Neggers Y. Plasma ferritin and pregnancy outcome. Am J Obstet Gynecol 1996;175:1356-9. Aisen P, Cohen G, Kang JO. Iron toxicosis. Int Rev Exp Pathol 1990;31:1-46. Andorn AC, Britton RS, Bacon BR. Ascorbate-stimulated lipid peroxidation in human brain is dependent on iron but not on hydroxyl radicals. J Neurochem 1996;67:717-22. Sobotka TJ, Whittaker P, Sobotka JM, Brodie RE, Quander DY, Robl M, et al. Nurobehavioral dysfunctions associated with dietary iron overload. Physiol Behav 1996;59:213-9. Friedriksson A, Schröder N, Eriksson P, Izquierdo I, Archer T. Maze learning and motor activity deficits in adult mice induced by iron exposure during a critical postnatal period. Develop Brain Res 2000;119:65-74. Castelnau PA, Garrett RS, Palinski W, Witztum JL, Campbell IL, Powell HC. Abnormal iron deposition associated with lipid peroxidation in transgenic mice expressing interleukin-6 in the brain. J Neuropathol Exp Neurol 1998;572:268-82. Bothwell TH, Charlton RW, Motulsky AG. Hemochromatosis. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic and molecular bases of inherited disease. 7th ed. New York: McGraw-Hill; 1995. p. 2237-69. Garn SM, Keating MT, Falkner F. Hematological status and pregnancy outcomes. Am J Clin Nutr 1981; 34:115-7. Lu ZM, Goldenberg RL, Cliver SP, Cutter G, Blankson M. The relationship between maternal hematocrit and pregnancy outcome. Obstet Gynecol 1991;77:190-4.
Cord serum ferritin concentrations and mental and psychomotor development of children at five years of age Tsunenobu Tamura, MD, Robert L. Goldenberg, MD, Jinrong Hou, MD, Kelley E. Johnston, BS, Suzanne P. Cliver, BA, Sharon L. Ramey, PhD, and Kathleen G. Nelson, MD Objective: Our purpose was to evaluate the association between fetal iron status and mental and psychomotor development at 5 years of age. Study design: We evaluated the association of fetal iron status (umbilical cord serum ferritin concentrations) with test scores of mental and psychomotor development of 278 children. Six tests were given, including full-scale intelligence quotient (FSIQ), language ability, fine- and gross-motor skills, attention, and tractability. Results: Compared with children with cord ferritin in the 2 median quartiles, those in the lowest quartile scored lower on every test and had significantly worse language ability, fine-motor skills, and tractability. They were also 4.8-fold more likely to score poorly in fine-motor skills and 2.7-fold more likely to have poor tractability than children in the median quartiles. FSIQ in the highest quartile was slightly, but not significantly, lower than the median quartiles, but the odds ratio for having a FSIQ score of less than 70 for children in the highest quartile was 3.3 (95% CI 1.2-9.1). Conclusion: Poor iron status (low ferritin) in utero appears to be associated with diminished performance in certain mental and psychomotor tests. The reason for the association between high ferritin concentrations and low FSIQ scores is unknown. (J Pediatr 2002;140:165-70) Iron deficiency is considered to be a serious problem worldwide among infants and women of childbearing age.1,2 Iron is involved in a wide range of biochemical reactions and is essential for normal growth and development of fetuses and infants. It is well-established in animals
that the prenatal exposure to inadequate iron nutritional status results in abnormal brain development.2-6 Many investigators evaluated the association between iron nutritional status and mental and psychomotor development during early infancy and childhood.7-23
From the Departments of Nutrition Sciences, Obstetrics and Gynecology, and Pediatrics, and Civitan International Research Center, University of Alabama at Birmingham.
Supported by a grant and contract from National Institutes of Health (HD32901 and N01-HD4-2811) and by the Agency for Health Care Policy Research Contract (DHHS 282-92-0055). Submitted for publication Oct 9, 2000; revisions received Feb 1, 2001, and July 3, 2001; accepted Sept 27, 2001. Reprint requests: T. Tamura, MD, 218 Webb Bldg, Department of Nutrition Sciences, University of Alabama at Birmingham, Birmingham, AL 35294-3360. Copyright © 2002, Mosby, Inc. All rights reserved. 0022-3476/2002/$35.00 + 0 9/21/120688 doi:10.1067/mpd.2002.120688
The results of several observational trials suggest that iron deficiency early in life results in a long-lasting delay in mental and psychomotor development. However, trials to evaluate the effect of iron supplementation on development during the first 2 years of life have not
See editorial, p 145. provided unanimous results.7-20 In the 14 preventive and treatment intervention trials published to date, only some indicated that iron supplementation has a positive effect either by improving mental and psychomotor development or preventing its delay. In the majority of the studies, the patients were infants (aged <24 months) with iron deficiency, and Bayley Scales of Infant Development were used for the assessment of their development.7-20 A consensus on the relationship between iron nutritional status or iron supplementation and mental and psychomotor development during infancy and childhood has yet to be BMI FMS FSIQ GMS SGA TACL
Body mass index Peabody Fine Motor Scales Full-scale intelligence quotient Peabody Gross Motor Scales Small for gestational age Test for Auditory Comprehension of Language-Revised YAL-A Yale Children’s Inventory for testing attention YAL-T Yale Children’s Inventory for testing tractability
achieved. To our knowledge, there has been no human investigation presented in the literature to evaluate the association between fetal iron nutritional status and mental and psychomotor development of children. We, therefore, under165
TAMURA ET AL
THE JOURNAL OF PEDIATRICS FEBRUARY 2002
took this study to evaluate the iron status of fetuses as assessed by cord serum ferritin concentrations and correlated these values with the development of the same children at 5 years of age.
the test. The study was approved by the Institutional Review Board for Human Use of the University of Alabama at Birmingham, and consent was obtained from each child’s mother.
METHODS
Mental and Psychomotor Tests
Patients The mothers of the children evaluated in this investigation were selected from more than 1500 participants in a study to identify risk factors for pregnancies with small-for-gestational-age (SGA) infants and to evaluate mental and psychomotor development of their children at 5 years of age.24 The women were seeking care through the public health system in the Birmingham, Alabama area. In this population, <5% of the women were breast-feeding their infants and virtually all were eligible for the special supplemental program for Women, Infants, and Children. They had various risk factors for having an SGA infant, including smoking, short stature, low body mass index (BMI), and a history of having an infant with low birth weight. Birth weight was measured within 1 hour of delivery, and the gestational age was based on the information of the last menstrual period and confirmed by ultrasonography. Maternal blood samples were obtained at the mean of 25 and 36 weeks’ gestation, and umbilical cord blood samples were collected at delivery with plastic syringes. Serum was stored at –70°C until it was analyzed. Because of resource constraints, approximately 600 of more than 1500 children received mental and psychomotor tests at 5 years of age. Of these 600 cases, cord blood samples were available from only 278 children (130 females), and they were evaluated in this study. The tests were performed in a single 3-hour session at a mean of 5.5 (± 0.4, SD) years of age. None of these children had major malformations, known chromosomal anomalies, or known congenital viral infection, and all children appeared normal at the time of 166
Mothers took the Home Screening Questionnaire test, a measure of home environment for children, and the Peabody Picture Vocabulary Test-Revised, an evaluation of cognitive function given as a standardized score at the same time. The child’s tests included the Wechsler Intelligence Scale for Children-Revised, which yields a full-scale intelligence quotient (FSIQ), the Test for Auditory Comprehension of Language-1985 0Revised (TACL), the fineand gross-motor scales of the Peabody Developmental Motor Scales (FMS and GMS, respectively), and the Yale Children’s Inventory for attention and tractability (YAL-A and YAL-T, respectively) (ability to obey rules and follow general orders of decorum) of children.25,26 For FSIQ, TACL, GMS, and FMS, the higher the score, the better the performance, whereas for YAL-A and YAL-T, the lower the score, the better the performance. These tests were given by 2 evaluators with extensive training in the testing methodology, who had no knowledge of the ferritin concentrations or birth status of children, including gestational age, birth weight, Apgar score, and any neonatal complications.
Ferritin Assay Cord and maternal serum ferritin concentrations were determined using MAGIC Ferritin [125I] Radioimmunoassay kit (Ciba-Corning Diagnostic, Irvine, Calif), and the radioactivity was measured by a Gamma 400 counter (Beckman, Palo Alto, Calif). The interassay coefficient of variation was approximately 5.8% by repeated analyses of control samples supplied by the manufacturer.27
Statistical Analyses When appropriate, data are presented as mean ± SD. The association
between the test scores of child development and cord serum ferritin concentrations was evaluated by comparing the lowest and highest quartiles independently with the 2 median quartiles of ferritin concentrations. The scores of mental and psychomotor tests were compared after adjusting for the factors that could possibly affect scores, including birth weight, ethnicity, sex, maternal prepregnancy BMI (the body weight in kilograms divided by the square of the height in meters), education, age, smoking status, and scores of the Peabody Picture Vocabulary Test and the Home Screen Questionnaire. SAS statistical software (SAS Institute, Cary, NC) was used for χ2 test, Student t test, Pearson correlation coefficients, and multiple logistic regression. The comparisons of various characteristics between the ferritin quartiles were evaluated with analysis of variance with Tukey procedure as a post hoc test. A P value of <.05 was considered to be statistically significant.
RESULTS The characteristics of the mothers and the children (at birth and 5 years of age) are shown in Table I for groups divided by quartiles of cord ferritin concentrations. The mothers were African American (71.6%) and white (28.4%) and had a mean age of 29.9 (± 4.4) years at the time of mental and psychomotor tests, and mean maternal education was 12.0 (± 1.9) years. The mothers’ mean scores on the Peabody Picture Vocabulary Test and the Home Screening Questionnaire were 79 (± 9.8) and 36.5 (± 6.4), respectively, and both scores are typical of those in inner-city, resource-poor populations. The mothers delivered singleton infants at a mean gestational age of 38.7 (± 2.1) weeks with an average birth weight of 3135 (± 595) g. The prevalence of preterm birth (<37 weeks’ gestation) was 13.3%, and the prevalence of SGA infants was 22% as defined by <15th percentile birth weight for gesta-
TAMURA ET AL
THE JOURNAL OF PEDIATRICS
VOLUME 140, NUMBER 2 Table I. Characteristics of children and their mothers based on cord serum ferritin quartiles
Lowest quartile (<76 µg/L) (n = 66)
Characteristics Mothers Age (y) African American (%) Prepregnancy BMI Education (y) Peabody Picture Vocabulary Test (standardized score) Home screening questionnaire Smokers (%) Alcohol use (%) Illicit drug use (%) Children at birth Males (%) Gestational age (wk) Birth weight (g) Head circumference (cm) Children at age 5 Height (cm) Weight (kg) Head circumference (cm) Low birth weight (<2500 g) infants (%)
2 median quartiles (76–187 µg/L) (n = 142)
Highest quartile (≥188 µg/L) (n = 70)
30.1 ± 4.7 74.2 23.4 ± 5.4 12.2 ± 1.9
30.0 ± 4.3 71.8 24.1 ± 6.5 12.0 ± 1.8
29.5 ± 4.4 68.6 23.3 ± 5.7 12.0 ± 2.1
81 ± 11 35.6 ± 6.6 41.5 35.4 5.8
79 ± 10 36.3 ± 6.3 39.3 31.4 3.3
78 ± 9 37.8 ± 6.5 47.1 32.9 3.4
62.1* 38.2 ± 2.5 2981 ± 720† 33.3 ± 2.2
54.9 38.8 ± 1.9 3221 ± 517 33.7 ± 1.9
41.4* 39.0 ± 2.1 3105 ± 589 33.8 ± 1.8
112 ± 5 19.6 ± 2.8 51.6 ± 1.5 21.2
114 ± 6 20.6 ± 3.5 51.8 ± 1.5 9.2
112 ± 6 20.4 ± 5.0 51.5 ± 1.6 14.3
*Different from the 2 median groups at P < .05. †Different from the other 2 groups at P < .05.
Table II. Mean ± SD test scores in 3 groups on the basis of cord serum ferritin concentrations
Tests FSIQ TACL GMS FMS YAL-A YAL-T
Lowest quartile (<76 µg/L) (n = 66)
2 median quartiles (76–187 µg/L) (n = 142)
Highest quartile (≥188 µg/L) (n = 70)
80 ± 14 74 ± 21* 159 ± 12 76 ± 8* 70 ± 17† 75 ± 21*
84 ± 12 82 ± 17 162 ± 11 79 ± 6 65 ± 16 66 ± 20
83 ± 12 79 ± 21 160 ± 13 79 ± 7 68 ± 17 69 ± 20
All patients combined (n = 278) 83 ± 12 80 ± 19 161 ± 12 78 ± 7 67 ± 17 69 ± 20
*P < .01 for comparison with the 2 median quartiles. †P = .05 for comparison with the 2 median quartiles.
tional age on the basis of Alabama standards, which corresponds roughly to the 10th percentile established by Williams et al.24,28 There were no significant differences in the characteristics of patients in various quartiles, with the exception that the mean birth weight in the lowest ferritin quartile was significantly lower than the birth weights of the 2 median groups (P < .05). The percentage of male infants was highest in the lowest ferritin quar-
tile and lowest in the highest ferritin quartile (P < .05), and this finding most likely reflects the significantly higher cord serum ferritin concentrations in female infants than in male infants as previously reported.27 The percentage of smokers was slightly higher in the mothers with children in the highest ferritin quartile. Because of the differences (even if not significant) in birth weight, child sex, and smoking among results in the ferritin quartiles, we ad-
justed for these and other factors in subsequent analyses. The mean cord ferritin concentration was 142.8 (± 98.4) µg/L, which was 7to 10-fold higher than the mean maternal serum ferritin concentrations (22.3 [± 26.7] and 15.0 [± 18.9] µg/L at 25 and 36 weeks’ gestation, respectively). These values are comparable to values reported by other investigators.27,29 Cord ferritin concentrations correlated significantly but weakly with maternal 167
TAMURA ET AL
THE JOURNAL OF PEDIATRICS FEBRUARY 2002
Fig 1. Comparison of the percentage of children who had lower than the 15th percentile of mental and psychomotor test scores and 3 groups based on quartiles of cord serum ferritin concentrations.The star mark indicates the value is significantly different from that in the II + III group. I, lowest quartile; II and III, 2 median quartiles; IV, highest quartile. Table III. Adjusted odds ratios (95% CI) for test scores <15th percentile in the lowest and highest quartiles compared with the 2 median quartiles of cord serum ferritin
Tests (cutoff score for the 15th percentile) FSIQ (≤70) TACL (≤58) GMS (≤148) FMS (≤70) YAL-A (≥90) YAL-T (≥98)
Lowest quartile vs 2 median quartiles
Highest quartile vs 2 median quartiles
2.1 (0.8–5.3) 1.9 (0.7–4.8) 1.6 (0.6–3.9) 4.8 (1.9–12.0) 2.1 (1.0–4.7) 2.7 (1.2–5.9)
3.3 (1.2–9.1) 2.1 (0.8–5.4) 1.8 (0.7-4.4) 1.3 (0.4–3.9) 1.4 (0.6–3.3) 1.5 (0.6–3.4)
The odd ratios for poor scores in the lowest and highest groups were calculated independently by comparing with the 2 median quartiles. These were adjusted for maternal age, education, prepregnancy BMI, Peabody score, smoking, and race, as well as Home Screen Score, birth weight, and gender of children.
ferritin concentrations at 36 weeks’ gestation (r = 0.17, P < .02), whereas there was no correlation between cord and maternal serum ferritin concentrations at 25 weeks’ gestation. The children in the lowest cord ferritin quartile scored the worst on every test, had significantly lower scores in 168
TACL and FMS, and worse scores on YAL-A and YAL-T than children in the 2 median quartiles (Table II). Although lower, the scores for FSIQ and grossmotor function were not significantly different among these groups. There were also no significant differences in scores on any test between the children
in the highest ferritin quartile and children in the 2 median quartiles. To further evaluate the degree of association between low cord ferritin concentrations and low scores on these tests, we evaluated the relationship between cord ferritin concentrations and scores below the 15th percentile for each test. This percentile value was arbitrarily selected because no established cutoffs have been available to define abnormal test scores. The use of this value for the FSIQ test led to a cutoff of 70, which is commonly used to define mental retardation. The Figure shows percentages of children with test scores lower than the 15th percentile in the lowest, highest, and 2 median cord ferritin quartiles. For every test, the possibility of a U-shaped relationship with cord ferritin concentrations was observed, indicating that children in the 2 median quartiles uniformly performed better than children in either the lowest or highest quartile. For the FMS, YALA, and YAL-T, the most pronounced difference is between the low ferritin group and the group in the middle quartiles of ferritin. Adjusted odds ratios for scores below the 15th percentile on each test were calculated by comparing the lowest or highest quartile with the 2 median quartiles (Table III). For children in the lowest ferritin quartile for every test, the likelihood of having a low score was increased by at least 60%, and it was increased 2 to 5 times for FMS, YAL-A, and YAL-T. A 3.3-fold increase in the odds for having an FSIQ score lower than 70 (15th percentile cutoff score) was found in the highest ferritin quartile compared with the 2 median quartiles of cord serum ferritin concentrations.
DISCUSSION The growth and development of the human brain are extremely rapid during the last few months of pregnancy and the first several months of neonatal life.30 Iron deficiency in either fetal or
TAMURA ET AL
THE JOURNAL OF PEDIATRICS
VOLUME 140, NUMBER 2 neonatal life is believed to have a significant effect on mental development, including cognitive and behavioral performance, which is associated with long-lasting developmental differences in later life.2-4 To our knowledge, however, the investigation to evaluate direct associations between iron nutritional status in utero and multiple indicators of development in later life has not been conducted in humans. We found that the scores of various developmental tests were significantly compromised in children in the lowest cord ferritin quartile than children in the 2 median quartiles. Our observation of poorer performance in children in the lowest cord ferritin quartile supports the findings by many researchers that adequate iron nutritional status is important for normal brain development.2-4 Furthermore, we found that high ferritin, reflecting either excessive iron or an acute-phase reaction in intrauterine life, may be associated with poor FSIQ. Although the exact mechanistic effect of iron deficiency on brain development remains unclear, investigators have demonstrated that prenatal iron deficiency leads to alterations in neurotransmitter or neuronal metabolism and myelin formation in the brain of experimental animals.3,5,6 Our data indicate that poor prenatal iron status may be detrimental to language ability, motor functions, attention, and tractability. However, it is unknown whether iron supplementation early in neonatal life in our patients could have reversed poor development, and only early iron supplementation intervention trials can answer this question. If the results of such a trial are positive, the prevention of poor development by iron supplementation would then be a high priority in terms of public health policy. No report in the literature explains our observation of the association between the high risk for lower FSIQ among children in the highest cord ferritin quartile compared with children in the 2 median quartiles. The following 2 hypotheses may be offered. First, fer-
ritin is not only one of the best indicators of body iron stores,31 but also acts as an acute-phase reactant, such as in the case of infection or trauma.32 If elevated cord ferritin concentrations are indicative of an acute-phase reaction in fetal life, it is possible that an infection, perhaps of the maternal upper-genital tract or of the fetus itself, was the cause of low FSIQ. Maternal infection at delivery has been associated with cerebral palsy and mental retardation.33 Furthermore, elevated maternal ferritin concentrations have been associated with preterm delivery in a similar socioeconomic population.34,35 Second, our data may indicate that the increased cord ferritin concentrations may reflect excessive fetal iron stores that may be detrimental to the developing brain; it is known that iron-catalyzed oxidative damage of lipids occurs.36 This occurs in cultured human brain cells and in brain cells of experimental animals.37,38 Furthermore, neurobehavioral dysfunction, maze learning, and motor activity deficits induced by oral iron overload in early life were recently reported in laboratory rodents.39,40 Similar damage may occur during the late intrauterine and early neonatal life, which could lead to poor intellectual performance in humans. The possible mechanisms by which excessive iron stores reach the fetus may include (1) an excessive maternal iron intake during pregnancy, (2) a genetic disposition (eg, hemochromatosis),41 and (3) excess placental transfer of iron by an unidentified mechanism. Further studies are warranted to evaluate the nature of the relationship between high ferritin concentrations and low FSIQ scores. The U-shaped relationship between cord ferritin and mental and psychomotor developmental status (Figure) is similar to the association between maternal hemoglobin or hematocrit values and pregnancy outcome, including preterm delivery and fetal growth restriction.42,43 In each study, investigators demonstrated an increased incidence of unfavorable pregnancy outcome at both
ends of the hemoglobin or hematocrit distribution. This finding suggests that both high and low hemoglobin or hematocrit values may be associated with fetal and maternal morbidity and mortality. It is unknown whether the mechanisms in effect are similar, because a high hemoglobin or hematocrit value may reflect vasoconstriction, whereas ferritin concentration may be elevated for other reasons. Low socioeconomic status is known to affect cognitive and language development of children. However, the differences in developmental status in relation to cord ferritin concentrations observed in this study are independent from bias regarding socioeconomic status because they generally appeared to be uniform among our patients (Table I). Also, regression analyses adjusting for some of the most important maternal socioeconomic characteristics did not change any of the outcomes.
REFERENCES 1. Looker AC, Dallman PR, Carroll MD, Gunter EW, Johnson CL. Prevalence of iron deficiency in the United States. JAMA 1997;277:973-6. 2. Walker ARP. The remedying of iron deficiency: what priority should it have? Br J Nutr 1998;79:227-35. 3. Beard JL, Connor JR, Jones BC. Iron in the brain. Nutr Rev 1993;51:157-70. 4. Pollitt E. Iron deficiency and cognitive function. Annu Rev Nutr 1993;13:52137. 5. 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. 6. Kwik-Uribe C, Gietzen D, German JB, Golub MS, Keen CL. Chronic marginal iron intakes during early development in mice result in persistent changes in dopamine metabolism and myelin composition. J Nutr 2000;130:2821-30. 7. Oski FA, Honig AS. The effects of therapy on the developmental scores of iron-deficiency infants. J Pediatr 1978; 92:21-5. 8. Lozoff B, Brittenham GM, Viteri FE, Wolf AW, Urrutia JJ. The effects of 169
TAMURA ET AL
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
170
short-term oral iron therapy on developmental deficits in iron-deficient anemic infants. J Pediatr 1982;100:351-7. Oski FA, Honig AS, Helu B, Howanitz P. Effect of iron therapy on behavior performance in nonanemic, iron-deficient infants. Pediatrics 1983;71:877-80. Walter T, Kovalskys J, Stekel A. Effect of mild iron deficiency on infant mental development scores. J Pediatr 1983; 102:519-22. Aukett MA, Parks YA, Scott PH, Wharton BA. Treatment with iron increases weight gain and psychomotor development. Arch Dis Child 1986; 61:849-57. Deinard AS, List A, Lindgren B, Hunt JV, Chang P-N. Cognitive deficit in iron-deficient and iron-deficient anemic children. J Pediatr 1986;108:681-9. Lozoff B, Brittenham GM, Wolf AW, McClish DK, Kuhnert PM, Jimenez E, et al. Iron deficiency anemia and iron therapy effects on infant developmental test performance. Pediatrics 1987; 79:981-95. Walter T, De Andraca I, Chadud P, Perales CG. Iron deficiency anemia: adverse effects on infant psychomotor development. Pediatrics 1989;84:7-17. Idjradinata P, Pollitt E. Reversal of developmental delays in iron-deficient anaemic infants treated with iron. Lancet 1993;341:1-4. Moffatt MEK, Longstaffe S, Besant J, Dureski C. Prevention of iron deficiency and psychomotor decline in highrisk infants through use of iron-fortified infant formula: a randomized clinical trial. J Pediatr 1994;125:527-34. Lozoff B, Wolf AW, Jimenez E. Irondeficiency anemia and infant development: effects of extended oral iron therapy. J Pediatr 1996;129:382-9. Roncagliolo M, Garrido M, Walter T, Peirano P, Lozoff B. Evidence of altered central nervous system development in infants with iron deficiency anemia at 6 mo: delayed maturation of auditory brainstem responses. Am J Clin Nutr 1998;68:683-90. Williams J, Wolff A, Daly A, MacDonald A, Aukett A, Booth IW. Iron supplemented formula milk related to reduction in psychomotor decline in infants from inner city
THE JOURNAL OF PEDIATRICS FEBRUARY 2002
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
areas: randomized study. BMJ 1999;318:693-7. Morley R, Abbott R, Fairweather-Tait S, MacFadyen U, Stephenson T, Lucas A. Iron fortification follow on formula from 9 to 18 months improves iron status but not development or growth: a randomized trial. Arch Dis Child 1999; 81:247-52. Deinard A, Gilbert A, Dodds M, Egeland B. Iron deficiency and behavioral deficits. Pediatrics 1981;68:828-33. Lozoff B, Jemenez E, Wolf AW. Longterm developmental outcome of infants with iron deficiency. N Engl J Med 1991;325:687-94. Lozoff B, Jimenez E, Hagen J, Mollen E, Wolf AW. Poorer behavioral and developmental outcome more than 10 years after treatment for iron deficiency in infancy. Pediatrics 2000;105:e51. Goldenberg RL, DuBard MB, Cliver SP, Nelson KG, Blankson K, Ramey SL, et al. Pregnancy outcome and intelligence at age five years. Am J Obstet Gynecol 1996;175:1511-5. Test critiques. In: Keyser DJ, Sweetland RC, editors. Vol. I, V, and VI. Kansas City: Test Corporation of America; 1984, 1986, and 1987. Shaywitz SE, Schnell C, Shaywitz BA, Towle VR. Yale Children’s Inventory (YCI): an instrument to assess children with attention deficit and learning disabilities. I. Scale development and psychometric properties. J Abnorm Child Psychol 1986;14:347-64. Tamura T, Hou J, Goldenberg RL, Johnston KE, Cliver SP. Gender difference in cord serum ferritin concentrations. Biol Neonate 1999;75:343-9. Williams RL, Creasy RK, Cunningham GC, Hawes WE, Norris FD, Tashiro M. Fetal growth and perinatal viability in California. Obstet Gynecol 1982; 59:624-32. Rois E, Lipschitz DA, Cook JD, Smith NJ. Relationship of maternal and infant iron stores as assessed by determination of plasma ferritin. Pediatrics 1975;55: 694-9. Dobbing J, Sands J. Quantitative growth and development of human brain. Arch Dis Child 1973; 48:757-67. Cook JD, Finch CA, Smith NJ. Evaluation of iron status of a population. Blood 1976;48:449-55. Gabay C, Kushner I. Acute-phase proteins and other systemic responses to
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
inflammation. N Engl J Med 1999; 340:448-54. Leviton A, Paneth N, Reuss ML, Susser M, Allred EN, Dammann O, et al. Maternal infection, fetal inflammatory response, and brain damage in very low birth weight infants. Pediatr Res 1999;46:566-75. Tamura T, Goldenberg RL, Johnston KE, Cliver SP, Hickey CA. Serum ferritin: a predictor of early spontaneous preterm birth. Obstet Gynecol 1996; 87:360-5. Goldenberg RL, Tamura T, DuBard M, Johnston KE, Copper RL, Neggers Y. Plasma ferritin and pregnancy outcome. Am J Obstet Gynecol 1996;175:1356-9. Aisen P, Cohen G, Kang JO. Iron toxicosis. Int Rev Exp Pathol 1990;31:1-46. Andorn AC, Britton RS, Bacon BR. Ascorbate-stimulated lipid peroxidation in human brain is dependent on iron but not on hydroxyl radicals. J Neurochem 1996;67:717-22. Sobotka TJ, Whittaker P, Sobotka JM, Brodie RE, Quander DY, Robl M, et al. Nurobehavioral dysfunctions associated with dietary iron overload. Physiol Behav 1996;59:213-9. Friedriksson A, Schröder N, Eriksson P, Izquierdo I, Archer T. Maze learning and motor activity deficits in adult mice induced by iron exposure during a critical postnatal period. Develop Brain Res 2000;119:65-74. Castelnau PA, Garrett RS, Palinski W, Witztum JL, Campbell IL, Powell HC. Abnormal iron deposition associated with lipid peroxidation in transgenic mice expressing interleukin-6 in the brain. J Neuropathol Exp Neurol 1998;572:268-82. Bothwell TH, Charlton RW, Motulsky AG. Hemochromatosis. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic and molecular bases of inherited disease. 7th ed. New York: McGraw-Hill; 1995. p. 2237-69. Garn SM, Keating MT, Falkner F. Hematological status and pregnancy outcomes. Am J Clin Nutr 1981; 34:115-7. Lu ZM, Goldenberg RL, Cliver SP, Cutter G, Blankson M. The relationship between maternal hematocrit and pregnancy outcome. Obstet Gynecol 1991;77:190-4.