Iron-deficiency anemia and infant development: Effects of extended oral iron therapy

Iron-deficiency anemia and infant development: Effects of extended oral iron therapy Betsy Lozoff, Mb, Abraham W. Wolf, PhD, and Elias Jimenez, MD From the Center for Human Growth and Development and the Department of Pediatrics and Communicable Diseases, University of Michigan, Ann Arbor, the Department of Psychiatry, MetroHealth Medical Center, Case Western Reserve UniversitySchool of Medicine, Cleveland, Ohio, and the Hospital Nacional de Nifios, San Jose, Costa Rica

Objective: To determine whether extended oral iron therapy corrects lower developmental test scores in infants with iron-deficiency anemia. Study design: Double-blind, controlled trial in Costa Rica involving 32 12- to 23month-old infants with iron-deficiency anemia ,and 54 nonanemic control subjects. Anemic infants were treated with orally administered iron for 6 months; half the nonanemic children were treated with iron and half with placebo. Developmental test scores and hematologic status were evaluated before treatment, after 3 months, and after 6 months. Results: Iron-deficient anemic infants received lower mental test scores than nonanemic infants at all three time points (p <0.05 pretreatment and at 3 months, p = 0.07 at 6 months). There were no significant differences in motor test scores. More of the anemic infants were rated as unusually fearful and unhappy. Anemic infants came from families with lower maternal education and less support for child development and were less likely to be breast fed, were weaned earlier, and consumed more cow milk. Conclusions: Lower mental test scores persisted in infants with iron-deficiency anemia despite extended oral iron therapy and an excellent hematologic response. Iron-deficiency anemia may serve as a marker for a variety of nutritional and family disadvantages that may adversely affect infant development. (J Pediatr 1996;129:382-9)

Several recent studies have shown that infants with irondeficiency anemia receive lower scores on tests of mental and motor development than peers with better iron status.l-7 Test scores have generally still been lower in a majority of anemic infants after a full course of iron treatment (2 to 3 months).4, 5, 8 However, a recent study with a strong exper-

Supportedby grantsfrom the NationalInstitutesof Health (HD 14122 and I-ID31606). Submitted for publication Oct. 12, 1995; accepted April 23, 1996. Reprint requests: Betsy Lozoff, MD, Director, Center for Human Growth and Development, 300 N. Ingalls, University of Michigan, Ann Arbor, MI 48109-0406. Copyright © 1996 by Mosby-Year Book, Inc. 0022-3476/96/$5.00 + 0 9121/74507

382

imental design reported correction of lower test scores. 6 Because this study involved 4 months of iron therapy, the resuits raise the possibility that correcting lower developmental test scores requires a longer course of iron therapy than that needed to treat anemia. We evaluated this question by testing anemic and nonanemic infants before treatment and again after 3 and after 6 months of carefully supervised oral iron therapy.

METHODS Population. The study was conducted between 1986 and 1990 in Costa Rica, a country with an excellent record of infant health but where iron supplementation was not routine at the time. Health policy changed in 1990 so that supplemental iron is now recommended for all infants.

The Journal of Pediatrics Volume 129, Number 3

The sample was drawn from the periurban community o f Desamparados, located at an elevation of 1185 m, near San Jose, the capital of Costa Rica. Participant identification started with door-to-door screening of the entire community, which was predominantly lower middle class. A finger-stick blood specimen was obtained for hemoglobin determinations for all 12- to 23-month-old infants (total number screened = 441). Those whose finger-stick hemoglobin levels were either 105 gm/L or less or 120 gm/L or more, whose birth weight was 2.5 kg or more, whose singleton birth was uncomplicated, and who were free of acute or chronic medical problems were asked to come to Hospital Nacional de Nifios for a physical examination and a venipuncture blood specimen (179 children were eligible, and 85 % were brought to the hospital for the further examination). Iron status was determined by venous levels of hemoglobin, transferrin saturation, erythrocyte protoporphyrin concentrations, and serum ferritin concentrations. Hemoglobin values at this altitude would be approximately 4 gm/L higher than those at sea level. 9 Infants with moderate iron-deficiency anemia (hemoglobin concentration -<100 gm/L and two of three iron measures indicating iron deficiency: serum ferritin level -< 12 ~ag/L, erythrocyte protoporphyrin concentration >100 ~g/dl packed erythrocytes, or transferrin saturation -< 10%) and nonanemic infants (hemoglobin concentration ~ 12.5 gin/L) were invited to participate in the study. These criteria were designed to produce clearly anemic and clearly nonanemic groups. There were no specific criteria for iron status in the nonanemic group, because our criterion for hemoglobin level was high and iron deficiency without anemia was not associated with lower developmental test scores or behavioral alterations in our previous studies in Central America.4, 10 The study was approved by the committees on human investigation at Case Western Reserve University in the United States and at Hospital Nacional de Nifios, the Social Security Administration, and the National Ministry of Health in Costa Rica. The study was explained to the infants' parents, and signed informed consent was obtained. The original study design called for matching each iron-deficient anemic infant by age and sex with two nonanemic babies. However, the number of infants who met the study's strict entrance criteria for health and hematologic parameters was limited for both the anemic and nonanemic groups, and it soon became clear that~a number of qualifying children would be lost if matches were required. We therefore decided to enroll all qualifying infants and to control for age and sex statistically, if necessary. Of the 39 infants who met the criteria for irondeficiency anemia, parents of 82% agreed to participate, giving 32 iron-deficient anemic infants. Of the 61 infants who met the criteria for the nonanemic comparison group, 89% were enrolled, giving 54 nonanemic infants in the

Lozoff, Wolf, and Jimenez

383

comparison group. There were no differences in sex ratio between the groups, but anemic infants were 1 month younger, on average, than the nonanemic group. During the 6 months of study participation, 6% dropped out (5 children: 4 nonanemic, 1 anemic), primarily because of moving out of the community. Procedure. Nonanemic infants were randomly assigned to receive either an oral iron preparation (3 mg/kg per dose) or placebo drops (a similar volume of carder alone) twice a day for 6 months. All anemic infants received orally administered iron twice a day (3 mg/kg per dose) for 6 months. Neither study families nor project personnel were aware of a child's group or treatment until the end of study participation. Project personnel personally administered the dose of iron or placebo three times per week and recorded the times that the remaining doses of medication were given by family members. The Bayley Scales of Infant Development 11 were administered before and after both 3 months and 6 months of oral iron or placebo treatment. A physical examination and a venipuncture blood specimen were also repeated after 3 and 6 months to monitor the children's health and their hematologic response to iron therapy. Comprehensive information was collected about the child and family, including demography, birth history, nutrition, socioeconomic status, stimulation in the home (HOME Inventoryl2), and parental IQ. Statistical analysis. Differences between anemic and nonanemic groups in demographic, family, hematologic, and nutritional characteristics were analyzed by the Student t test for independent means for continuous variables and by the chi-square test for categorical variables. The Bayley mental and motor test scores were analyzed by analysis of covariance, with group (anemic, nonanemic) as the independent variable and age as the covariate. The Bayley Infant Behavior Record was analyzed as previously described,13, 14 with ratings in the suspect range for abnormality identified on each scale and composite measures of test-affect and taskorientation created.

RESULTS Hematologic and biochemical data (Table I). The anemic infants' initial hemoglobin levels averaged 94 gin/L, whereas those for the nonanemic group averaged 132 gm/L. By definition, all the anemic infants had at least two iron measures in the deficient range. Among the nonanemic group, 30% were iron sufficient (all three iron m e a s l e s in the normal range) and 70% have one or more measures of iron status in the deficient range (32 had a low ferritin level as the sole abnormality, one had a low transferrin saturation, and five had two abnormal measures). Nonanemic infants with some indication of iron deficiency were represented in the iron and placebo conditions in comparable proportions.

384

Table

Lozoff, Wolf, and Jimenez

The Journal of Pediatrics September 1996

I. Hematologic status Anemic (n = 32)

Hemoglobin(gm/L) Before treatment 94.3 + 5.9 After 3 mo 128.5 + 7.6 After 6 mo 132.6 _+7.9 Free erythr0cyteprotoporphyrin (pg/dl packed erythrocytes) Before treatment 335.3.+ 173.6 After 3 mo 65.2 .+ 24~3 After 6 mo 35.8 _+ 18.4 Ferritin (pg/L) Before treatment 4.4 _+4.7 After 3 mo 22,0 _+ 16.5 After 6 mo 27.5 ,+ 19.7 Transferrin saturation(%) Before treatment 8.4 .+ 2.6 After 3 mo 19.5 ,+ 11.0 After 6 mo 22,8 -+ 10.6

Nonanemic (n = 54)

131.6 --. 5.4 133.0 -+ 8.4 133.8 .+ 9.2

p

<0.001 <0.05

59.6 _+24.1 46.5 -+ 23.2 38,9 _+ 19.2

<0.001 <0.001

13.0 .+ 17.1 27.2 -+ 23.6 34.6 .+ 34.8

<0.001

16.8 ,+ 1.9 21.6 -+ 8.3 22.2 _+6.4

Values are means- SD.

The infants showed an excellent response to iron therapy. Among the anemic group, 100% increased their hemoglobin levels by 10 gm/L or more after 3 months and maintained that response at 6 months. The average increase in hemoglobin concentration above initial levels was 34 grn/L at 3 months and 35 gm/L at 6 months. This response confirms that all anemic infants were indeed iron deficient. Anemia was corrected (hemoglobin concentration -> 120 gm/L) in all but one infant at 3 months, with correction of that infant by 6 months. However, two anemic infants who had hemoglobin concentrations of 120 gm/L or more at 3 months dropped their levels to 116 and 113 gm/L at 6 months, With respect to biochemical evidence of iron deficiency, 71% of the anemic group were iron sufficient at both 3 months and 6 months. Among the nonanemic group, 16% showed an increase in hemoglobin concentration of 10 gm/L or more at 3 months and 24% did so at 6 months. The average increase was 7 gm/L at 3 months and 10 gm/L at 6 months. Three infants had a drop in hemoglobin level to less than 120 gm/L at 3 months and two had a similar drop at 6 months (none <110 grn/L). Iron status improved primarily among nonanemic infants who received iron therapy: 78% started the study with one or more measures of iron status in the deficient range, whereas only 16% and 17% had a deficiency at 3 and 6 months, respectively. In contrast, among nonanemic infants receiving placebo, 63% started with at least one abnormal iron value, and 62% and 60% continued to have an abnormal value at 3 and 6 months, respectively. The differential response between the iron- and placebo-treated nonanemic infants confirms that even infants with normal hemoglobin levels in this population could absorb and respond to therapeutic ironJ s

We were unable to identify a reason or reasons that some iron-treated infants became iron sufficient whereas others did not. There were no statistically significant differences in initial hematologic status, growth, feeding, or family variables to distinguish those who became iron sufficient, within either the anemic group or the iron-treated nonanemic group. Overall, the posttreatment iron status of both groups was excellent (Table I), and a much higher percentage were iron sufficient than after 3 months of treatment in our previous study.4 These otherwise healthy infants had an average lead level of 10.4 ± 4.1 ggM1 whole blood (range, 5.6 to 23.8 pg/dl) and normal hemoglobin electrophoreses and A2 levels, had no evidence of deficiencies in folic acid or protein (except for a nonanemic child with a folate level of 1.4 ng/L), were growing in the normal range by U.S. standards, and were free of parasites (except for Giardia lamblia in 3%, Entamoeba histolytica in 1%, and Ascaris in 2%). Developmental test scores and behavior ratings. There were no differences in developmental test scores between nonanemic infants who received iron and those who received placebo, either at study entry or after treatment. Therefore all nonanemic infants were placed into a single nonanemic comparison group. Infants with iron-deficiency anemia had Bayley mental test scores that averaged 6.1 points lower than those of the nonanemic group at study entry and continued to receive lower scores after 3 and 6 months of treatment (Fig. 1). Mental test scores of anemic and nonanemic infants, adjusted for age in analysis of covariance and expressed as themean +_ SEM, wereas follows; at study entry, 99.1 +_ 2.2 versus 105.2 ± 1.7 (F[1,83] = 5.10;p = 0.03); after 3 months, 91.3 -+ 2.1 versus 98.1-+ 1.6 (F[1,78] = 6.50;p = 0.01); and after 6 months, 91.6 +- 2.3 versus 96.9 -+ 1.8 (F[1,76] = 3.28; p = 0.07). Anemic and nonanemic groups both showed some decline in mental test scores with time, which has been fiequently reported in disadvantaged infants in the second year of life. 1°, 16,17 No differences in motor test scores were observed either at study entry or after treatment (Fig. 2). Motor scores were high in both groups at all three time points, averaging around 110 (adjusted for age in analysis of covafiance and expressed as the mean _+ SEM: 109.3 +-_3.1 vs 111.3 _+ 2.4 at study entry; 107.7 - 3.2 vs 110.6 _+ 2.5 after 3 months; and 111.5 +- 3.4 vs 115.1 +_ 2.6 after 6 months). There was no evidence that infants who became iron sufficient after either 3 or 6 months of treatment showed test-score improvements, compared with those who still had biochemical indicators of iron deficiency, regardless of whether they were initially anemic or nonanemic. The lead levels of anemic infants were not significantly different from those of nonanemic infants (11.5 --- 4.0 Dg/dl vs 10.0 __+_4.1 pg/dl; t(1,62) = 1.36, p = 0.18), and lead levels were not correlated with either mental or motor test scores

The Journal of Pediatrics Volume 129, Number 3

Lozoff, Wolf, and Jimenez

"O-

115

Iron-Deficient

Anemic

115

- - ~ - - Nonanemic

"~

110

385

p=.03 = .01

7

110

105 ¢J

nS ns

100 -O-

95 O

Nonanemic

95

9O I

Study Entry

I

3 Months

m

I

6 Months

Fig. 1. Mental development test scores. The Mental Development Index (adjusted mean + SEM with age as covariate) of the Bayley Scales of Infant Development are shown for the iron-deficientanemic and nonanemic groups before treatment, after 3 months, and after 6 months. The anemic group's scores were lower than those of the nonanelNc group at all three time points: statistically significant before treatment and at 3 months, and showing a suggestive trend at 6 months.

(r--0.02 and r = 0.01, respectively). However, anemic infants were disadvantaged in several other respects, compared with the nonanemic group (Table II). They came from families with less well educated mothers and lower HOME scores (there were no statistically significant differences in parental IQ). Although almost all infants (90%) were initially breast fed, seven of eight infants who were totally bottle fed were in the anemic group (two-tailed Fisher Exact Test, p = 0.004). Among those who were breast fed, anemic infants were weaned earlier. Anemic infants also consumed more cow milk. Despite these disadvantages in feeding, there were no statistically significant differences in growth with or without control for age (weight, length, weight-for-age, length-for-age, weight-forqength, head circumference, or left upper ann circumference). To take into account the effects of family and nutritional disadvantages on test scores, we performed a multiple regression analysis, entering these variables into the equation before considering the effects of iron-deficiency anemia. None of the individual background variables except for infant age contributed significantly to the model, and the effect of iron deficiency was no longer statistically significant after all the background factors were considered (p = 0.27). On the Infant Behavior Record, there were no differences between anemic and nonanemic groups in the composite summaries of test affect and task orientation, either before or after treatment. However, analysis of the individual scales that make up these composites indicated abnormalities in

Iron-Deficient Anemic

Study Entry

3 Months

6 Months

Fig. 2, Motor development test scores. The Psychomotor Development Index (adjusted mean _+ SEM with age as covariate) of the Bayley Scales of Infant Development are shown for the iron-deficient anemic and nonanemic groups before treatment, after 3 months, and after 6 months. There were no statistically significant differences at any time point.

affect before treatment. A greater proportion of anemic than nonanemic infants were rated as unusually fearful (53% vs 30%; chi-square value = 4.69, p = 0.03) and unhappy (38% vs 11%; chi-square value = 8.46, p <0.01), and there was a suggestive trend that more of them were excessively wary or hesitant with the examiner (53% vs 35%; chi-square value=2.68, p = 0.10). No differences between anemic and nonanemic groups on Infant Behavior Record ratings were observed after 3 or 6 months of treatment. DISCUSSION This study found that infants with iron-deficiency anemia received lower mental test scores and showed affective abnormalities before treatment, did not improve their mental test scores after iron therapy, and were disadvantaged compared with nonanemic infants with respect to family background and feeding practices. The observations of lower mental test scores and alterations in affect form a consistent pattern with other studies of infants with iron-deficiency anemia. All seven studies that included careful definitions of iron status and appropriate comparison groups have found lower mental test scores before treatment. I-6, this studyFive of seven reported lower motor scores as well. 14, 6 Affective differences, such as wariness, fearfulness, or unhappiness, have been reported in all studies that rated infmlt behavior (Lozoff B, Klein NK, Nelson EC, et al.: unpublished observations).2, 5, 18, 19 Affective abnormalities were no longer observed at 3 or 6 months in the present study. It is tempting to conclude that this improvement was an effect of iron treatment. However,

386

Lozoff, Wolf, and Jimenez

The Journal of Pediatrics September 1996

II. Demographic, family, and nutritional characteristics

Table

Sex (%male) Age (mo) Birth order Birth weight (kg) Days to discharge Mother's education (yr) Mother's IQ score HOME score Breast fed Age at weaningif breast fed (mo) Milk consumption (oz/day) Weight (kg) Length (cm) Weight-for-length percentile Head circumference (cm) Left upper arm circumference (cm)

Anemic (n = 32)

Nonanemic (n = 54)

50% 15.5 _+3.1 3.2 _+2.0 3.21 +- 0.29 1.5 _+ 1.0 7.3 -+ 2.6

48% 17.0 + 3.5 2.7 + 1.6 3.35 -+ 0.52 i.4 +- 0.8 9.1 -+ 2.8

<0.01

79.8 _+ 11.4 25.4 _+6.6 78% 5.0 ~+4.1

83.7 -+ 12.8 29.0 + 5.5 98% 7.2 _+4.6

<0.01 <0.01 <0.05

32.7 + 11.2

22.2 +_ 1.8

<0.001

10.2 _+ 1.4 77.2 + 5.0 53.3 _+27.1

10.4 _+ 1.4 79.2 -+ 4.4 43.6 -+ 27.3

46.1 _+ 1.5

46.6 + 1.5

15.3 _+ 1,1

15.4 +- 1.1

p

<0.05

Values are means-+ SD.

evidence from our previous studies suggests that other factors explain the change. In both the Guatemala study 18 and the earlier Costa Pica study (unpublished data), affective abnormalities in iron-deficient anemic infants improved after 1 week regardless of iron or placebo treatment. Thus it appeared that infants with iron-deficiency anemia respond to unfamiliar, somewhat stressful situations with wariness or hesitance, which becomes less apparent on repeated contact with the setting, people, and procedures. In the present study, however, affective improvement cannot be attributed to iron therapy or another factor with confidence, because no placebo-treated anemic group was included because of ethical concerns about withholding iron therapy for 3 to 6 months. The lack of a treatment effect is clear in the present study with respect to mental test scores. No improvement relative to nonanemic infants was observed after 3 or 6 months of iron therapy. This lack of improvement in mental test scores is consistent With previous studies in Costa Pica, Chile, and England, in which the majority of anemic infants continued to show lower test scores after 2 to 3 months of treat• merit.4, 5, 8 These findings contrast with a recent study in Indonesia in which Idjradinata and Pollitt6 reported dramatic increases in test scores after iron treatment--19 points in mental scores and 24 points in motor scores (compared with

minimal change in placebo-treated anemic infants). Because the Indonesia study and the present study are the only ones so far to include extended iron treatment (4 to 6 months), trying to understand the differing results is important. Duration of treatment seems unlikely to be the explanation, because we found no improvement in test scores even after 6 months of treatment in the present study, despite an excellent hematologic response. Severity of iron-deficiency anemia and infant age are also unlikely explanations, because hematologic status and infant age were similar in both studies. Pollitt (personal communication, 1995) suggests that environmental factors may be critical. The Indonesian families were solidly middle class, whereas the Costa Pican and Chilean families may have been more disadvantaged (though still better off than in most developing countries). Another difference was that the pretreatment scores of the Indonesian anemic children, relative to the nonanemic group, were lower than those in Costa Pica and Chile, hence providing more room for improvement. However, it is unclear why iron-deficient anemic infants in Indonesia had much lower relative scores, especially if they came from middle-class families. Feeding practices may be relevant. Although breast-feeding is common in both Indonesia and Costa Pica, it is possible that early weaning or use of unmodified cow milk explained even more of the anemia in Indonesia than in Costa Pica. Since several recent studies indicate that breast-feeding is linked to better mental development,2°-23 a difference in feeding practices could account for lower initial scores in iron-deficient anemic infants in Indonesia. These uncertainties notwithstanding, the results in Indonesia and the fact that a minority of iron-deficient anemic infants did show improvement with treatment in two other studies4, s suggest that there may be some conditions under which iron therapy is effective in correcting lower developmental test scores. Response to iron therapy seemed to be a promising explanation. In an earlier study in Costa Pica, we found that the minority of anemic infants who became iron sufficient (about one third) showed a 10-point improvement in motor test scores and no decline in mental scores. 4 Similarly, a study in England found improvement in a subgroup of iron-deficient anemic infants (31%) who had the most improvement in hemoglobin levels.8 However, no improvement in test scores, regardless of hematologic response to iron therapy, was noted in the present study or in a study in Chile. 5 Thus, after 2 to 6 months of treatment, one study reported an overall improvement in mental and motor test scores, 6 two studies observed continued lower test scores in the majority of anemic infants, with improved scores in a minority who showed the best response to iron therapy, 4, 8 and two found no improvement at all, regardless of hematologic response. 5' this studyTherefore the ability of iron ther-

The Journal of Pediatrics Volume 129, Number 3

apy to correct lower test scores or the conditions under which iron therapy can produce test score improvements remain open questions. Concerns about reversibility with treatment are especially worrisome, because all available long-term follow-up studies show that children treated for iron-deficiency anemia as infants still test lower in mental and motor functioning years later. 24-2saA lack of correction of lower test scores with iron treatment has several potential explanations. A theoretical possibility is that the effects of early iron-deficiency anemia are irreversible. This explanation would parallel observations in the developing rat that a reduction in brain iron and some behavioral changes caused by early iron-deficiency anemia are not corrected with treatment. 29-33Another potential explanation is that the lower test scores are not caused by iron-deficiency anemia but, rather, by some other closely associated factor(s), such as deficiency of another nutrient or environmental disadvantage. 36 No other nutrient deficiency has yet been described,4, 5 but the possibility must be kept in mind. However, differences in other factors that could account for the observed behavioral alterations have consistently been found in studies of iron-deficient anemic infants. Although healthy and born at term, anemic infants nonetheless averaged 150 to 250 gm less at birth I' 2,4 and were breast fed for a shorter period. 4, 37. thisstudy These observations may point to preexisting behavioral differences: babies who are small for gestational age, even if their birth weights are in the normal range, may differ in behavior from heavier babies, 38, 39 and difficult behavior, regardless of its cause, might lead to earlier weaning. There is also consistent evidence that iron-deficiency anemia is more common in less advantaged families.4, 40-43 For instance, in both studies in Costa Rica, the mothers of anemic infants had less education or lower IQ scores than the comparison group and lower scores on the HOME Inventory. 4, thisstudyThe effects of iron-deficiency anemia on developmental test scores were still statistically significant after control for these differences in the first study but not in the present one. The reason may be the smaller smnple size in the current study. Power analysis indicated that we had a greater than 90% chance of detecting an 11-point difference between anemic and nonanemic groups after control for background variables but could not have detected a smaller difference with confidence. Nonetheless, poorer developmental outcome in iron-deficiency anemia may be inextricably linked to other health, nutrition, and environmental disadvantages,36 as is the case in several other risk conditions.44-48 Experimental manipulation of iron status is the most effective way to control for potential behavioral and environmental differences. The requisite study design is a prospective clinical trial in which iron-deficiency anemia is experimentally prevented in one group of infants but not in another

Lozoff Wolf, and Jimenez

387

in a double-blind, placebo-controlled, randomized design. Two preventive trials have been published, and a third large trial is nearing completion in Chile.49 The first, by Heywood et al.5° in Papua New Guinea, compared 1-year-old infants, half of whom had received intramuscularly administered iron at 2 months of age, and half a placebo injection at the same age. Although the design of the study was strong, the results are difficult to interpret because malaria was endemic, all groups were anemic at 12 months, and iron status measures did not clearly indicate iron deficiency. However, it seemed that iron-treated infants who had no evidence of malaria showed better attentional abilities. The other preventive trial, recently published by Moffatt et al.,5~ included 283 almost entirely Native American infants in Canada, half of whom had received iron-fortified formula and half unfortified formula from birth, with follow-up until 15 months. The groups were initially comparable in iron status, development, and family background but diverged in hemoglobin concentrations, iron status, and psychomotor test scores at 9 and 12 months of age, with poorer outcome in the unfortified group. This study, because of its design, provides convincing evidence that iron deficiency causes lower developmental test scores in infancy. Two findings qualify this conclusion, however. (1) No differences in mental scores were noted, even though they have been found in all available case-control studies. 1-6' this study (2) The differences in motor scores resolved spontaneously by 15 months of age. These results also raise the possibility that some other factor(s) accounts for the differences in mental and motor development consistently observed in iron-deficient anemic infants. In sum, the present study, combined with earlier research, documents that iron-deficient anemic infants, in comparison with peers with better iron status, receive lower developmental test scores, show affective abnormalities before treatment, and are disadvantaged with respect to breastfeeding and family background. The lack of improvement in developmental test scores after iron therapy in the majority of anemic infants in this study, despite 6 months of treatment and an excellent hematologic response, is similar to findings of other studies in Costa Rica, Chile, and England. Differing results regarding the effectiveness of iron therapy and the role of other disadvantages suggest that environmental factors or unidentified factors other than iron-deficiency anemia may be essential in understanding the behavioral and developmental effects of iron-deficiency anemia in infancy. Although there remain unanswered questions about a causal relationship between iron-deficiency anemia and lower developmental test scores, preventing iron-deficiency anemia in infancy still seems to be the safest course, given that irondeficiency anemia affects more than 20% to 25% of babies worldwideS2, 53 and many poor or minority children in the United States. 54

388

Lozoff, Wolf, and Jimenez

We are grateful to the study families and to the project's dedicated team: the pediatrician and psychologists doing the community field work, Yvonne Gomez, MD, Angela Radan, and Patricia A1varado; the hematology laboratory personnel, Rafael Jimenez, MQC, MSc, and Luis Mora, MQC; Maria Elena Chacon, who administered developmental tests; and Denney Artavia, who performed data entry. REFERENCES 1. Lozoff B, Brittenham GM, Vitefi FE, Wolf AW, Urrutia JJ. The effects of short-term oral iron therapy on developmental deficits in iron- deficient anemic infants. J Pediatr 1982; 100:351-7. 2. Walter T, Kovalskys J, Stekel A. Effect of mild iron deficiency on infant mental development scores. J Pediatr 1983;102:51922. 3. Gfindulis H, Scott PH, Belton NR, Wharton BA. Combined deficiency of iron and vitamin D in Asian toddlers. Arch Dis Child 1986;61:843-8. 4. Lozoff B, Brittenham GM, Wolf AW, et al. Iron deficiency anemia and iron therapy: effects on infant developmental test performance. Pediatrics 1987;79:981-95. 5. Walter T, de Andraca I, Chadud P, Perales CG. Iron deficiency anemia: adverse effects on infant psychomotor development. Pediatrics 1989;84:7-17. 6. Idjradinata P, Pollitt E. Reversal of developmental delays in iron-deficient anaemic infants treated with iron. Lancet 1993; 341:1-4. 7. Wasserman G, Graziano JH, Faetor-Litvak P, et al. Independent effects of lead exposure and iron deficiency anemia on developmental outcome at age 2 years. J Pediatr 1992;121: 695-703. 8. Aukett MA, Parks YA, Scott PH, Wharton BA. Treatment with iron increases weight gain and psychomotor development. Arch Dis Child 1986;61:849-57. 9. Cartwfight GE. Normal values and corpuscular constants. In: Cartwfight GE, editor. Diagnostic laboratory hematology. New York: Grune & Stratton, 1968:109-19. 10. Lozoff B. Developmental deficits in iron deficient infants: effects of age and severity of iron lack. J Pediatr 1982;101:94852. 11. Bayley N. Bayley scales of infant development. New York: Psychological Corp., 1969. 12. Caldwell BM. Instruction manual: home inventory for infants (revised edition). Little Rock: University of Arkansas Center for Early Child Development, 1975. 13. Wolf AW, Lozoff B. A clinically interpretable method for analyzing the Bayley Infant Behavior Record. J Pediatr Psyehol 1985;10:199-214. 14. Matheny AP Jr. Bayley's Infant Behavior Record: behavioral components and twin analyses. Child Dev 1980;51:1157-67. 15. Jimenez E, Lozoff B, Jimenez R. Hematologic response to iron in nonanemic infants [abstract]. Washington (DC): Ambulatory Pediatric Association, 1985. 16. Brown RE, Halpefin F. The variable pattern of mental development of rural black children. Clin Pediatr 1971;10:404-09. 17. Grantham-McGregor SM, Powell C, Fletcher P. Severe malnutrition, stunting and mental development in young children. Eur J Clin Nutr 1989;43:403-9. 18. Lozoff B, Wolf AW, Urrutia JJ, Viteri FE. Abnormal behavior and low developmental test scores in iron-deficient anemic infants. J Dev Behav Pediatr 1985;6:69-75.

The Journal of Pediatrics September 1996 19. Lozoff B, Klein NK, Prabucki KM. Iron-deficient anemic infants at play. J Dev Behav Pediatr 1986;7:152-8. 20. Rodgers B. Feeding in infancy and later ability and attainment: a longitudinal study. Dev Med Child Neurol 1978;20:421-6. 21. Rogan JW, Gladen BC. Breast feeding and cognitive development. Early Hum Dev 1993;31:181-93. 22. Ferguson DM, Beautris AL, Silva PA. Breast-feeding and cognitive development in the ftrst seven years of life. Soc Sci Med 1982;16:1705-8. 23. Lucas A, Morley R, Cole TJ, Lister G, Leeson-Payne C. Breast milk and subsequent intelligence quotient in children born preterm. Lancet 1992;339:261-4, 24. Lozoff B, Jimenez E, Wolf AW. Long-term developmental outcome of infants with iron deficiency. N Engl J Med 1991; 325:687-94. 25. Walter T, de Andraca I, Castillo M, Rivera F, Cobo C. Cognitive effect at 5 years of age in infants who were anemic at 12 months: a longitudinal study [abstract]. Pediatr Res 1990;28: 295. 26. Palti H, Pevsner B, Adler B. Does anemia in infancy affect achievement on developmental and intelligence tests? Hum BiN 1983;55:189-94. 27. Patti H, Meijer A, Adler B. Learning achievement and behavior at school of anemic and non-anemic infants. Early Hum Dev 1985;10:217-23. 28a. Hurtado E. Cross-linking a third data set: long-term effects of iron deficiency during infancy and early childhood [abstract]. Indianapolis (IN): Society for Research in Child Development, 1995. 28. Dommergues MP, Archambeand B, Ducot Y, et al. Carence en fer et tests de drveloppement psychomoteur: 6tude longitudihale entre l'~ge de 10 mois et l'~ge de 4 ans. Arch Fr Pediatr 1989;46:487-90. 29. Dallman PR, Siimes M, Manies EC. Brain iron: persistent deficiency following short-term iron deprivation in the young rat. Br J Haematol 1975;31:209-15. 30. Dallman PR, Spirito RA. Brain iron in the rat: extremely slow turnover in normal rats may explain long-lasting effects of early iron deficiency. J Nutr 1977;107:1075-81. 31. Weinberg J, Levine S, Dallman PR. Long-term consequences of early iron deficiency in the rat. Pharmacol Biochem Behav 1979; 11:631-8. 32. Youdim MBH, Green AR, Bloomfield MR, Mitchell BD, Heal D J, Grahame-Smith DG. The effects of iron deficiency on brain biogenic monoamine biochemistry and function in rats. Neuropharmacology 1980;19:259-67. 33. Ben-Shachar D, Ashkenazi R, Youdim MBH. Long-term consequence of early iron-deficiency on dopaminergic neurotransmission in rats. Int J Dev Neurosci 1986;4:81-8. 34. Yehuda S, Youdim MBH, Mostofsky DI. Brain iron deficiency Causes reduced learning capacity in rats. Pharmacol Biochem Behav 1986;25:141-4. 35. Felt BT, Lozoff B. Brain iron and behavior are not normalized by treatment of iron deficiency anemia during early developmerit. J Nutr 1996;126:693-701. 36. Lozoff B. Has iron deficiency been shown to cause altered behavior in infants? In: Dobbing J, editor. Brain, behavior, and iron in infant diet. London: Springer-Verlag, 1990:107-31. 37. Walter T. Iron deficiency and behavior in infancy: a critical review. In: Dobbing J, editor. Brain, behavior, and iron in the infant diet. London: Springer Verlag, 1990:133-55. 38. Beeghly M, Flannery K, Birss S, et al. Cognitive and psycho-

The Journal of Pediatrics Volume 129, Number 3

39.

40.

41.

42. 43.

44.

45.

Lozoff, Wolf, and Jimenez

logical development of small-for-gestational age (SGA) infants: a follow-up study [abstract]. Kansas City (MO): Society for Research in Child Development, 1989. Gorman KS, Pollitt E. Relationship between weight and body proportionality at birth, growth during the first year of life, and cognitive development at 36, 48, and 60 months. Infant Behav Dev 1992;15:279-96. Life Sciences Research Office. Assessment of the iron nutrition status of the U.S. Population based on data collected in the Second National Health and Nutrition Survey, 1976-1980. Bethesda: Federation of American Societies for Experimental Biology, 1984. Czaika-Narins DM, Haddy TB, Kallen DJ. Nutrition and social correlates in iron deficiency anemia. Am J Clin Nutr 1978; 31:955-60. Owen GM, Lubin AH, Garry PJ. Preschool children in the United States: who has iron deficiency? J Pediatr 1971 ;79:563 - 8. Zee P, DeLeon M, Roberson P, Chen CH. Nutritional improvement of poor urban preschool children. JAMA 1985;253: 3269-72. Escalona SK. Babies at double hazard: early development of infants at biological and social risk. Pediatrics 1982;70: 670-6. Barrett DE. Nutrition and social behavior. In: Fitzgerald HE, Lester BM, Yogman MW, editors. Theory and research in be-

46. 47. 48. 49.

50.

51.

52. 53. 54.

389

havioral pediatrics; vol 3. New York: Plenum Press, !986:14798. Levitsky DA. Malnutrition, environment, and behavior. Ithaca, New York: Comell University Press, 1979. Dobbing J. Early nutrition and later achievement. London: Academic Press, 1987. Lozoff B. Nutrition and behavior. Am Psycho1 1989;4:231-6. Lozoff B, de Andraca I, Walter T, Pino P. Does preventing iron-deficiency anemia (IDA) improve developmental test scores [abstract]? Pediatr Res 1996;39:136A. Heywood A, Oppenheimer S; Heywood P, Jolley D. Behavioral effects of iron supplementation in infants in Madang, Papua New Guinea. Am J Clin Nutr 1989;50(Suppl):630-40. Moffatt MEK, Longstaffe S, Besant J, Dureski C. Prevention of iron deficiency and psychomotor decline in high-risk infants through iron-fortified infant formula: a randomized clinical trial. J Pediatr 1994;125:527-34. Bothwell TH, Charlton RW, Cook JD, et al. Iron metabolism in man. Oxford: Blackwell Scientific Publications, 1979. deMaeyer E, Adiels-Tegman M. The prevalence of anaemia in the world. World Health Stat Q 1985;38:302-16. Yip R, Parvanta I, Scanlon K, Bortand EW, Russell CM, Trowbridge FL. Pediatric nutrition surveillance system-United States, 1980-1991. MMWR Morb Mortal W!dy Rep 1992;41(SS-7):1-24.

AVAILABILITY OF JOURNAL BACK ISSUES As a service to our subscribers, copies of back issues of The Journal of Pediatrics for the preceding 5 years are maintained and are available for purchase from Mosby until inventory is depleted, at a cost of $11.00 per issue. The following quantity discounts are available: 25% off on quantities of 12 to 23, and one third off on quantities of 24 or more. Please write to Mosby-Year Book, Inc., Subscription Services, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, or call (800)453-4351 or (314)453-4351 for information on availability of particular issues. If unavailable from the publisher, photocopies of complete issues may be purchased from UMI, 300 N. Zeeb Rd., Ann Arbor, MI 48106, (313)761-4700.