QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pollitt, E. Search for Related Content PubMed PubMed Citation Articles by Pollitt, E.

---

Supplement

The Developmental and Probabilistic Nature of the Functional Consequences of Iron-Deficiency Anemia in Children^{1 ,2}

Department of Pediatrics and Program of International Nutrition, University of California, Davis, CA 95616.

	ABSTRACT

TOP ABSTRACT INTRODUCTION Nature of psychobiological... Differential effects and... Time of measurement Psychometrics of development Summary and conclusions REFERENCES

 
It is often assumed that the psychometric tools currently available measure accurately the effects of iron-deficiency anemia (IDA) on cognition in young children and that such effects are rooted in cerebral changes. It is also assumed that snapshots of development within a clinical trial can document such effects. I challenge these assumptions on the basis of four considerations. The first is that there are multiple biological, physical and social-psychological factors that reorient the trajectory of different psychobiological domains in early life after intense and prolonged stress. Further, psychobiological development changes are not necessarily caused by brain changes; there are other mechanisms that also affect development (e.g., biomechanics). A second consideration focuses on intraindividual, interindividual and intergroup differences concerning the nature of the effect of IDA and the response to iron treatment. Individual and group factors can moderate the effects of IDA; for example, different stages of iron deficiency involve different systemic changes, which in turn affect different psychobiological domains. The third consideration is that differences in the time of measurement of an intervention within a randomized trial could lead to detecting effects in different domains or effects of different intensity within the same domain. Finally, developmental assessments with the traditional developmental scales during the first 18 mo of life yield equivocal findings. Snapshots of development will overlook the course of effects of a nutrition intervention over time. Repeated measures over time within the same domain are considered particularly useful to draw the course of development.

KEY WORDS: • anemia • iron • development • clinical trials • Bayley Scales

	INTRODUCTION

TOP ABSTRACT INTRODUCTION Nature of psychobiological... Differential effects and... Time of measurement Psychometrics of development Summary and conclusions REFERENCES

 
Three ideas hide behind the research done on the effects of iron-deficiency anemia (IDA)³ on mental and cognitive development (Grantham-McGregor and Ani 2001 ). One is that the psychometric tools used yield aggregate scores that accurately represent the nature and size of the effect. A second idea is that this comparatively poor psychometric score is explained by particular changes in the architecture and biochemistry of the brain. The third is that randomized clinical trials aimed at the measurement of a main effect (e.g., intergroup difference of an intragroup change) from the nutritional factor are the optimal way in which to assess the effects of iron deficiency on cognition in children. These ideas overlook the complexity of psychobiological development and lead to unrealistic expectations of how much we know about the functional effect of micronutrient deficiencies (e.g., iron or zinc) and about the potential success of intervention programs.

The comments that follow challenge some aspects of the validity of these ideas and offer alternatives to formulate the problem. To this end, I will discuss four interrelated conceptual and methodological issues, i.e., the nature of psychobiological differentiation, differential responsivity, time of measurement and psychometrics of development.

	Nature of psychobiological differentiation

TOP ABSTRACT INTRODUCTION Nature of psychobiological... Differential effects and... Time of measurement Psychometrics of development Summary and conclusions REFERENCES

 
This section focuses on the multiple factors that are potentially capable of influencing the development of children with IDA and on different pathways through which nutritional factors can affect development.

Multiplicity of factors.

Infants and toddlers, the subjects of particular concern to researchers in the field, undergo fast structural and functional cerebral changes that make it possible for the organism to compensate for or adjust to stressful events, that is, to reorient the trajectory of different psychobiological domains, including cognition, toward a normal developmental path, even after having experienced intense and prolonged biological and social-psychological stress (Gottlieb 1991 , Pollitt 2000 ). Several sets of factors can either protect or compromise the expected changes and the trajectories of such domains. In the present context and in the particular case of cognition, it is useful to recall the scope of such factors (e.g., social, biological or personal) and recognize the breadth of the scientific literature that documents their influential roles (Wachs 2000 ). The determinants of development within the social realm include culture, community and family environment, parental education and occupation, early childhood education and formal schooling. The biological realm contains, in addition to genetics, variables such as the maturation of neural structures and neurochemical systems, and pre- and postnatal health and nutrition. Individual life histories include particular experiences, life events and microenvironmental influences that are unique to each person across the person’s life span (Johnson 1997 ). The nature and timing of the factors that contribute to orient the organism toward a developmental course are not fixed, and their combined effects are generally neither additive nor subtractive (Gottlieb et al. 1998 , Johnson 2000 , Sameroff 1995 ). Complex bidirectional, vertical and horizontal interactions among the levels of these factors and the relationships among the factors themselves often do not lawfully follow the codes from the genome (Wachs 2000 ).

System theories of psychobiological development (Gottlieb 1991 , Gottlieb et al. 1998 , Lewis 2000 , Sameroff 1995 , Thelen and Smith 1998 ) are perhaps the closest approximations there are today to a general framework for the study of human development. It is proposed that in addition to the cerebral changes and brain-behavior relationships, other powerful influential forces shape the course of development. I am referring to particular intraorganism relations among different psychobiological domains and to interactions of the organism with the social and physical environment. Cerebral changes and these internal and external processes, in tandem, orient the paths of development and determine in part the changing probabilities of deviance (Pollitt 2000 ). Delays in the timing and interference with these processes (see below) will delay the gradual transformation of behavioral systems into higher levels of functional complexity.

Pathways of influence.

Causes of deviations of development secondary to a nutritional deficit begin sometimes, but not always, with alterations or changes in the brain. This claim is partially justified in this section by focusing selectively on unidirectional and bidirectional relations from motor development to physical activity to emotional regulations.

Motor development and physical activity.

A study in Pangalengan, Indonesia, depicts the relationships among three motor variables over a 12-mo period in a cohort of 12-mo-old infants (Jahari et al. 2000 ). The variables were gross motor milestones (e.g., crawling or walking), the score on the Bayley Scales of Motor Development (i.e., items of both fine and gross motor movements) and motor activity expressed as a multiple of basal metabolic rate. To determine the degree of association, Spearman correlation coefficients were calculated for each pair of these three variables every 2 mo. Consistently, the age of the subjects was negatively related to the size of the correlation coefficients. For example, the correlation between the motor milestones and motor activity ranged from 0.7 to 0.8 at 12 and 14 mo of age, respectively. This correlation fell to about 0.4 at 18 mo and to close to 0 at 24 mo. However, as we have shown elsewhere (Pollitt et al. 1999 ), the level of motor activity in this 12-mo-old cohort did not decrease at any time during the experiment—it either increased or reached a plateau.

The children studied in Pangalengan were separated randomly into three groups that received one of the following daily supplements for 6 mo: E, 1171 kJ + 12 mg Fe; M, 209 kJ + 12 mg of Fe; and S, 104 kJ. All of the bivariate correlations (motor milestones and activity, Bayley motor and motor milestones, Bayley motor and activity) showed that the pattern of the correlations between any two variables was similar to the pattern observed for all groups combined. However, in every bivariate correlation case, the size of the correlations of the children who received E dropped first. In fact, the correlations dropped to zero, months before the same occurred in the other two groups. Moreover, the children who received the E supplement displayed more physical activity than did the other groups (Jahari et al. 2000 ).

Thus, a nutritional factor determined in part the timing at which two different closely related behavioral processes broke away from each other. Entrance to y 2 of life was accompanied by a reduction in the contribution that new motor skills made to locomotion. Other motor changes (e.g., neuromotor organization) and factors (e.g., motivational or contextual) were likely to take the roles of main determinants of motor actions.

IDA constrains the oxygen supply and the respiratory capacity of muscles in humans as observed in studies of exercise physiology and work performance (Zhu and Haas 1999 ) and experiments with animals in the laboratory (Brooks 1994 ). It is plausible that for children, these same effects might delay the acquisition of new motor skills and explain the lag in motor development and physical activity observed in the presence of IDA (Harahap et al. 2000 , Stoltzfus et al., unpublished data, 2000). This limitation of the biomechanical system would then extend to the area of cognitive development because motor activity contributes to the alignment of visual spatial perception (Thelen and Smith 1998 ) and intersensory organization (Gottlieb et al. 1998 ).

Physical activity and emotional regulation.

During the first 2 y of life, creeping, crawling, walking and even running decrease the physical dependence of the young on the caretakers. Physical activity contributes to the activation of mechanisms for the control of emotions. In one study, for example, after the effects of age were controlled for, walking without assistance was followed by emotional changes reflecting autonomy and assertiveness. Children became more sociable and affectionate after the acquisition of walking skills (Birengen et al. 1995 , Campos et al. 1992 ). Thus, delays in locomotion and motor activity delay access to external sources of emotional regulation and constrain the development of self-sufficiency and independence. It follows that anemic children are at risk of delaying the acquisition of developmentally appropriate systems of emotional regulation.

This last hypothesis and others, such as the proposition of linear relationships between zinc supplementation and undifferentiated increments of physical activity (Bentley et al. 1997 , Black 1998 ) and enhanced exploratory behavior, should be tested. A cross-sectional study (Walka and Pollitt 2000 ) and a longitudinal study (Pollitt et al. 2000 ) based on several hours of observation over several months observed a negative correlation between physical activity and the manipulation of objects in toddlers (Pollitt et al. 2000 , Walka and Pollitt 2000 ). Object manipulation as a cognitive action involves perceptual information processing, which requires calmness (i.e., neutral or positive emotions) and motor control (Celsing et al. 1986 ). Further, the different levels of play behavior (e.g., manipulative, relational or symbolic) in infants, toddlers and preschool children involve different degrees of motor activity (Walka and Pollitt 2000 ).

At another level of analysis, the biological plausibility of a particular biobehavioral relationship in this field of study is neither a justification for a claim of a mechanism or even for having identified a pathway of causation. Alterations in the dopamine system, particularly in the dopamine receptors (Nelson et al. 1997 ) and the identification of lags in myelin formation (Connor and Menzies 1996 , Roncagliolo et al. 1998 ), among other neurobiological findings, represent major scientific advancements in our understanding of the effects of IDA. They shed light on the mechanisms behind the reductions of motor activity or delays in motor development that have been observed among IDA laboratory animals (Hunt et al. 1994 ) and children (Harahap et al. 2000 ). However, there is still an ample gap between the evidence on cerebral changes and the documented behavioral delays of IDA subjects. For example, we have no empirical evidence concerning whether the lags in myelination explain the delays of locomotion observed in IDA children or whether they account for alterations in information processing through any of the sensory channels active in early life.

	Differential effects and responsivity

TOP ABSTRACT INTRODUCTION Nature of psychobiological... Differential effects and... Time of measurement Psychometrics of development Summary and conclusions REFERENCES

 
This section focuses on interindividual and intergroup differences concerning the nature of the functional effects of iron deficiency and of the different responses to iron therapy. It is plausible that such individual and group differences are related to the physiological spectrum of IDA and to the different factors (e.g., comorbidity) that moderate effects and responses to treatment.

A spectrum of iron deficiency.

The several stages of iron depletion lead to an incremental involvement of different systems in the body that activate distinct mechanisms within and among domains. It is plausible, for example, that some cerebral changes occur soon after the reserve of iron is depleted and there is a decreased activity in nutrient-dependent enzymes (Bruner et al. 1966, Oski et al. 1983 ). This could be followed by constraints in motor development and physical activity secondary to a drop in the oxygen supply to muscle fibers (Brooks 1994 ). At a final stage, severe anemia would alter higher cognitive functions (Stoltzfus et al., unpublished data, 2000).

Factors that moderate effects and responses to treatment.

A study in Bandung, Indonesia (Idjradinata and Pollitt 1993 ) is the only randomized, therapeutic trial of iron among infants and toddlers that showed large and significant changes in the mental development of anemic [hemoglobin (Hb) <100.5 g/L] infants and toddlers (14.2 mo) selected from middle-class homes. There was a change of 38 g/L (from 95.6 to 129.4 g/L) in the mean hemoglobin of those IDA subjects over the 4 mo of treatment [3 mg/(kg · d)]. The hemoglobin change in the anemic subjects who received a placebo was 10 g/L. Similarly, the mean score obtained in the Bayley mental scale of those who received iron changed from 88.8 to 108.1. There was a negligible change in those who received a placebo (92.4 to 92.9). The intergroup difference in intragroup changes was highly significant.

An unexpected finding in that same study (Idjradinata et al. 1994 ) was the difference between the weight gain of the iron-sufficient (Hb >120 g/L; serum ferritin >12 µg/L) controls who did and did not receive ferrous sulfate. The placebo group gained significantly more weight than did the iron-supplemented group (mean ± SEM; 0.016 ± 0.010 vs. 0.070 ± 0.011 kg every 2 wk). There were no differences in linear growth rate.

The findings from the Indonesian study contrast sharply with the findings from two studies in Costa Rica. A 3-mo iron intervention [1 wk at 10 mg/(kg · d) followed by 12 wk at 6 mg/(kg · d)] given to infants (mean age 16.5 mo) with moderate anemia (Hb <100.5 g/L) resulted in an average hemoglobin increase of 37 g/L and a remission of the anemia (Lozoff et al. 1987 ). However, 64% of the subjects with IDA still had elevated erythrocyte protoporphyrin (EP; mean 1.63 mg/L packed RBC) at the end of the treatment. From the pre- to post-treatment evaluation, the mean EP of the anemic subjects changed from (mean ± SD) 5.79 ± 0.44 to 1.63 ± 0.10 mg/L RBC (SD =). On average, this subgroup (64%) was also more likely to show a developmental lag.

EP responded slowly to iron (Dagg and Goldberg 1973 ), but the dosage of iron given to the Costa Rican children was sufficient to produce a larger fall in the EP than the one actually observed during the 13 wk of treatment. It remained to be determined whether the limited response of EP was related to the absence of a therapeutic response in the Bayley Scales of mental development. Later, Lozoff et al. (1996) conducted a similar study of anemic (Hb <100.5 g/L) toddlers (12–23 mo), extending the oral iron over 6 mo. Anemic (Hb <95 g/L) toddlers received 3 mg/kg iron administered orally twice per day for 6 mo. Three and 6 mo after baseline, the anemic subjects experienced an average hemoglobin change of 34.2 and 38.3 g/L, respectively; the EP changed from 5.96 to 1.17 mg/L RBC and then to 636 µg/L RBC. At these two last points, the mean EP of the subjects was within normal levels (cut-off point for 1–2 y: >1.42 mg/L RBC); however, the previously anemic subjects still showed a significant lag in mental development 6 mo after baseline. This 1996 experiment suggested that the limited response of EP observed in 1988 was not a reason for the failure of a therapeutic response in mental development in both Costa Rican studies.

In addition, in contrast to what was observed in Indonesia, there was no evidence that the oral iron preparation [6 mg/(kg·d)] given to the nonanemic control subjects in the Costa Rica studies limited weight gain. Because the adverse effects iron supplementation have generally been restricted to cases of iron overload (Bothwell et al. 1979 ), this interpopulation difference in growth response places the reliability of the Indonesian data in doubt. Recently, however, other investigators have reported limited weight gain of iron-sufficient subjects exposed to iron fortification in different populations (Dewey et al. 2000 ).

The data reported from Indonesia and Costa Rica do not account for the differences in the responses to similar treatments. In fact, similarities in the level of severity of the IDA, length of the iron intervention, age and social and economic background of the subjects are reasons to support the expectation of similarity in the developmental results. One notable difference between studies was the design, i.e., the Costa Rica study followed a quasi-experimental design and the Indonesian study was a randomized trial. Such a difference lies at the very essence of internal validity (Cook and Campbell 1979 ), i.e., the threats to internal validity were higher in Costa Rica than in Indonesia. However, the two studies in Costa Rica and the trial in Indonesia carried the markers of sound, solid research. The differences in the findings between sites are likely to reflect true differential responsivity to the iron treatment rather than to differences in design. Briefly, timing of the deficiencies, dietary, health or even social psychological factors through unknown pathways could well account for the interpopulation differences in the findings on mental development and physical growth.

An example of an atypical response of hemoglobin to a long-term iron intervention was found in a recently completed community-based, randomized trial on the island of Pemba in Zanzibar (Stoltzfus et al., unpublished data, 2000). This study tested the effects of low dose daily iron supplementation (10 mg iron) and regular anthelmintic treatment (500 mg mebendazole) for 1 y on the iron status, anemia, growth, morbidity and development of children 6–71 mo at the start of the trial. Malaria (Plasmodium falciparum) is holoendemic with year-round transmission in Pemba. Hookworm (Ancylostoma duodenale and Necatur americanus) as well as Ascaris lumbricoides, Trichuris trichiura, and Schistoma hematobium are highly endemic. About 35% of the children were underweight and 5.5% were wasted. The degree of infection, anemia and general undernutrition observed in this sample was probably the most severe among all studies reported in this field of research. [For other studies of children with moderate and severe anemia, see Boivin and Giordani (1993) and Heywood et al. (1989) .]

The mean hemoglobin at baseline was the same (mean ± SD; 86 ± 15 g/L) for those who received placebo and iron. Their respective hematological changes during the 12 mo of treatment were 11 and 13 g/L, respectively. The intergroup differences in intragroup changes were significant (P = 0. 003) for EP; the mean values of those who received iron changed from a geometric mean of 162 to 72 µmol/mol, and the controls changed from 153 to 86 µmol/mol.

Iron treatment was associated with higher post-treatment motor scores in children with low baseline hemoglobin, but this treatment effect disappeared at higher baseline hemoglobin concentration. The benefit from iron treatment was apparent at baseline hemoglobin concentration <90 g/L and became significant at a concentration of <80 g/L. The effect of the iron treatment was also significant in language development across children with the wide range of hemoglobin concentration.

The restoration of iron in the organism leading to improved motor and language development without reversing the anemia discriminated the effects of iron deficiency from those of anemia. However, the correlation between the baseline hemoglobin and the salutary developmental changes observed prevented the conclusion that the hematological derangement did not play a causal role in the developmental delays observed before treatment.

Another example of the way health and nutritional factors moderate the responses to a micronutrient intervention was found in Pangalengan, Indonesia (Harahap et al. 2000 ) in the research project already cited (Pollitt et al. 2000 ). In this study focusing on micronutrients, we tested the differences between the effects of 6 mo of a micronutrient supplement that included 12 mg of iron given daily to 12-mo-old anemic infants and the effects of skimmed milk given to nonanemic infants of the same age. The subjects in the two groups were matched by sex and age and were selected from the same daycare centers. Every 2 mo, motor activity and motor and mental development were tested, and behavior was observed under natural conditions. Six months after baseline, all indicators of iron status improved significantly in the anemic children, and the initial differences in iron status were no longer significant.

In contrast to what most other studies have reported (Grantham-McGregor and Ani 2001 ), in this study, there were no differences between anemic and nonanemic children in the Bayley Scales of Mental Development, the Fagan Response to Novelty Test and an object concept test. However, motor development and activity discriminated between such groups. The anemic children who received iron improved their motor performance and became more active than the nonanemic children. Because of the data from other studies, it had been reasonable before launching the study to postulate that the iron therapy would benefit both motor and mental development. The original postulate was therefore wrong, i.e., the motor but not the mental area showed the expected response. This particular pattern in the findings was probably due to the poor overall nutritional status of the children under study.

In summary, IDA and its treatment cause several different responses in individuals and populations. With some exceptions, we are still far removed from a discovery of the moderating factors and the respective mechanisms. Within the field of developmental psychobiology, we must look more carefully at the physiological changes that accompany the different levels of depletion of iron and propose specific hypotheses concerning how such changes can alter particular psychobiological domains, including cognition. A challenge ahead is to account for the course and processes of the relationship between the causal risk factor and the target outcome and for the role played by other risk factors (Ciccetti and Cohen 1997, Kazdin et al. 1997 , Kopp 1994 ). A natural extension of this idea is that there are differences among regions in the nature, breadth and intensity of the developmental effects produced by IDA.

	Time of measurement

TOP ABSTRACT INTRODUCTION Nature of psychobiological... Differential effects and... Time of measurement Psychometrics of development Summary and conclusions REFERENCES

 
As noted, several sets of factors that affect psychobiological development and the canalization of the organism [as defined by Gottlieb (1991) ] must moderate the functional consequences of early IDA. The developmental nature of the moderating roles of such factors justifies the study of functional deficits over time to draw the contours in the course of the deficits and to identify and understand the pathways leading to such effects. This section discusses the idea that single snapshots cannot capture the course of development within domains and that the timing at which the adverse developmental changes are manifested varies among domains.

A recent study (Williams et al. 1999 ) on the effects of an iron-supplemented formula on the performance on the Griffith developmental scale is illustrative. Infants 6–8 mo of age were randomly assigned to receive either an iron-supplemented formula or unmodified cow’s milk. At 18 mo, the infants who received the iron-supplemented formula were changed to cow’s milk, and the two groups continued taking the cow’s milk for another 6 mo. Both groups took the Griffith test at 18 and 24 mo of age and both showed an intragroup negative function between performance and age. As the children grew older, their performance declined compared with a criterion. At 18 mo of age, there were no intergroup differences in intragroup changes; however, by 24 mo, the decline in the group that received the cow’s milk exclusively was significantly (P < 0.05) larger than that observed in the group that received the iron-supplemented formula. One explanation for the age difference in the responses to the treatment is that the level of neurobiological maturation and the organization of the cerebral functions that were tested at each age moderated the effects of the treatment. Thus, the benefits were delayed for abilities that were not present in normal children before a certain age. Another explanation is that the test was more sensitive to the intervention in older than younger children.

A second example on the importance of the time and timing of testing is found in Semarang, Indonesia (Soemantri et al. 1985 ). Anemic children whose school achievement was significantly behind that of nonanemic children improved their achievement scores after an iron intervention; no such improvement was seen among anemic children who received placebo. However, the iron had only limited effects because by the end of the study, the anemic children who were treated were still significantly behind the nonanemic children. Quite likely, the treatment did not compensate for the learning time lost. A second set of post-treatment measurements may have shown a larger salutary effect of iron.

Different psychobiological domains respond at different times to the same treatment. For example, the energy and micronutrient supplement given in Pangalengan, Indonesia (Jahari et al. 2000 , Pollitt et al. 2000 ) increased the motor activity and the verbal output of the children 2 mo after baseline. This same group of children engaged in more social play 4 mo after baseline but it took 6 mo before the effects of the energy supplement were detected on the Bayley Scales.

Lozoff and collaborators (2000) recently stated that there might be different mechanisms that vary according to domain. In some cases, the effect might be the direct expression of an insult in a particular brain region, whereas long-lasting motor effects could result from lags in myelination. In other cases, there would be indirect effects involving both neural and behavioral processes. As an example of a specific deficit, they speculated that the poorer visual-spatial working memory and a delay in developing the ability to attend selectively and inhibit attention to the irrelevant might be related to a deficiency in iron in the prefrontal-striatal and hippocampal systems.

An 8-y follow-up study of a 3-mo trial of energy supplementation showed the differential responsivity of the organism as a function of the time of evaluation and timing of intervention (Pollitt et al. 1997 ). Eight years after the trial ended, there was a specific effect on the time-to-scan working memory in the children in the experimental group who were 18 mo or younger. However, there were no effects on several other tests that assessed higher cognitive outcomes, such as quantitative thinking or language development. It thus seems reasonable to propose that IDA during infancy may well have long-lasting adverse effects on basic and elementary cognitive functions that are rooted in early brain structures. On the other hand, it is dubious that early IDA will affect higher cortical functions in adolescents that depend on several distinct but interrelated cognitive processes and cumulative learning.

	Psychometrics of development

TOP ABSTRACT INTRODUCTION Nature of psychobiological... Differential effects and... Time of measurement Psychometrics of development Summary and conclusions REFERENCES

 
A psychometric approach has dominated the selection of tools for the assessment of development in the studies under consideration for infants, toddlers, preschool and school-age children. My comments in this brief review are limited to the youngest groups. I will first address issues related to the predictive and construct validity of the mental development scales used and then address the yield of snapshots compared with repeated measures of development.

Mental development scales.

Three of the four prospective trials published on iron fortification in infants and toddlers used the Bayley Scales of Mental Development, which is generally considered a test of infant intelligence (Lozoff 1997 , Moffat et al. 1994 , Morley et al. 1999 , Williams et al. 1999 ). This scale yields an aggregate score (mental development index) that allegedly depicts several mental abilities and processes present in early childhood. The fourth study (Williams et al. 1999 ) showed differences in the changes of the general quotient of the Griffith Scales from 7 to 24 mo in favor of the subjects who took the fortified formula. An analysis by subscales showed that the only significant difference was found in the social and personal subscale. The four trials started the nutrition intervention a few months after birth and included a development assessment at least once before age 12 mo.

Probable reasons for the unexpected results of the Bayley Scales are as follows: 1) iron has no effect on mental development; 2) iron affects mental development among toddlers with IDA but in these three studies, the prevalence of this micronutrient deficiency in the respective communities was too low for the effect of the intervention to surface; 3) times of measurement were inappropriate (see previous section); and 4) tests were not sensitive enough to the intervention. In my view all of these reasons, except the first, lie behind such results.

Psychometric data currently available strongly suggest that despite its suggestive name, the mental developmental scales for infants and toddlers in general and the Bayley Scales in particular are not tests of intelligence as generally assumed (Neisser et al. 1996 ). Specifically, the Bayley Scales, particularly during approximately the first 18 mo of life, does not measure the same intellectual abilities and functions that are assessed by IQ tests during the preschool or school period. Later abilities to understand complex or abstract ideas, be socially adaptive, learn from experience or do well in school are not rooted in the kinds of abilities or behaviors tested with infant development scales (Goodman 1990 , McCall and Mash 1994 , Meisels and Atkins-Burnett 1999 ). For this reason, the mental development scales administered during the first 18 mo have little if any power to predict performance on intelligence tests during the preschool and school years. This limitation, however, is not restricted to healthy, well-nourished populations (McCall and Mash 1994 ). It is also true for rural populations in low income countries in which the prevalence of growth retardation and micronutrient deficiencies is high (Pollitt and Triana 1999 ). For example, in rural West Java, the power of the Bayley Scales administered some time between 6 and 30 mo of age to predict a verbal intelligence scale administered during the preschool (36–84 mo) or during the school period was zero. Similarly, the scale had no power to predict performance in an arithmetic test administered in school.

The psychometric problems go beyond a lack of predictive validity. Mental development before age 18 mo also lacks stability, as indicated by repeated measures of the same subjects with the same test during the first 18 mo of life. For example, in the United States, the correlation between the scores obtained at 4–6 mo and at 13–18 mo of age is 0.39; the correlations between scores at 7–12 mo and 13–18 mo is 0.46 (McCall and Mash 1994 ). In rural Guatemala, the test-retest correlations between an infant mental development scale administered at 6 and 15 mo were 0.08 for boys and 0.01 for girls (Lasky et al. 1981 ). In Indonesia the correlations between assessments at 12 and at 14, 16 and 18 mo of age were 0.53, 0.39 and 0.22, respectively (Pollitt and Triana 1999 ). The weak stability of the developmental scales during the first months of life tells us that its construct validity (Wainwright and Ward 1997 ) is also weak. Briefly, the little convergence there is between testing and retesting the same subject during the early months makes the findings of intergroup differences uninterpretable because we do not know what construct we are actually assessing. From these arguments, it is reasonable to claim that the findings of the three prospective studies (Lozoff 1997 , Moffat et al. 1994 , Morley et al. 1999 ) on the mental development scale were deceptive.

McCall and Mash (1994) state that standardized assessments of developmental status administered during the first 2 y of life have such short-term stability that any differences between groups observed at one age would not be expected to persist even across a few months during infancy. They further state that the prediction of later IQ is so modest that it has little scientific or practical significance. Similarly, Escalona and Moriarty noted: "When applied at age levels below 18 mo, the term ‘intelligence test’ is misleading ...the true relationship between that which is measured by infant tests and that which we later call intelligence remains largely unknown" (Meisels and Atkins-Burnett 1999 ).

Repeated measurement of development.

Pre- and post-treatment snapshots from developmental assessments are neither reliable nor valid measures for detecting intergroup differences of intragroup changes in infants and toddlers that are due to nutritional factors. What other criteria besides test reliability and validity should be met in order to assess whether nutritional interventions shape a developmental trajectory? One criterion, as noted, is the time of post-treatment measurement to allow for the manifestation of a treatment effect. Several repeated measures of a basic and universal process of cognition that changes over time without losing some of its primary purposes, e.g., language, are an attractive option.

	Summary and conclusions

TOP ABSTRACT INTRODUCTION Nature of psychobiological... Differential effects and... Time of measurement Psychometrics of development Summary and conclusions REFERENCES

 
The putative effects of IDA on infants and toddlers cannot be measured with the psychometric tools that have generally been used. Mental development scales have neither the stability nor the predictive power to attribute much scientific merit to the data generated by these scales during the first 18–24 mo of life. Therefore, there is no psychometric justification to draw the conclusion that IDA diminishes the mental capacity of this age group. This does not negate, however, that IDA produces cerebral changes in early life that compromise particular cognitive functions in early and late development.

The nature of the cerebral changes resulting from experimentally induced IDA in laboratory animals suggests that these changes contribute to the developmental lags observed in children with IDA through direct and indirect pathways. However, the nature of psychobiological development suggests that, with exceptions, there is not likely to be a one-to-one relationship between the changes observed in particular neurotransmitter systems (e.g., dopamine or -aminobutyric acid) or myelin formation in early life and alterations of higher cognitive functions and school achievement. Relationships among psychobiological domains internal to the organism and transactions between the organism and the environment are also likely to make critical contributions to cognitive development. They will either moderate the effects from cerebral changes or operate as pathways through which development is affected.

Clinical trials are considered to be the optimal design to assess the effects of IDA on development. However, the approach adopted has not been optimal. The spectrums of physiological changes that occur at different stages of iron depletion have generally been disregarded, and there have been no attempts to assess the moderator role other risk factors may play. Repeated measures over time within the same domain are considered of particular importance to draw the course of development.

	ACKNOWLEDGMENTS

 
The author gratefully acknowledges the critical comments of Peggy Gionnoni, Linda L. Kelly, Michael Lawler, Deanna Olney, Erin Reid and Janet Talley.

	FOOTNOTES

 
¹ Presented at the Belmont Meeting on Iron Deficiency Anemia: Reexamining the Nature and Magnitude of the Public Health Problem, held May 21–24, 2000 in Belmont, MD. The proceedings of this conference are published as a supplement to The Journal of Nutrition. Supplement guest editors were John Beard, The Pennsylvania State University, University Park, PA and Rebecca Stoltzfus, Johns Hopkins School of Public Health, Baltimore, MD.

² The Nestle Foundation funded all of the studies in West Java, Indonesia, cited in this study.

³ Abbreviations: EP, erythrocyte protoporphyrin; Hb, hemoglobin; IDA, iron-deficiency anemia.

	REFERENCES

TOP ABSTRACT INTRODUCTION Nature of psychobiological... Differential effects and... Time of measurement Psychometrics of development Summary and conclusions REFERENCES

 

1. Bentley M. E., Caulfield L. E., Ram M., Santizo M. C., Hurtado E., Rivera J. A., Ruel M. T., Brown K. H. Zinc supplementation affects the activity patterns of rural Guatemalan infants. J. Nutr. 1997;127:1333-1338[Abstract/Free Full Text]

2. Birengen Z., Emde R. N., Campos J. J., Applebaum H. Affective reorganization in the infant, the mother and the dyad—the role of upright locomotion and its timing. Child Dev 1995;66:499-514[Medline]

3. Black M. M. Zinc deficiency and child development. Am. J. Clin. Nutr. 1998;68:464-469

4. Boivin M. J., Giordani B. Improvements in cognitive performance for school children in Zaire, Africa, following an iron supplement and treatment for intestinal parasites. J. Pediatr. Psychol. 1993;18:249-264[Abstract/Free Full Text]

5. Bothwell T. H., Charlton R. W., Cook J. D., Finch C. A. Iron Metabolism in Man 1979 Blackwell Scientific Oxford, UK.

6. Brooks G. Hierro, metabolismo muscular y actividad fisica. O’Donnell A. M. Viteri F. E. Carmuega E. eds. Deficiencia de Hierro 1994:67-86 Centro de Estudios Sobre Nutricion Infantil Buenos Aires, Argentina.

7. Bruner A., Joffe A., Duggan A., Casella J., Brandt J. Randomized study of cognitive effects of iron supplementation in non-anemic iron-deficient adolescent girls. Lancet 1996;348:992-996[Medline]

8. Campos J. J., Kermoian R., Zumbahlen M. R. Socioemotional transformation in the family system following infant crawling onset. Eisenberg N. Fabes R. eds. Emotions and Its Regulation in Early Development. New Directions in Child Development 1992:110 Josey Bass San Francisco, CA.

9. Celsing F., Bloomstrand E., Werner B., Pihlstedt P., Ekblom B. Effects of iron deficiency on endurance and muscle enzyme activity in man. Med. Sci. Sports Exerc. 1986;18:156-161[Medline]

10. Cicchetti D., Cohen D. J. Perspectives on developmental psychopathology. Cicchetti D. Cohen D. eds. Developmental Psychopathology 1995;I:3-22 Wiley New York, NY.

11. Connor J. R., Menzies S. L. Relationship of iron to oligodendrocytes and myelination. Glia 1996;17:83-93[Medline]

12. Cook T. D., Campbell D. T. Quasi-Experimentation 1979 Design and Analysis Issues for Field Settings. Houghton Mifflin Boston, MA.

13. Dagg J. H., Goldberg A. Detection and treatment of iron deficiency. Callender S. eds. Clinics in Hematology 1973;2 W.B. Saunders & Co London, UK.

14. Dewey K. G., Domellof M., Cohen R. J., Landa Rivera L., Hernell O., Lonnerdal B. Effects of iron supplementation on growth and morbidity of breastfed infants: a randomized trial in Sweden and Honduras. FASEB J 2000;14:A509(abs.)

15. Goodman J. F. Infant intelligence: do we, can we, should we assess it?. Reynolds C. R. Kamphaus R. W. eds. Handbook of Psychological and Educational Assessment of Children 1990:183-208 Guilford New York, NY.

16. Gottlieb G. The experiential canalization of behavioral development. Theory Dev. Psychol. 1991;27:4-13

17. Gottlieb G., Wahlsten D., Lickliter R. The significance of biology for human development: a developmental psychobiological system. Damon W. Lerner R. M. eds. Handbook of Child Psychology, Vol. 1. Theoretical Models Of Human Development 1998:233-274 Wiley New York.

18. Grantham-McGregor S., Ani C. A review of studies on the effect of iron deficiency on cognitive development in children. J. Nutr. 2001;131:649S-668S[Abstract/Free Full Text]

19. Harahap H., Jahari A. B., Husaini M. A., Saco-Pollitt C., Pollitt E. Effects of an energy and micronutrient supplement on iron deficiency anemia, physical activity, and motor and mental development in undernourished children in Indonesia. Eur. J. Clin. Nutr. 2000;54:S114-S119

20. 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;50:630S-637S[Abstract]

21. Hunt J. R., Zito C. A., Erjavee J., Johnson L. K. Severe or marginal iron deficiency affects spontaneous physical activity in rats. Am. J. Clin. Nutr. 1994;59:413-418[Abstract/Free Full Text]

22. Idrjadinata P., Pollitt E. Reversal of developmental delays in infants treated with iron. Lancet 1993;341:1-4[Medline]

23. Idrjadinata P., Watkins W., Pollitt E. Adverse effect of iron supplementation on weight gain of iron-replete young children. Lancet 1994;343:1252-1254[Medline]

24. Jahari A. B., Saco-Pollitt C., Husaini M. A., Pollitt E. Effects of energy and micronutrient supplementation on motor development and motor activity in undernourished children in Indonesia. Eur. J. Clin. Nutr. 2000;54:S60-S68

25. Johnson M. H. Developmental Cognitive Neuroscience 1997 Blackwell Great Britain.

26. Johnson M. H. Functional brain development in infants: elements of an interactive specialization framework. Child Dev 2000;71:75-81[Medline]

27. Kazdin A. E., Kraemer H. C., Kessler R.F.C., Kupfer D. J., Offorfd D. R. Contribution of risk factor research to developmental psychopathology. Clin. Psychol. Rev. 1997;17:375-406[Medline]

28. Kopp C. K. Trends and directions in studies of developmental risk. Nelson C.A. eds. Threats to Optimal Development: Integrating Biological, Psychological, and Social Risk Factors 1994;27:1-34 The Minnesota Symposia on Child Psychology Hillsdale, N.J.

29. Lasky R. E., Klein R. E., Yarbrough C., Kallio K. D. The predictive validity of infant assessments in rural Guatemala. Child Dev 1981;52:847-856[Medline]

30. Lewis M. D. The promise of dynamic systems approaches for an integrated account of human development. Child Dev 2000;71:36-43[Medline]

31. Lozoff B. Does preventing iron deficiency anemia (IDA) improve developmental test scores?. Pediatr. Res. A 1997;39:136(abs.)

32. Lozoff B., Brittenham G. M., Wolf A. W., McClish D. K., Kuhnert P., Jimenez E., Jimenez R. L., Mora L.R.F., Gomez I., Krauskoph D. Iron deficiency anemia and iron therapy effects on infant developmental test performance. Pediatrics 1987;79:981-995[Abstract/Free Full Text]

33. Lozoff, B., Jimenez, E., Hagen, J., Mollen, E. & Wolf, A. W. (2000) Poorer behavioral and developmental outcome more than 10 years after treatment for iron deficiency in infancy. Pediatrics 105: e51 (electronic version: http://www.pediatrics.org/).

34. Lozoff B., Wolf A. W., Jimenez E. Iron deficiency anemic and infant development: effects of extended oral iron therapy. J. Pediatr. 1996;29:382-389

35. McCall R., Mash C. W. Infant cognition and its relation to mature intelligence. Ann. Child Dev. 1994;10:27-56

36. Meisels S. J., Atkins-Burnett S. Assessing intellectual and affective development before age three: a perspective on changing practices. Food Nutr. Bull. 1999;20:23-33

37. Moffatt M.E.K., Longstaffe S. W., Besant J., Dureski C. Prevention iron deficiency and psychomotor decline in high-risk infants through use of iron-fortified infant formula: a randomized clinical trial. J. Pediatr. 1994;125:527-534[Medline]

38. Morley R., Abbott R., Fairweather-Tait S., MacFadyen U., Stephenson T., Lucas A. Iron fortified formula from 9 to 18 months improves iron status but not development or growth: a randomized trial. Arch. Dis. Child. 1999;81:247-252[Abstract/Free Full Text]

39. Neisser U., Boodoo G., Bouchard T. J., Boykin A.W., Brody N., Ceci S. J., Halpern D. F., Loehlin J. C., Perloff R., Sternber R. S., Urbana S. Intelligence: known and unknowns. Am. Psychol. 1996;51:72-101

40. Nelson C., Erkison K., Pinero D. J., Beard J. L. In vivo dopamine metabolism in altered iron-deficient anemic rats. J. Nutr. 1997;127:2282-2288[Abstract/Free Full Text]

41. Oski F. A., Hong A. S., Heel B., Howanitz P. Effects of iron therapy on behavior performance in nonanemic, iron-deficient infants. Pediatrics 1983;71:877-880[Abstract/Free Full Text]

42. Pollitt E. A developmental view of the undernourished child: background and purpose of the study in Pangalengan, Indonesia. Eur. J. Clin. Nutr. 2000;54:S1-S9

43. Pollitt E., Chuang J. F., Jahari A. A developmental function of motor activity among nutritionally at risk children. Food Nutr. Bull. 1999;20:100-107

44. Pollitt E., Saco-Pollitt C., Jahari A., Husaini M. A., Huang J. Effects of an energy and micronutrient supplement on mental development and behavior under natural conditions in undernourished children in Indonesia. Eur. J. Clin. Nutr. 2000;54:S80-S90

45. Pollitt E., Triana N. Stability, predictive validity and sensitivity of mental and motor development scales and pre-school cognitive tests among low-income children in developing countries. Food Nutr. Bull. 1999;20:45-52

46. Pollitt E., Watkins W., Husaini M. Three-month nutritional supplementation in Indonesian infants and toddlers benefits memory function 8 y later. Am. J. Clin. Nutr. 1997;66:1357-1363[Abstract/Free Full Text]

47. Roncagliolo M., Garrido M., Walter T., Peirano P., Lozoff B. Evidence of altered central system nervous development in infants with iron deficiency anemia at 6 mo: delayed maturation of brainstem responses. Am. J. Clin. Nutr. 1998;68:683-690[Abstract]

48. Sameroff A. General systems theories and developmental psychopathology. Cicchetti D. Cohen D. J. eds. Developmental Psychopathology 1995;1:659-695 Theory and Methods Wiley, New York, NY.

49. Soemantri A. G., Pollitt E., Kim I. Iron deficiency anemia and educational achievement. Am. J. Clin. Nutr. 1985;42:1221-1228[Abstract/Free Full Text]

50. Thelen, E. & Smith, L. B. (1998) Dynamic system theories. In: Handbook of Child Psychology (Damon, W. & Lerner, R. M., eds.), Vol.1. Theoretical Models of Human Development, pp. 563–634. Wiley, New York, NY.

51. Wachs T. D. Necessary but not sufficient: the respective roles of single and multiple influences on individual development 2000 American Psychological Association Washington, DC.

52. Wainwright P. E., Ward G. R. Early nutrition and behavior: a conceptual framework for critical analysis of research. Dobbing J. eds. Developing Brain and Behaviour: The Role of Lipids in Infant Formula 1997:387-414 Academic Press New York, NY.

53. Walka H., Pollitt E. A preliminary test of a developmental model for the study of undernourished children in Indonesia. Eur. J. Clin. Nutr. 2000;54:S21-S27

54. Williams J., Wolff A., Daly A., MacDonald A., Aukett A., Booth I. W. Iron supplemented formula milk related to reduction in psychomotor decline in infants from inner city areas: randomized study. Br. Med. J. 1999;318:693-698[Abstract/Free Full Text]

55. Zhu Y. I., Haas J. D. Altered metabolic response of iron depleted nonanemic women during a 15-km time trial. J. Physiol. 1999;84:1768-1775

This article has been cited by other articles:

				J. C McCann and B. N Ames An overview of evidence for a causal relation between iron deficiency during development and deficits in cognitive or behavioral function Am. J. Clinical Nutrition, April 1, 2007; 85(4): 931 - 945. [Abstract] [Full Text] [PDF]

				P. E Wainwright and J. Colombo Nutrition and the development of cognitive functions: interpretation of behavioral studies in animals and human infants. Am. J. Clinical Nutrition, November 1, 2006; 84(5): 961 - 970. [Abstract] [Full Text] [PDF]

				E. Pollitt Are the psychological tests valid? Am. J. Clinical Nutrition, July 1, 2005; 82(1): 201 - 201. [Full Text] [PDF]

				C. M. Smuts, C. J. Lombard, A. J. S. Benade, M. A. Dhansay, J. Berger, L. T. Hop, G. Lopez de Romana, J. Untoro, E. Karyadi, J. Erhardt, et al. Efficacy of a Foodlet-Based Multiple Micronutrient Supplement for Preventing Growth Faltering, Anemia, and Micronutrient Deficiency of Infants: The Four Country IRIS Trial Pooled Data Analysis J. Nutr., March 1, 2005; 135(3): 631S - 638S. [Abstract] [Full Text] [PDF]

				A. B. Perez-Exposito, S. Villalpando, J. A. Rivera, I. J. Griffin, and S. A. Abrams Ferrous Sulfate Is More Bioavailable among Preschoolers than Other Forms of Iron in a Milk-Based Weaning Food Distributed by PROGRESA, a National Program in Mexico J. Nutr., January 1, 2005; 135(1): 64 - 69. [Abstract] [Full Text] [PDF]

				P. A. HOLDING and P. K. KITSAO-WEKULO DESCRIBING THE BURDEN OF MALARIA ON CHILD DEVELOPMENT: WHAT SHOULD WE BE MEASURING AND HOW SHOULD WE BE MEASURING IT? Am J Trop Med Hyg, August 1, 2004; 71(2_suppl): 71 - 79. [Abstract] [Full Text] [PDF]

				K. G. Nead, J. S. Halterman, J. M. Kaczorowski, P. Auinger, and M. Weitzman Overweight Children and Adolescents: A Risk Group for Iron Deficiency Pediatrics, July 1, 2004; 114(1): 104 - 108. [Abstract] [Full Text] [PDF]

				J. Liu, A. Raine, P. H. Venables, C. Dalais, and S. A. Mednick Malnutrition at Age 3 Years and Lower Cognitive Ability at Age 11 Years: Independence From Psychosocial Adversity Arch Pediatr Adolesc Med, June 1, 2003; 157(6): 593 - 600. [Abstract] [Full Text] [PDF]

				E. Pollitt, A. Jahari, M. Husaini, P. Kariger, and C. Saco-Pollitt Developmental Trajectories of Poorly Nourished Toddlers that Received a Micronutrient Supplement with and without Energy J. Nutr., September 1, 2002; 132(9): 2617 - 2625. [Abstract] [Full Text] [PDF]

				D. Totin, C. Ndugwa, F. Mmiro, R. T. Perry, J. B. Jackson, and R. D. Semba Iron Deficiency Anemia Is Highly Prevalent among Human Immunodeficiency Virus-Infected and Uninfected Infants in Uganda J. Nutr., March 1, 2002; 132(3): 423 - 429. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar