Keywords

Introduction

Iron deficiency is the most prevalent single nutrient deficiency worldwide primarily affecting infants, children, and women of reproductive age [1]. Although individuals of all ages and classes are susceptible to iron deficiency, those who are poor and less educated are most vulnerable. Iron deficiency has been identified as a biological risk factor for the failure of >250 million children in reaching their full developmental potential, resulting in reductions in economic productivity [2]. The association between iron status and brain functioning has been examined through investigations of central nervous system biochemical changes using animal studies and through the assessment of cognitive functioning and behavior in humans. There is clear evidence from animal studies that iron is critical for myelination, neuronal morphology, neuronal metabolic activity, and the synthesis of monoamines [3]. In humans, the evidence points to cognitive, motor, behavioral, and affective alterations in iron-deficient individuals [4]. In most age groups, iron repletion appears to ameliorate these alterations. The exception is infancy where the negative consequences of iron deficiency seem to persist despite iron repletion, perhaps indicating critical periods of development during which a deficiency in iron leads to irreversible effects.

A heterogeneous distribution of iron exists in the human brain with differential patterns in children versus adults [5]. The accumulation of iron in different brain regions is a function of the stage of brain development [6]. The highest concentrations of brain iron are found in the substantia nigra, deep cerebellar nuclei, the red nucleus, the nucleus accumbens, and portions of the hippocampus [7, 8]. Dopaminergic, serotonergic, and noradrenergic systems have been identified as sensitive to brain iron status [9].

This chapter summarizes cognitive and behavioral alterations resulting from iron deficiency and iron deficiency anemia in children and women of reproductive age. Multiple reviews have been published on these topics [10, 11], so here we focus on reviewing randomized controlled studies published in the past decade, examining both the short- and long-term cognitive and behavioral consequences of iron deficiency. Studies where the control group contained approximately the same levels of iron as the treatment group (for instance, control group given a multiple micronutrient (MMN) treatment or comparing oral iron to intravenous iron) were excluded from evaluation.

Effects of Iron Deficiency on Cognition and Behavior in Children

Most studies examining the association between iron status and cognitive/behavioral outcomes have been conducted in infants and young children. Over 40 years of accumulated evidence from infant studies reveals that iron deficiency anemia is associated with impaired performance on developmental tests as well as behavioral differences such that iron-deficient anemic infants are more wary, hesitant, and clingy when compared to their iron-sufficient counterparts [12]. During the preschool years, iron deficiency anemia has been associated with impairments in learning and language acquisition; motor development; and in school-aged children, iron deficiency anemia has been shown to be related to impaired academic performance (especially on verbal and math tests) and memory [13]. Many of the recent studies on the association between iron status and cognition/behavior have focused on the long-term consequences of early-life iron treatment. The studies that we include here were conducted in apparently healthy children; we have divided the studies by primary outcome of interest (cognitive or behavioral outcomes).

Effects of Iron Treatment on Child Cognitive Outcomes

Ten recent publications were identified in which children or adolescents received iron treatment or a true control to assess cognitive outcomes [14,15,16,17,18,19,20,21,22,23]. Sample sizes across all ten studies ranged from 140 to 1933 with supplementation periods ranging from 4.5 to 8.5 months. As far as vehicle of supplementation, four provided iron supplements (drops or tablets) [14, 16, 19, 21], three provided iron via infant formula [15, 22, 23], and three provided iron through either fortification [17, 18] or biofortification [20] of food. Five of the studies assessed outcomes immediately following the end of the treatment period [14, 17, 18, 20, 21] and the other five assessed outcomes several years later [15, 16, 19, 22, 23]. As far as data analysis, eight used the original randomized groups [14,15,16,17,18,19,20, 23] while two ran the analyses with children stratified by iron status in infancy [22] or by iron status in the fetal-neonatal and infancy periods [21]. All studies used multiple biomarkers specific to iron to determine iron status although the 3.5-year follow-up study conducted in Sweden [16] did not measure iron status. Since iron status at follow-up may have affected the interpretation of the findings, contextualization of those results are more difficult. Of note, the study in China included measures of serum ferritin concentrations but did not collect a measure of inflammation and, therefore, ferritin levels were not adjusted [21].

Table 23.1 summarizes the main outcome variables and findings from these studies. Full scale intelligence quotient (IQ), memory, and motor function were the main outcome domains assessed using both manual and computerized tasks. Of the eight studies which used the original randomized groups, three reported no significant differences between the intervention and the control group at endline or follow-up [16, 17, 19]. Two of these studies provided supplementation (iron drops) to infants in Sweden [16, 19] while the other used fortified biscuits, provided to Moroccan school-aged children [17]. All three of these studies used measures of cognition that require highly trained administrators and that are subjective in nature. In addition, these measures were developed and standardized in a Western context. As such, using Western-based standardized scores may not have been appropriate for the study conducted in Morocco. Three other studies which ran analyses using the original randomization groups reported improvements in cognition favorable to the iron treatment group, and there was indication that those with a lower iron status at baseline experienced the greatest gains in terms of cognitive scores [14, 18, 20]. It may be noteworthy that supplementation occurred during the school years in these studies and not during the infancy period, when supplementation was not related to improvements in cognition [16, 19]. The study conducted in India [20] used computerized tests which are less subjective and allow for a finer assessment of cognition (for instance, measuring time in milliseconds). Nevertheless, two studies [14, 18] used more subjective measures, similar to those used in the Sweden and Morocco studies described above and yet, reported significant differences after supplementation. Again, timing of supplementation may be at play, but it may also be important to consider the way in which the cognitive test scores were calculated. The study in Cambodia used raw scores for the three administered tests (Raven’s Colored Progressive Matrices and the Block Design and Picture Completion subtests from the Wechsler Intelligence Scale for Children III), citing that no standardized scores appropriate to that setting were available [18]. The study in South Africa used Western standardized scores for some of the outcomes (subscales of the Kaufman Assessment Battery for Children) and raw scores for others (where no standardized scores exist; Hopkins Verbal Learning Test) [14]. Interestingly, the differences between groups were only seen for tests where the authors used raw scores. Whether or not a lack of association when using Western-based standardized scores is an indication that applying these scores in these contexts is inappropriate is a question that remains to be answered. One final possibility should be mentioned here. The iron treatments used in the studies conducted in Cambodia [18] and India [20] provided higher levels of other micronutrients (such as vitamin B12 or zinc) compared to the placebo groups. The possibility that these nutrients positively influenced cognition cannot be ruled out. Of the remaining studies that conducted analyses using the original randomization groups, both reported worse cognitive outcomes for children who had been treated with higher iron formula vs. those treated with a low iron formula during infancy [15, 23]. These follow-up studies assessed children at 10 and 16 years of age and used cognitive assessments that require a highly trained administrator. Interestingly, when accounting for hemoglobin levels at enrollment, both studies reported higher scores (for spatial memory assessed with the Kaufman Assessment Battery for Children at 10 years of age and for visual motor integration assessed with the Beery-Buktenica Developmental Test of Visual-Motor Integration at both 10 and 16 years of age) after iron treatment for those children whose hemoglobin was low at baseline (6 months of age) but lower scores after iron treatment for those children whose hemoglobin was high at baseline.

Table 23.1 Studies examining the effects of iron treatment on cognitive domains in children and adolescents (2010–2020)
Full size table

The two studies that ran analyses based on iron status in infancy both reported lower scores (motor or language) in children who were iron deficient during the fetal/neonatal and/or infancy periods [21, 22]. One of these studies assessed outcomes immediately at the end of the treatment (9 months of age) [21] while the other assessed outcomes several years after the treatment ended (at 5.5 and 10 years of age) [22].

Overall, results from recent studies, which assessed the association between iron and cognition in children and adolescents, are mixed. In general, the studies point to an association between poor iron status and lower cognitive scores. However, the benefits of treating with iron on cognitive outcomes are not clearly established. Studies that provided iron treatment during infancy seem to indicate no benefit or even worse outcomes with higher doses of iron. Alternatively, studies that provided the iron treatment during the school-age years appear to show a benefit of the supplementation. Given the findings that early life iron deficiency may have irreversible effects, it appears that preventing iron deficiency in infancy should be a top priority. While a limited number of observational studies that assess the association between iron status in children and adolescents and neurophysiology exist, no such randomized controlled trials were found in our assessment of articles published in the past decade. Additional studies are needed to better understand the role of timing, duration, and severity of iron deficiency on cognitive outcomes and neurophysiology as well as the effect of timing, duration, dose, and vehicle of supplementation on these outcomes in children. Type of testing conducted and manner in which the tests are scored may also affect the findings.

Long-Term Effects of Iron Treatment in Early Life on Child Behavioral and Affective Outcomes

Over the past decade, seven publications were identified in which children had received iron treatment or a control during infancy and were then followed up at later ages during childhood to assess behavioral and affective (outward expression of an individual’s internal emotions) outcomes [16, 19, 24,25,26,27,28]. The seven included studies represent follow-up from three original studies [29,30,31] with follow-up sample sizes ranging from 161 to 1116 and supplementation periods of approximately 6 months for six of the studies [16, 18, 25,26,27,28] and 12–20 months for one [24]. As far as the vehicle for supplementation, Chun-Ming et al. provided a sachet of multiple micronutrients to be added to complementary foods (comparison group received a sachet of rice flour and vegetable oil), Berglund et al. provided iron drops (1 or 2 mg/kg body weight/day with comparison group receiving 0 mg/kg body weight/day), and Lozoff et al. provided a high or low-iron formula (12.7 mg/L and 2.3 mg/L, respectively; comparison group received a formula with no added iron). Chun-Ming and colleagues only measured hemoglobin and, as such, there is no indication of whether or not the anemia measured in their study was due to iron deficiency. At baseline [29], no child was excluded due to their hemoglobin concentration while at follow-up [24], the authors excluded individuals who were anemic. The studies conducted by Berglund et al. and Lozoff et al. used multiple biomarkers which are specific for iron status and, as such, were able to classify the children as iron-sufficient, iron-deficient, and iron-deficient anemic. Of importance, the 3.5-year follow-up [16] conducted by Berglund et al. did not assess iron status at follow-up; as such, the interpretation of the findings is less clear. The study conducted by Berglund et al. excluded anemic children at baseline [30] and the study conducted by Lozoff et al. [31] was a prevention trial, randomizing children who were iron sufficient or iron deficient but not anemic but supplementing all children who were anemic at baseline. Age of children at follow up ranged from 3.5 to 17 years. Of the seven follow-up studies assessing behavior, four analyzed the data by using the original randomized groups [16, 19, 25, 27]. The others ran the analyses with the children stratified by iron status in infancy (iron-sufficient, iron-deficient, or iron-deficient anemic [26, 28] or by whether or not the deficiency was corrected by the original treatment [24]) to assess the impact of early-life iron status on later behavioral outcomes.

Table 23.2 summarizes the main outcome variables and findings from these follow-up studies. Affect, behavior, and social difficulties were the three main outcome domains assessed through researcher observation, parental report, and/or child self-report, depending on the study. The studies that utilized the original randomization groups in their analyses reported a higher prevalence of behavioral problems (overall behavior as assessed with the Child Behavior Checklist (CBCL), externalizing problems (CBCL), and conduct disorder (CBCL)) in the control group (no additional iron provided), vs. the groups who received iron when behavior was assessed through parental report [16, 19, 27], and more positive affect in children who received iron vs. those who did not when assessed via observer rating [25]. These studies assessed the children years after the original supplementation (at 3.5, 7, 10, and 14 years of age). However, when behavior was assessed through child self-report (at ~14 years of age), higher scores on the ADHD symptoms subscale were found in the groups who received iron vs. the group who did not [27]. As only one study used child self-report to assess behavior, this finding will need to be replicated in future studies before it can be properly interpreted. All of the remaining analyses assessed the association of early-life iron status to later behavioral/affective outcomes (regardless of randomization group although they controlled for group) and reported that iron deficiency anemia in infancy was related to worse affect (less positive affect, more dull affect) and worse behavior (higher externalizing problems, excessive alcohol use, risky sexual behavior) even after controlling for possible confounders (such as SES, maternal education, maternal depressive symptoms, child sex, family stressors, the home environment) when compared to children who were iron sufficient in infancy [24, 26, 28]. One study compared children whose anemia was corrected in infancy to those who were never anemic in infancy and found no differences in behavior or affect at the 4-year follow-up [24]. Another study found that children who were iron deficient (but not anemic) in infancy had higher adolescent behavior problems compared to children who were iron sufficient in infancy [28].

Table 23.2 Long-term follow-up of studies examining the effect of iron treatment in infancy on child and adolescent behavior and affect (2010–2020)
Full size table

Together, these studies indicate negative consequences of iron deficiency and iron deficiency anemia in infancy on behavior and affect, years later. The differences reported among adolescents who were formerly iron deficient are especially troubling, given the serious potential consequences of the behaviors (excessive alcohol consumption and risky sexual behaviors). The finding that behavioral differences might persist in children who were iron deficient, but not anemic, in infancy is cause for concern as the prevalence of iron deficiency without anemia is high and iron deficiency, in the absence of anemia, typically goes undetected. Whether or not iron supplementation in early life can reverse these negative consequences is still in question since few randomized controlled trials exist and most of the long-term studies have assessed the outcomes based on early-life status as opposed to randomization group.

Effects of Iron Deficiency on Cognition and Behavior in Women of Reproductive Age

While decades of research indicate an association between iron status and cognitive/behavioral outcomes in infants and young children, few studies have been conducted in adults. This was due to the belief that the brain was resistant to changes in iron once the blood–brain barrier reached maturity [32]. However, animal studies revealed that the uptake of iron into the brain is dependent on iron status, as there is an increased rate with low iron status and a decreased rate with high iron status [33]. Furthermore, the uptake process is not reflective of overall blood–brain barrier permeability [34, 35]. Since this knowledge emerged, there has been an increased interest in understanding the association between iron status and cognition/behavior in adults, particularly in women of reproductive age, given their susceptibility to iron deficiency. Studies conducted prior to the past 10 years included both observational and intervention designs and most were conducted in developed countries. The observational studies suggested a relation between iron status and cognition such that higher iron levels are associated with better cognitive functioning, specifically, spatial ability, attention, memory, and executive functioning. These studies also report a relation between iron status and affect such that higher iron levels are associated with fewer depressive symptoms and a higher quality of life [4]. All of the intervention studies assessing cognition as the outcome, reported an improvement in cognitive functioning with iron supplementation. Likewise, in intervention studies assessing affect as the outcome, improvements were reported following iron supplementation with the greatest improvements found among women who had poor baseline iron status. However, few studies were randomized controlled trials and few studies specifically examined the effect of iron deficiency, in the absence of anemia, on cognition, behavior, or affect. Here, we focus on recent randomized controlled studies that examined the association between iron supplementation and cognition/behavior in women of reproductive age.

Effects of Iron Treatment on Cognitive and Neurophysiological Outcomes in Women of Reproductive Age

Four manuscripts, representing three randomized controlled trials, of the effects of iron treatment on cognition in adult women of reproductive age have been published over the past decade [36,37,38,39]. Three recruited university-attending women as the participants [36, 38, 39] and one recruited women who worked on a tea plantation [37]. The vehicle for supplementation was beef in one study (comparison group received non-beef foods), provided for 16 weeks [36], double-fortified salt (comparison group received single-fortified salt), provided for 10 months [37], and biofortified beans (comparison group received conventional beans), provided for 18 weeks [38, 39]. All studies included multiple iron status biomarkers which were assessed at baseline and endline and the study in Rwanda was restricted to women with a ferritin ≤20 ng/mL at baseline. The studies analyzed the data using an intent to treat approach and also included secondary data analysis approaches.

Table 23.3 summarizes the main outcome variables and findings from these randomized controlled trials. The main outcome domains were memory and attention and one study also used electroencephalogram (EEG) measurements to assess electrophysiology [39]. The study conducted in the United States did not find any difference in cognitive outcomes between the groups at endline [36]. Both groups improved their iron status over time but changes in cognitive outcomes did not differ by group. On the other hand, when classifying the women as those who had a positive change in ferritin vs. those who did not, the authors report greater improvements in all three cognitive domains tested (memory, attention, spatial planning) in the “ferritin responders” vs. the “ferritin non-responders.” In contrast, the studies conducted in India and Rwanda found significant differences between groups at endline on the cognitive domains tested (memory, attention, perception) with women in the treatment arms having greater improvements [37, 38]. Although the exact tests given differed in these studies, all studies utilized computerized cognitive tests. The study in Rwanda was limited to women who were iron deficient at baseline, the study in India included only the subgroup of women who had the lowest ferritin values from the larger parent trial, and the study conducted in the United States did not limit enrollment based on ferritin concentrations. Indeed, at baseline, the mean ferritin concentrations of those in the US study were nearly four times higher than the mean ferritin concentrations of those in the Rwanda study and almost 40% higher than those in the India study. It is possible that women with a lower iron status at baseline experience a greater benefit of increased iron status. The Rwanda study also included measures of electrophysiology and found greater improvements in EEG amplitude and spectral power in the group who consumed the biofortified beans vs. those who consumed the conventional beans [39]. An important contribution of this study is the finding that changes in brain activity (EEG) mediate the relation between changes in iron biomarkers and changes in cognition (memory and attention). Studies which use electrophysiological measurements and relate the findings to changes in cognition are especially helpful in terms of providing a link between the mechanistic studies conducted with animal models and the behavioral and affective outcomes that are typically measured in human studies.

Table 23.3 Randomized controlled trials assessing the effect of iron treatment on cognition and neurophysiology in women of reproductive age (2010–2020)
Full size table

Recent observational studies (not reviewed in-depth here) are supportive of an association between iron status and cognitive functioning in women of reproductive age, indicating that better executive functioning and attention scores are found in women who are iron sufficient vs. those who are deficient [40,41,42]. Two recent observational studies have also examined the association between iron status and neurobiology. One revealed differences in left EEG alpha activity in prefrontal regions between iron-deficient (non-anemic) and iron-sufficient women of reproductive age [43]. The other study found differences in brain connectivity (using functional magnetic resonance imaging) between women who had been iron-deficient anemic in infancy vs. controls (iron sufficient or mild iron deficiency in infancy) [44]. Specifically, formerly iron deficient anemic subjects had decreased connectivity from the posterior Default Mode Network (DMN) to the left posterior cingulate cortex (PCC) and increased connectivity from the anterior DMN to the right PCC. They also exhibited differences in the left medial frontal gyrus.

Although these observational studies are supportive of a relation between iron and cognition in women of reproductive age, the number of recent randomized controlled trials assessing the effects of iron treatment on cognition in this population is extremely limited. While the studies provide evidence that iron deficiency is related to cognitive alterations and changes in electrophysiology, more work is needed to fully understand these associations in this age group. Optimal cognitive functioning is necessary for performing day-to-day duties. Alterations in maternal cognitive functioning may have significant implications for maternal–child interactions with subsequent negative effects on child development, as women are often the primary caregiver for children. It is therefore crucial that these types of studies continue to be conducted in women of reproductive age.

Effects of Iron Treatment on Behavioral and Affective Outcomes in Women of Reproductive Age

Three recent randomized controlled trials were identified in which women of reproductive age received iron treatment or a control and behavioral/affective variables were assessed as the outcomes of interest [45,46,47]. Two of the studies were conducted in developed countries [45, 46] and the other was conducted in Iran [47], a semi-developed country. Sample sizes ranged from 70–198 with intervention periods ranging from 2 weeks (intravenous iron) to 12 weeks. As for the vehicle for supplementation, the study conducted in Switzerland utilized intravenous iron (comparison group received intravenous placebo) while the other two provided oral iron supplements as tablets (with comparison groups receiving oral placebo tablets). As far as iron status assessment and inclusion criteria, all of the studies included multiple iron status biomarkers which were assessed both at baseline and endline and all of the studies excluded women who were anemic at baseline. Additionally, the studies conducted in Switzerland and France included only women whose baseline ferritin levels were ≤ 50 ng/mL and the study conducted in Iran included only women with postpartum depression. All of the studies analyzed the data using an intent to treat approach and it is important to note that the study conducted in France was observer blinded while the other studies were double blinded.

Table 23.4 summarizes the main outcome variables and findings from the randomized controlled trials conducted over the past decade. The main outcome domains were affect (specifically, anxiety, depression, and quality of life) and fatigue. In both studies where fatigue was assessed, the authors found significantly larger decreases in fatigue in the women who received iron vs. those who received a placebo [45, 46]. For one of these studies, this association was only true for women whose ferritin concentrations were ≤ 15 ng/mL at baseline [45]. For the studies that assessed affect, one reported no significant differences between groups on measures of anxiety, depression, or quality of life [46]. The other study reported significant improvements in depression scores in women who were treated with iron compared to those treated with placebo (improvement rate of 42.8% vs. 20.0% for iron vs. placebo treated, respectively; p = 0.03) [47]. Several differences exist between these studies which may contribute to the discrepant finding: (1) different instruments were used to assessed depression, (2) participants in the Iran study were all 1-week postpartum while the participants in France were not in the postpartum period, and (3) the study conducted in Iran only enrolled women who had been diagnosed with postpartum depression.

Table 23.4 Randomized controlled trials assessing the effect of iron treatment on behavior and affect in women of reproductive age (2010–2020)
Full size table

Although the number of randomized controlled trials assessing iron status and behavior in women of reproductive age is limited, the studies provide evidence that iron status is related to behavior and that iron repletion may ameliorate the negative findings. Of importance, all of these studies excluded anemic women and, therefore, the findings indicate alterations in behavior in those who are iron deficient but not anemic. In other words, mild iron deficiency has behavioral and affective consequences in women of reproductive age. As mentioned above in our review of the findings in children, this is especially concerning, given the lack of identification of iron deficiency in the absence of anemia, in most settings.

Conclusion

The brief review provided in this chapter reveals an association between iron deficiency (with and without anemia) and alterations in cognition and behavior for both children and women of reproductive age. Findings of this association in the absence of anemia are particularly troubling, given the magnitude of iron deficiency without anemia and the fact that it goes largely undetected. The findings of long-term cognitive and behavioral consequences of iron deficiency in infancy despite iron repletion are also of particular concern. Finally, the fact that higher iron doses used for repletion may be related to worse outcomes when supplementation occurs in infancy needs to be further investigated. These findings indicate that preventing iron deficiency in infancy should be a top priority. The magnitude of cognitive and affective changes reported with iron deficiency vary by study but, in general, indicate levels that are likely to impact daily activities. Although studies of the functional consequences of iron deficiency continue, there is a clear need for well-designed randomized controlled studies in order to better understand the effects of timing, duration, and severity of the deficiency as well as the optimal treatment timing, duration, and dose. Studies that link changes in neurophysiology to changes in these cognitive and behavioral outcomes are especially needed.