REVIEW ARTICLE

published: 17 August 2012

doi: 10.3389/fnhum.2012.00238

The impact of poverty on the development of brain

networks
Sebastián J. Lipina1,2 * and Michael I. Posner3
1
Unidad de Neurobiología Aplicada, Centro de Educación Médica e Investigaciones Clínicas Norberto Quirno (CEMIC), Consejo Nacional de Investigaciones
Científicas y Técnicas CONICET, Buenos Aires, Argentina
2
Centro de Investigaciones Psicopedagógicas Aplicadas CIPA, Universidad Nacional de San Martin (UNSAM), San Martín, Argentina
3
Department of Psychology, University of Oregon, Eugene, OR, USA

Edited by: Although the study of brain development in non-human animals is an old one, recent imag-
Rajeev D. S. Raizada, Cornell ing methods have allowed non-invasive studies of the gray and white matter of the human
University, USA
brain over the lifespan. Classic animal studies show clearly that impoverished environments
Reviewed by:
Leslie J. Carver, University of
reduce cortical gray matter in relation to complex environments and cognitive and imaging
California at San Diego, USA studies in humans suggest which networks may be most influenced by poverty. Studies
Rajeev D. S. Raizada, Cornell have been clear in showing the plasticity of many brain systems, but whether sensitivity to
University, USA learning differs over the lifespan and for which networks is still unclear. A major task for cur-
Bernd Weber, Rheinische
Friedrich-Wilhelms-Universität,
rent research is a successful integration of these methods to understand how development
Germany and learning shape the neural networks underlying achievements in literacy, numeracy, and
*Correspondence: attention. This paper seeks to foster further integration by reviewing the current state of
Sebastián J. Lipina, Unidad de knowledge relating brain changes to behavior and indicating possible future directions.
Neurobiología Aplicada, Centro de
Educación Médica e Investigaciones Keywords: childhood poverty, brain networks, plasticity, attention, literacy, numeracy
Clínicas, Consejo Nacional de
Investigaciones Científicas y Técnicas,
Av. Galván 4102, C1431FWO Buenos
Aires, Argentina.
e-mail: [email protected]

INTRODUCTION at rest (Fair et al., 2009, 2011; Gao et al., 2009). The results
The study of the influence of material and social deprivation on the to date have shown evidence of sparse connectivity between
central nervous system (CNS) has been an issue of interest in the brain structures during infancy and a strong increase in long
agenda of neuroscience since the first half of the twentieth century. range connectivity at 2 years (Gao et al., 2009) and later (Fair
Early neuroscience studies analyzed how the exposure to complex, et al., 2007, 2009). In studies of neonates, the parietal areas,
standard, or deprived environments modifies the brain of exper- prominent in the orienting of attention network, show strong
imental animals (Hebb, 1949; Mohammed et al., 2002; Grossman connectivity to lateral and medial frontal areas. The mid frontal
et al., 2003; Markham and Greenough, 2004; Sale et al., 2008). cortical area and the anterior cingulate cortex (ACC), which
The development of neuroimaging has provided non-invasive have been implicated in self-regulation (Posner and Rothbart,
methods for examining changes in gray and white matter of the 2007b) and down-regulation of the amygdala during the pro-
human brain (Posner and Raichle, 1994; Raichle, 2009) and may cessing of threatening social cues (McEwen and Gianaros, 2010),
now serve to better integrate animal work with the century long show a marked increased connections to frontal areas and to
efforts to specify lifespan changes in critical areas of development. lateral parietal areas during childhood. In older children (Fair
Applying neuroscience methods to issues of child poverty has et al., 2009), these tendencies continue and the ACC becomes
emerged from these developmental efforts (Hackman and Farah, increasingly differentiated from the orienting network as one
2009; Lipina and Colombo, 2009; Hackman et al., 2010). Since the approaches adulthood. One view of control systems was sug-
mid 1990s different researchers have used behavioral and imaging gested recently by Fair et al. (2011). They argued, “data suggested
to compare the cognitive and academic performance of children that there might be at least two control networks functioning in
with disparate socioeconomic status (SES; e.g., Korenman et al., parallel. Based on the differences in their functional connectiv-
1995; Guo, 1998; Lipina et al., 2004, 2005; Mezzacappa, 2004; ity and activation profiles we suggested that each network likely
Noble et al., 2005; Farah et al., 2006). Advances in neuroimag- exerts distinct types of control on differing temporal scales. The
ing have made it possible to incorporate neural network analysis fronto-parietal network was proposed to be important for rapidly
in studies of the influence of poverty (e.g., Noble et al., 2007; adaptive control and to work on a shorter timescale. The cingulo-
D’Angiulli et al., 2008; Raizada et al., 2008; Stevens et al., 2008, opercular network was thought to be important for more stable
2009; Kishiyama et al., 2009). set-maintenance, and to operate on a longer timescale. Since this
A major achievement in the use of fMRI to study human initial work there have now been several reports supporting this
development has arisen through the study of brain connectivity framework.”

Frontiers in Human Neuroscience www.frontiersin.org August 2012 | Volume 6 | Article 238 | 1

Lipina and Posner Poverty and brain networks

In the present paper, we specifically focus on the issue of neuro- receptors, the Alzheimer amyloid precursor protein, immediate
science approaches to childhood poverty using the recent data on early genes, serotonin receptors, α-amino-3-hydroxy-5-methyl-
brain networks including the resting state data. The resting state 4-isoxazolepropionic acid (AMPA) receptor binding, and neuro-
studies provide a unique perspective on the developing human genesis (Mohammed et al., 2002). Middle-aged rats that had at
brain, but their weakness is the limited evidence on how these least 1 year in complex environmental conditions showed higher
changes influence infant and child behavior. In the section of this levels of nerve growth factor (NGF) in the entorhinal cortex, com-
paper on attention and self-regulation, we try to establish such pared to isolated animals (Mohammed et al., 2002). Furthermore,
links, trying to highlight what should be taken in consideration several studies suggest an involvement of these neurotrophins in
regarding plasticity sensitivity to learning in the first stages of the synaptic plasticity of the hippocampus and other brain areas
development. The aim of our effort is to review current progress (McAllister et al., 1995). Several neurotrophic factors, such as
and identify target areas, which could help form a research agenda the NGF, the brain-derived neurotrophic factor (BDNF), and the
for the coming years. To carry this out we first examine the general neurotrophin-3 (NT-3), are abundantly expressed in pyramidal
plasticity of the nervous system from the cognitive neuroscience cells and dentate granule cells, a fact suggesting that neurotrophins
viewpoint. Then we consider in turn the neural networks underly- are involved in mediating changes in the dendritic morphology
ing the regulation of stress, attention and self-regulation, literacy, following exposure to complex environments. In this regard, the
and numeracy. These topics were chosen because of their impor- role of the hippocampus in learning and memory has been the
tance in normal development and according to one recent review focus of increased research interest involving the generation of
on the mechanisms of executive function and language as the new neurons. Using the proliferation marker bromodeoxyuri-
most influenced by SES (Hackman and Farah, 2009; Lipina and dine (BrdU), different studies showed that enrichment increases
Colombo, 2009; Hackman et al., 2010; Raizada and Kishiyama, neurogenesis in the dentate gyrus of adult mice and rats. This
2010). We consider the integration of animal and human studies, effect has been associated with learning (Shors et al., 2001), while
imaging of both gray and white matter, and the role of genes and stress experience has been related to the reduction of neurogenesis
specific experience in developing brain networks. We concentrate (Gould et al., 1997).
on linking neuroscience methods with behavioral and psycholog- The cerebellum has also been found to display plastic proper-
ical approaches to development. In the next to last section we ties in response to environmental influences. Specifically, learning
review a number of other current approaches to these same issues. of a complex motor skill leads to an increase in the synaptic
Since there is much we still do not know, in the final section numbers of the cerebellar cortex, and to changes in the mor-
we propose priorities for a research agenda to allow for further phology of cerebellar Purkinje cells in rodents and monkeys
progress. (Kleim et al., 2003).
Behavioral changes following exposure to enriched environ-
BRAIN PLASTICITY mental conditions are not limited to the improvement of cog-
In recent years, neuroscience has begun to take seriously that expe- nitive performance. One robust observation has been based on
rience shapes the brain in important ways. Plasticity has been used the effects of isolation. Rodents after being reared in isolated
as a blanket term to cover these changes. One area of plasticity environments interacted less with objects in a free exploration
related to the possible consequences of poverty is that experi- situation, while displaying at the same time an increased loco-
mental studies on rodents and non-human primates exposed to motor activity coupled to a reduced habituation. These findings
motor, sensory, and social stimulation, show structural and func- would suggest that, depending on the complexity of environmen-
tional changes in neuronal and non-neuronal components, in tal enrichment or deprivation, brain circuits would be affected
comparison with those subjects exposed to deprived environments with more, or less, specificity involving different cortical areas
(see a comprehensive approach to brain plasticity in Rosenzweig, and subcortical structures. Although the experimental condi-
2003; Markham and Greenough, 2004; Sale et al., 2008; Lipina tions of enrichment/deprivation cannot be identified directly with
and Colombo, 2009). These changes include synaptic number and human poverty they may illuminate some aspects of child poverty
morphology, dendritic arborization, cell morphology, the num- (Lipina et al., 2011).
ber of astrocytes and glial-synaptic contacts, myelination, glial cell The findings of effects from enriched environments make use-
morphology; brain vasculature; brain cortex weight and thickness, ful interdisciplinary efforts and in particular to the translational
rate of hippocampal neurogenesis, availability and metabolism of animal and human research. For instance, in the context of analy-
both neurotrophic factors and neurotransmitters in different brain sis of the neural reactivity to stress the study of failures to habituate
areas, and neurotrophic and neurotransmitter gene expression to repetition of the same stressor or to terminate adaptive
as well. autonomic and neuroendocrine response, are based upon the
Complex environments induce neuroanatomical and bio- experimental animal–human translational efforts (McEwen and
chemical changes in several brain areas of young and adult Gianaros, 2010).
rodents, including frontal, parietal, and entorhinal cortices, Neural plasticity in humans may also lead to structural adapta-
hippocampus, and cerebellum (Rosenzweig and Bennet, 1996; tion in cortical gray matter, in response to environmental demands
Mohammed et al., 2002; Grossman et al., 2003). A considerable (Bavelier and Neville, 2002). At the level of imaging studies, there is
body of literature links the hippocampus to various plasticity fac- evidence that the brain may adapt dynamically to reflect environ-
tors, and to learning and memory mechanisms, such as gene mental cognitive demands. Neuroimaging studies show evidence
expression and protein levels of neurotrophins, glucocorticoid of structural changes in specific areas after training in difficult

Frontiers in Human Neuroscience www.frontiersin.org August 2012 | Volume 6 | Article 238 | 2

Lipina and Posner Poverty and brain networks

motor tasks. For example, studies of professional musicians (Gaser subject to a number of limitations. For instance, researchers
and Schlaug, 2003; Imfeld et al., 2009) show increased size of rele- can make only indirect inferences about brain function from
vant motor areas, and selective increases in gray matter volume in behavioral tests. In addition, many of these tests are multi-
posterior hippocampus. In addition, concomitant spatial memory factorial and performance could be disrupted for reasons other
performances have been shown in licensed London taxi drivers than those resulting from a specific dysfunction. Moreover, corre-
(Woollett and Maguire, 2011). lations among these tests are low, which means that two tasks can
Current studies in the developmental neuroscience field con- engage the same system in different ways. So a deeper examina-
tinue to advance in the understanding of the plastic mechanisms tion of the impact of SES and poverty on the relationship between
through which experience and environmental influences interact cognitive processes and brain function is needed. To this end, neu-
with genes, more specifically with the DNA biochemical markers roimaging techniques can be applied to analyze the neural level of
and histone proteins that regulate gene activity. Specifically, post- analysis.
translational modifications of histones and DNA methylation are
the most frequently analyzed mechanisms, involved in changes BRAIN PLASTICITY AND STRESS
in gene activity with environmental factors (Zhang and Meaney, The stress system has been used as a model system for an exami-
2010; Roth and Sweatt, 2011). Learning and memory processes also nation of adverse early experience on a brain network (see Lupien
evoke alteration of epigenetic markers in the adult CNS, as shown et al., 2009 for a review). The basic brain network involved in the
by animal models. For instance, Miller and Sweatt (2007) have stress response is the hypothalamus–pituitary–adrenal (HPA) axis.
used the contextual-fear conditioning paradigm to analyze epige- It is vertically organized and involves the autonomic and the CNSs
netic modulation of hippocampal genes. They found that during and hormonal as well as neural function.
a period of fear memory formation, adult rats have a demethy- Stress-related studies have been carried out in a variety of
lation and transcriptional activation of the memory-enhancing non-human animals as well as in humans (Lupien et al., 2009).
gene reelin, and an increase in methylation and transcriptional These studies rely heavily on animal models to examine the effects
silencing of the memory suppressor gene protein phosphatase 1. of early environmental effects such as maternal care, caregiver
These multiple changes in neural structure are correlated with maltreatment, mother–infant separation, and prenatal stress. The
functional changes in motor, cognitive, and emotional systems evidence suggests that early stress can produce lasting epigenetic
and behaviors (Mohammed et al., 2002; Grossman et al., 2003; modifications, stable changes in the CNS gene activity, as well
Markham and Greenough, 2004). Thus, development and learning as on behavior. For example, adult offspring raised by moth-
alter the local neural environment and that lead to the observed ers providing high level of pup licking and grooming showed
changes in brain and behavior (Galván, 2010). molecular changes in hippocampal glucocorticoid receptors, tran-
For humans poverty is associated with a restricted environ- scription of neural growth factor, corticotrophin releasing factor
ment and thus it would be expected that SES would be correlated expressions, and glucocorticoid feedback sensitivity (Zhang and
with differences in brain and behavior. Researchers have reported Meaney, 2010; Roth and Sweatt, 2011). Human studies by Roth
that the modulation of performance by SES is not the same in all et al. (2009) suggest that caregiver maltreatment can result in mod-
areas of behavior nor uniform at all ages (e.g., Lipina et al., 2004; ified methylation of the BDNF in the prefrontal cortex, and a DNA
Noble et al., 2005; Farah et al., 2006). Actually, SES is a multifaceted hypermethylation, which paralleled a lasting deficit in expression
construct hard to capture by single or multifactor indexes, and of the gene. These effects could be treated in part by a DNA
each component (e.g., income, education, and occupation) repre- methylation inhibitor administered to adults. Similar findings
sents resources that might benefit development in different ways in animal models resulted from maternal separation (Zhang and
(Duncan and Magnuson, 2012). Conceptually, this implies that Meaney, 2010).
poverty does not necessarily generate homogeneous and contin- The epigenetic analysis of the early experiences on brain devel-
uous changes in neurocognitive processing. These findings move opment in humans is in its first stages as many of the issues in the
away from the notion of low-SES performance as a deficit. Instead study of childhood poverty and brain development. For instance,
we need to develop measures for different brain networks; ages recently McGowan et al. (2009) examined the gene expression and
and SES/poverty conditions to contribute understand how poverty DNA methylation of the human glucocorticoid receptor (Nr3c1)
and SES shape brain networks development. For instance, Evans gene in hippocampal samples from suicide victims with a his-
et al. (2005) have pointed to the role of chaotic unpredictable fam- tory of childhood maltreatment. They found decreases levels of
ily environments – a specific aspect of SES, which is usually not mRNA hippocampal glucocorticoid receptor gene, correlated with
analyzed – as a threat to emotion regulation. With the complex- increases of cytosine methylation of the Nr3c1 promoter, which
ities of SES in mind developmental cognitive neuroscience may suggests that human caregiver experiences may program genes
serve as fertile ground for new ideas about the role of SES in through epigenetic modifications. Bueller et al. (2006) have found
development. evidences that carriers of the methionine allele of the Val66Met
In summary, the findings from behavioral studies indicate BDNF polymorphism express lower gray matter volume in the
that SES disparities and poverty can adversely affect cognitive hippocampus and prefrontal cortex compared with carriers of the
processes, such as language, executive function, attention, and valine–valine allele, suggesting a pathway of modulation of BDNF
memory. In addition, these findings suggest that specific brain secretion and intracellular functioning. In another study, Ober-
regions are associated with these cognitive functions. However, lander et al. (2008) found that infants of mothers with high levels
these tests are still behavioral in nature and as such, they are of depression and anxiety during the third trimester of pregnancy

Frontiers in Human Neuroscience www.frontiersin.org August 2012 | Volume 6 | Article 238 | 3

Lipina and Posner Poverty and brain networks

had increased methylation of the Nr3c1 gene promoter in cord to the ability to regulate other brain networks, thus exercising
blood cells. executive control over behavior (Fair et al., 2007, 2009, 2011; Gao
These studies support the hypothesis that the epigenome of et al., 2009; Posner et al., 2012). This control depends critically
prenatal developing infants is sensitive to the experiences of upon factors in the social environment such as parenting. Better
their mothers. Of course, many methodological issues should be understanding of the mechanisms by which control develops and
explored in future studies, such as whether peripheral measures of is exercised can provide guidance to parents and to society. Crit-
DNA methylation accurately reflect CNS methylation (Roth and ical to this is an understanding of the mechanisms, which switch
Sweatt, 2011). control from the caregiver during infancy to later self-regulation
In summary, we stress two important points from the work by the child.
on stress. First it shows the importance of a distributed systems One function that has been traced to the frontal midline (ACC)
of brain areas in the control of stress, involving prefrontal cortex, is monitoring and correction of errors. The ability to notice per-
hippocampus, amygdala, and the HPA axis (Lupien et al., 2009) ceptual errors occurs as early as 7 months (Wynn, 1992; Berger
which operate as a non-linear, interactive network in which multi- et al., 2006) and activates the anterior cingulate in infants just as it
ple mediators regulate each other, and promote adaptive activities does in adults. However, infants of this age are not able to use the
and coping with aversive situations and discrete stimuli, such as information to control their behavior. We had children play a sim-
those usually present in low-SES and poverty contexts (e.g., crowd- ple Simon game that asked them to execute a response command
ing, hunger, threats to mental and physical health). These areas and given by one puppet while inhibiting commands given by a second
the white matter connection between them are important contrib- puppet (Jones et al., 2003). Children of 36–38 months showed no
utors to the influence of stress on the person. Experimental animal ability to inhibit their response and no slowing following an error,
models of the hippocampus have revealed a mechanism by which but at 39–41 months children showed both an ability to inhibit
chronic stress leads to remodeling of hippocampus circuitry. These and slowing of reaction time following an error. These results sug-
changes consist in shortening of dendrites, loss of spine synapses, gest that between 38 and 39 months, performance changes based
and suppression of neurogenesis in the dentate gyrus. One of the upon detecting an error response. These dramatic changes in error
effects of these processes is impairing hippocampal involvement in correction relate to the changing connectivity of the brain during
episodic, declarative, contextual, and spatial memory, what in turn early child development, as was indicated by studies with neonates
leads to alter the ability to process information in new situations and older children in which parietal areas related to orienting of
and to make decisions about how to cope with new challenges attention show strong connectivity to lateral and medial frontal
(McEwen and Gianaros, 2010). areas, and the ACC shows increased connections to frontal areas
Second, is the notion that epigenetic changes underlie the long- and to lateral parietal areas (Fair et al., 2009), suggesting tenden-
term impact of early experiences, and that epigenetic alterations cies of increasing differentiation in the networks involving parietal
are potentially reversible or modifiable through pharmacological and frontal areas related to attentional processing.
and behavioral interventions. This means that the understanding However, changes in brain connectivity are not finished at age 3.
of the role of genes and of the epigenome in behavioral modifica- The Attention Network Test (ANT) has been used to examine the
tions driven by early experiences could contribute to the field of efficiency of the three brain networks (Fan et al., 2002) in older
childhood poverty and brain development. However, the presence children and adults. The task requires the person to press one key
of genetic variation in humans suggests that similar childhood if the central arrow points to the left and another if it points to
experiences could produce different outcomes depending upon the right. Conflict is introduced by having flankers surrounding
the exact version of the gene present in the individual. A child’s the target point in either point the same (congruent) or opposite
reaction to stress is an important factor in success in school and (incongruent) direction as the target. Cues presented prior to the
our understanding of the stress reaction may also guide us in ana- target provide information on where or when the target will occur.
lyzing other brain systems more directly involved in schooling as Reaction times for the separate conditions are subtracted, provid-
we try to do in the remaining sections of this paper. ing three measures that represent the efficiency of the individual
in alerting, orienting, and executive networks.
ATTENTION We have examined the ANT in children from 6 to 10 years of
Attention is a key factor in school readiness and success (Pos- age using a version specifically adapted to them. The results for
ner and Rothbart, 2007a). An understanding of the underlying children of this age are similar to those found for adults using
brain networks involved in attention has been a major contri- the same child version of the task. The child reaction times are
bution from research using neuroimaging (Posner and Petersen, much longer, but they show similar independence between the
1990; Posner and Rothbart, 2007b; Petersen and Posner, 2012). three networks. Children have larger scores than adults for alerting
Attention networks are involved in obtaining and maintaining the up to age 10 and for conflict up to age 7, suggesting that young
alert state (alerting network), orienting to sensory stimuli (orient- children have trouble in resolving conflict and even older children
ing network), and resolving conflict between responses (executive have trouble in maintaining the alert state when not warned of the
network). The executive network is a key to the role of ability of target (Rueda et al., 2004).
children and adults to regulate their thoughts and feelings. Adja-
cent areas of the ACC are involved in cognitive and emotional SES EFFECTS ON ATTENTION NETWORKS
control (Bush et al., 2000). Connectivity of these control systems In studies of preschoolers, first graders, and middle school chil-
develop over the early life of infants and young children, and lead dren, low-SES children had reduced performance on many tasks

Frontiers in Human Neuroscience www.frontiersin.org August 2012 | Volume 6 | Article 238 | 4

Lipina and Posner Poverty and brain networks

compared to middle SES children (Noble et al., 2005; Farah et al., A second type of training involves changes in brain state that
2006). These findings suggest that the prefrontal/executive sys- might influence some attentional networks as well as stress and
tem is one of the primary neurocognitive systems associated with immunoreactivity. For example, physical exercise has been shown
social inequalities in early experience. Similar results have been to have general advantages for improving cognitive function in
observed in studies using specific paradigms designed to measure adults and elderly (Hillman et al., 2008) and meditation has been
aspects of both executive function and attention. For example, shown to have a specific influence on executive attention and stress
Lipina et al. (2005) examined performance of low and middle SES in undergraduates (Tang et al., 2007). Meditation has been shown
infants using a task of a delayed-response paradigm, which incor- to be effective in children as young as 4 years of age (Tang et al.,
porates the evaluation of processes such as working memory and 2012), but the role of this method with low-SES children remains
inhibitory control. Findings showed that low-SES infants made to be demonstrated.
more errors associated with impairments of inhibitory control
and spatial working memory, and errors associated with attention LANGUAGE AND LITERACY
and search strategies. Language and literacy are important in school and are func-
The effects of socioeconomic disparities on attention have been tions found to be reduced in low-SES children. According to a
examined in several studies. For instance, Mezzacappa (2004) recent analysis (Hackman et al., 2010) one of the systems most at
used the ANT to investigate the effects of socioeconomic dispar- risk for low-SES children involves language. In recent years, the
ity on attentional processes in children of 6 years of age. This way in which experience shapes language development starting in
task can be used to assess alerting, orienting, and executive atten- infancy has been analyzed in detail. Below we review some of these
tion networks. Results showed that low-SES children had reduced findings.
measures of both speed and accuracy on measures of alerting In the 1970s, behavioral studies using habituation to a repeated
and executive attention, indicating that SES modulated response stimulus provided evidence that from birth infants are able to dis-
conflict and inhibit distracting information. criminate basic phonemes, the basic constituents of language, not
Another report on the difficulty low-SES children have in atten- only in their own language but also in other languages to which
tion comes from studies comparing the performance of high and they have never been exposed (Eimas et al., 1971; Streeter, 1976).
low-SES children attempting to listen to a story presented to one Studies using these behavioral methods together with electrical
ear while ignoring that on another ear. Performance is assessed by recording from the scalp have probed some of the early devel-
the amplitude of the P1 component of the auditory event-related opment of the phonemic structure underlying language. More
potential (ERP) known to be influenced by attention. Low-SES recently, infants have been exposed to language while resting in
children show little evidence of attention amplifying this EEG fMRI scanners to examine the brain mechanisms activated by
component while higher SES children of the same age show clear language (Dehaene-Lambertz et al., 2006).
evidence of P1 amplification for the attended ear. After a period of The infant language system appears to involve the same
training in classroom that provided practice in various attention left hemisphere language structures found in adults (Dehaene-
networks the SES children also showed the influence of attention Lambertz et al., 2006). In one study infants listened to sentences
on P1 (Stevens et al., 2009; Neville et al., 2011). presented aurally in their language. Brain areas in the superior
temporal lobe (Wernicke’s area) and in Broca’s area were active.
FORMS OF ATTENTION TRAINING When the same sentence was presented after a delay of 14 s Broca’s
There has been considerable evidence that various types of atten- area activity increased, as though this area was involved in the
tion training can be effective in children (Rueda et al., 2005, 2012; memory trace.
Diamond et al., 2007; Tang and Posner, 2009; Diamond and Lee, It has long been supposed that the early acquisition of lan-
2011; Neville et al., 2011; Klingberg, 2012; Segretin et al., 2012). guage might involve very different mechanisms than are active in
Most of the evidence involves practice with attention related tasks adults (Vicari et al., 2000). Left hemisphere lesions in infancy do
either using computerized tasks or classroom curricula. Although not produce a permanent loss of language function as they can
it is likely that many of these methods target executive attention in adults. Nonetheless, the new fMRI data suggests the left hemi-
network, some may train primarily orienting to sensory stim- sphere speech areas are involved in receptive language even at 3
uli. These methods have been shown to be effective in normal months of age and even though brain damage at this early age
preschool children and in children with disorders such as ADHD may allow the same functions to develop in the right hemisphere
or those with low-SES. It is not possible to say which methods (Dehaene-Lambertz et al., 2006).
are most effective. The computerized methods allow rather com-
pete specification of the training, while the classroom methods PHONEMES
provide more practical means of training. Most of the studies It has been possible to study changes in phonemic discrimination
have not examined how long the training is effective, but one due to exposure to the native language at least by 10 months of
study suggested little loss after 2 months (Rueda et al., 2012). In age (Kuhl, 1994; Saffran, 2002). Infants appear to acquire a sharp-
general, many studies of early preschool education have shown ened representation of the native phonemic distinctions (Kuhl
that the advantages for specific skills taught in school disap- et al., 2006). During this same period, they also lose the ability
pear in a few years, but there is some evidence that general to distinguish representations not made in their own language
benefits of training of executive skills last for many years (Moffitt (Werker et al., 1981). At least a part of the loss occurs when the
et al., 2011). non-native language requires a distinction that is within a single

Frontiers in Human Neuroscience www.frontiersin.org August 2012 | Volume 6 | Article 238 | 5

Lipina and Posner Poverty and brain networks

phonemic category in the native language. An example is the READING

ra–la distinction important in English is lost because it is within Reading is a high-level skill and in alphabetic languages such as
a single category in Japanese (McClelland et al., 2002). It is as English, it has properties related to the phonemic structure of
thought Japanese no longer hear this distinction and even when language. There have been many studies of adult reading and
exposed to English they may not improve in distinguishing ra much more is known than can be reviewed here (see Posner and
from la. In the McClelland et al.’s (2002) study an adaptive train- Rothbart, 2007a; Schlaggar and McCandliss, 2007 for reviews).
ing regime starting with initially easy stimuli was contrasted with Adult studies of reading have revealed a complex neural network
a fixed training regime using difficult stimuli, with some subjects involved in the translation of words into meaning. Two important
receiving feedback on the correctness of their responses and oth- nodes of this network are the visual word form area, of the left
ers receiving no feedback in both conditions. After three and six fusiform gyrus and an area of the left temporal–parietal junction
sessions of training, subjects received tests assessing identification for translating visual letters into sounds. The activation of the
and discrimination of /r/–/l/ stimuli as well as generalization. In left temporal–parietal junction is modulated by SES, among other
all cases except fixed training without feedback, subjects showed factors (Monzalvo et al., 2012).
clear evidence of learning, and several indicators suggested that Languages like English that are highly irregular in visual to
training affects speech perception, rather than simply auditory sound mappings are heavily dependent upon brain areas that
processes. Thus, training by several methods (McClelland et al., translate visual words to sound. Children who have difficulty in
2002; Iverson et al., 2005) seems to improve this form of phoneme learning to read show little activation in these phonological areas
discrimination even in adults, although it is not known how (Shaywitz et al., 2007). Many of the studies of learning to read use
well this knowledge can be incorporated into normal daily life dyslexic children who have shown specific difficulties in learning to
communication. read, some of what has been learned probably will apply to low-SES
It might be useful to find a way that will preserve the distinc- children as well, since the child’s ability in phonemic awareness,
tions originally made for the non-native language during infancy. (e.g., rhyming of auditory words), is a good predictor of their
One study showed that 12 sessions of exposure to a mandarin being able to learn to read alphabetic languages such as English.
speaker during the first year of life help to preserve a mandarin Training studies designed to improve decoding have shown that
phoneme in children whose native language was English (Kuhl children with low reading skill can improve and when this occurs
et al., 2003). A similar amount of exposure to a computerized ver- they show enhanced activation in areas related to phonological
sion of the speaker was not effective, suggesting the importance translation (McCandliss et al., 2003a,b).
of social interaction in this early form of learning. More needs The visual word form area is involved in integrating or chunk-
to be learned about how and whether media presentation can be ing visual letters into words (McCandliss et al., 2003b). Although
effective in learning. there has been some dispute about its unique operations it appears
There are many reasons why it is useful to know more than to be a part of the visual system that becomes expert in deal-
one language. Much of the world population lives in places where ing with letters as reading skill develops in later childhood (Price
speakers of two or more language live in close proximity. In addi- and Devlin, 2004; Ben-Shachar et al., 2011). It is thought that
tion, there is some reason to believe that proficiency in a second without the functioning of this area reading cannot become flu-
language provides improved performance in the ability to exer- ent. For example, patients with a lesion that interrupted the flow
cise executive control over thoughts (Bialystok and Martin, 2004), of information from the right hemisphere to the visual word
which might be one form of attention training. form area used letter by letter reading when words were pre-
There is also some reason to believe that the process of phone- sented left of fixation (going to the right hemisphere), while they
mic discrimination being developed in infancy is important for read words normally when the word was projected to the left
later efficient use of spoken and written language (Molfese, 2000; hemisphere and thus reached the visual word form area (Cohen
Guttorm et al., 2005), which is critical in childhood poverty stud- et al., 2004). Children from 7 to 18 who were deficient in both
ies since SES modulates the early language environments (Hoff, decoding and comprehension skills failed to activate this area, but
2003). Electrical recording taken in infancy during the course of were able to do so after extensive phonological training (Shaywitz
phonemic distinctions (Molfese, 2000; Guttorm et al., 2005) have et al., 2007).
been useful in predicting later difficulties in language and reading. The time course of development of the visual word form area
There is a history of using ERPs to assess infant deafness early in in English is important for the development of fluent reading.
life and being able to do so reliably have been very useful in the Phonics training often leaves the child with improved decoding
development of sign language and other interventions to hasten skill, but with a lack of fluency. Evidence that the visual word form
the infant’s ability at communication. Perhaps a similar role will develops rather late and first only for words with which the child is
prove to be possible for ERPs in the development of methods to familiar (Posner and McCandliss, 1999), suggests the importance
insure a successful phonemic structure in the native language. of continuous practice in reading to develop fluency (Shaywitz
There have been efforts to develop appropriate intervention in et al., 2007). More research is needed on the best methods for
later childhood for difficulties in the use of language and read- developing fluency particularly in non-alphabetic languages.
ing such as the widely used FastForward programs (Temple et al.,
2003). Although there are disputes about exactly why and for what LOW-SES CHILDREN LEARNING TO READ
populations this method works it remains important to develop A study by Noble et al. (2007), hypothesized that SES systemati-
remedies for language difficulties based upon research. cally influenced the relationship between phonological awareness

Frontiers in Human Neuroscience www.frontiersin.org August 2012 | Volume 6 | Article 238 | 6

Lipina and Posner Poverty and brain networks

skills and brain activity in areas involved in reading. Phonological in adults suggesting that the number line can be used by this age.
awareness, as we discussed in the Section “Phonemes” has been a There has been some dispute concerning the developmental course
key predictor for success in learning to read. To test this hypothesis, of the number line as some studies have suggested frontal struc-
researchers examined fMRI responses during a pseudoword read- tures (Ansari et al., 2005), rather than parietal structures mediate
ing task in first- to third-graders from diverse SES backgrounds. this decision (Cantlon et al., 2006).
Results showed a significant phonological awareness–SES interac- Apparently, there are linguistic and cultural differences in the
tion in the left fusiform visual word form area, indicating that at use of Arabic digits in the performance of calculations that could
similar low phonological awareness levels, children from higher have important consequences for the acquisition of language by
SES were more likely to evidence increased responses in the left children in different parts of the world. Using Arabic digits com-
fusiform cortical gyrus, while children from lower SES did not. monly employed by many cultures, the ability to make simple
In another recent study of healthy 5-year-old children perform- numerical comparisons were compared for Chinese and English
ing an auditory rhyme-judgment task, Raizada et al. (2008) found native speakers. Despite the identical input and tasks used, quite
a more direct relation: the higher the socioeconomic status, the different networks of brain areas were used by the two groups.
greater the degree of hemispheric specialization in Broca’s area, as English native speakers used the network of parietal and frontal
measured by the left-minus-right fMRI activation during rhyming areas discussed above. However, Chinese natives relied on premo-
tasks. This suggests that the maturation of Broca’s area in children tor areas not found active for English speakers (Tang et al., 2006).
may be governed by the complexity of the linguistic environments These fMRI findings indicate the different neural systems may
in which they grow up. be involved in dealing with very simple numerical tasks. We do
not know the reasons for these differences, but if they arise from
SUMMARY training, it may provide a way to improve the understanding of
Language development begins very early in infancy. Exposure to quantity, which has shown to be deficient in low-SES preschool
language shapes the phonemic structure in a way, which could children (Griffin et al., 1995).
influence later acquisition of literacy. Imaging studies have traced As the tasks were increased in difficulty by requiring addi-
the brain structure and connectivity involved in learning language tion as well as comparison the English natives speakers used
and acquiring literacy. Interventions for improving brain areas in language areas, as had been reported previously for exact calcula-
children with low learning skill have been developed and proven tion (Dehaene et al., 1999). In addition, English speakers activated
useful both for decoding of letters into their auditory form and limbic areas related to anxiety. However, Chinese native speakers
for chunking letters into a visual unit. It seems likely that children did not show activation of language areas, nor of areas related to
in poverty face difficulty in all of these operations. Their exposure anxiety and negative affect. Whether the differences between Chi-
to reduced input of language during infancy may cause problems nese speaking and English speaking children lies in genes, early
with perceiving phonemes, which in turn predicts performance in experience, educational method of some other difference is yet to
acquiring reading. Future studies may lead to further development be determined and is of great importance because of the strong
of interventions designed to improve these skills in all children. advantage which various Asian groups have shown in elementary
arithmetic tests. The imaging results show that the brain networks
NUMERACY used by Chinese children differ from English speakers suggesting
The human infant like other animals seems to have an inborn that more than mere effort is involved, but has not yet provided a
skill to recognize quantity. At least by a few months of age the clear reason for the difference.
infant seems to be able to discriminate changes when presented
with a small number of events presented. Wynn (1992) showed INFLUENCE OF SES
that infants of 7–9 months looked longer when simple addition There is wide agreement that learning of arithmetic operations
problems (presented as puppets) were in error than when they depends on the early skill in the ability of children to understand
were correct. Berger et al. (2006) compared this ability in 7- to quantity (Siegler, 2009). Significant differences in the numerical
9-month-old children and adults, using high-density electrical proficiency of preschoolers and kindergartners from different SES
recording from the scalp. They found the same electrodes over backgrounds have been described on a wide range of tasks such as
frontal midline areas discriminated between errors and correct in reciting the digits, counting sets of objects, counting up or down
both infants and adults. The adult brain made the discrimination from a given number other than 1, recognizing written numerals,
by about 250 ms and the infant brain was only delayed by about adding and subtracting, comparing numerical magnitudes, and
50 ms. The authors showed that error detection was signaled by an the ability to describe thinking and explain ideas in the context of
increase in theta rhythm in both groups. The electrodes in ques- mathematical problem solving (Crane, 1996; Pappas et al., 2003;
tion had been related to activity in the dorsal ACC in previous Ramani and Siegler, 2008).
studies (Dehaene et al., 1994). Studies using a program called Rightstart (Griffin et al., 1995)
The overall network of brain activity in processing number indicated that children from low-SES homes were at high risk for
has been studied in children and adults by high-density electrical failure in elementary school arithmetic, but training in numer-
recording in a task which required the person to indicate by press- ical quantity before the start of school could greatly reduce
ing keys whether a digit was above or below 5 (Dehaene, 1996; this deficit. Manual and computerized exercises based on this
Temple and Posner, 1998). Children as young as 5 years of age concept have been developed for young children. For instance,
showed similar brain mechanisms underlie the decision as found Ramani and Siegler (2008) have tested the prediction that playing

Frontiers in Human Neuroscience www.frontiersin.org August 2012 | Volume 6 | Article 238 | 7

Lipina and Posner Poverty and brain networks

linear number board games should enhance children’s numeri- to multiple risk factors (e.g., low housing and neighborhood qual-
cal knowledge by applying an intervention in which low-income ity, pollutants, toxins, crowding, and noise) that vary with SES.
preschoolers played a game for 1 h. Results showed increased pro- Walker et al. (2011) have reviewed the inequality between groups
ficiency on several numerical tasks (i.e., magnitude comparison, in developing countries that originate in early adverse experi-
number line estimation, counting and identification). They have ences. They describe the impact of risk factors (e.g., low cognitive
also verified that these gains remained 9 weeks after playing and stimulation, stunting, iodine and iron deficiencies, intrauterine
that home experience playing number board games were asso- growth restriction, malaria, lead exposure, HIV infection, mater-
ciated with numerical knowledge, suggesting that playing these nal depression, institutionalization, exposure to social violence)
games with children from low-SES or poor homes would increase with the aim of providing priorities for intervention programs and
their numerical knowledge. In addition, Wilson et al. (2009) tested policies. Mathews and Gallo (2011) attempt to revise psychological
the effectiveness of an adaptive game designed to improve number theories of SES and physical health, by reviewing psychobiologi-
sense in a sample of low-SES kindergartners. They have found that cal (i.e., biomarkers, neurotransmitters) and psychosocial factors
children improved their numerical competences in comparison of (i.e., stress and distress).
digits and words suggesting a change only in one aspect of the Using an interdisciplinary perspective that integrates eco-
number sense competences (i.e., access). nomics, neuroscience, and developmental psychology, Heckman
(2006, 2008) considers the rates of return to human capital invest-
SUMMARY ment for low-SES children. In his framework, skill formation
The study of attention, literacy and numeracy all point to common follows a hierarchical order in which later attainments build on
roots of school success in the experiences of infancy (Blair and earlier foundations, the author argues that developmental disad-
Razza, 2007). Of course as has always been thought the preschool vantage arises more from lack of early family stimulation than
period is important for preparing the child for a successful school from the lack of financial resources, so late remediation strategies
experience. We now have many more studies of how these advances are not effective.
arise in the shaping of brain networks. Explicit or implicit training All these approaches point to the following factors as medi-
in attention at preschool level may foster the learning of wide ators between SES/poverty and socioemotional and cognitive
variety of skills acquired in school including literacy and numeracy development: nutrition, access to health care, housing, stimu-
(Posner and Rothbart, 2007a). Brain-oriented research points to lating cognitive materials and experiences, parent expectations
both the specific experiences needed and methods to assay whether and styles, teacher attitudes and expectations, allostatic load (see
they have been achieved, as for example the adaptation of the below), and health-relevant behaviors.
approaches designed by Ramani and Siegler (2008) and Wilson Recent reviews have discussed SES related to neurocognitive
et al. (2009) for low-SES kindergartners, to the classroom context. differences. For example, Hackman and Farah (2009) and Lipina
and Colombo (2009) have reviewed studies in which behavioral,
OTHER CURRENT APPROACHES electrophysiological, and neuroimaging methods have been used
Several publications reviewing the role of SES and poverty in to characterize SES disparities in neurocognitive functions. Lan-
physical, cognitive, and socioemotional development have been guage and cognitive control showed the most sensitivity to SES.
published (Hackman and Farah, 2009; Lipina and Colombo, 2009; Hackman et al. (2010) examined pre- and post-natal levels of
Evans and Kim, 2010; Raizada and Kishiyama, 2010). For example, stress, the role of parental care in the development of hippocam-
from a developmental cognitive perspective, Bradley and Cor- pal structure and function. The epigenetics of regulation of the
wyn (2002) provide an overview of the association between SES HPA axis and the capacity of home environment to stimulate cog-
and children’s well-being in cognitive, socioemotional, and health nition are candidate mechanisms by which SES influences brain
development domains. In their approach, the cognitive domain development.
was considered in terms of school achievement, language profi- McEwen and Gianaros (2010) focused their review on the links
ciency and IQ; and the socioemotional domain was approached between stress-related processes in the social environment and
in terms of symptoms of psychiatric disturbance (e.g., internal- the brain (mainly in adults). The authors have illustrated the joint
izing/externalizing behaviors, use of substances) and maladaptive roles of amygdala, hippocampus, and prefrontal cortex as the brain
social functioning. systems mediating allostatic processes.
From a combined developmental and sociological perspective, Finally, Raizada and Kishiyama (2010) have focused their
Conger and Donnellan (2007) have addressed the relationship review on the open research opportunities in the area, and the
between SES and health disparities, over the life span. They have importance of integrating the neuroimaging dimension to behav-
analyzed three general theoretical approaches aimed at providing ioral approaches in the study of how SES disparities influences
explanations for SES and development associations: the social cau- cognitive and socioemotional development, and intervention
sation, social selection, and interactionist theories. Duncan and efforts as well.
Magnuson (2012), based on recent evidence involving associations
between manipulation of family income and children cognitive FUTURE DIRECTIONS
functions (Duncan et al., 2011), have noted the need to increase Biological and psychosocial risk factors associated with low-SES
the sophistication in the measurement and modeling of SES. and poverty conditions are related with inequalities in child
From a health perspective, Evans and Kim (2010) have cognitive and socioemotional development that poses a threat
approached the analysis of SES gradients in terms of exposure to educational attainment and adult productivity worldwide

Frontiers in Human Neuroscience www.frontiersin.org August 2012 | Volume 6 | Article 238 | 8

Lipina and Posner Poverty and brain networks

(Heckman, 2006; Walker et al., 2011; Marmot et al., 2012). Low- brain maturation. Finally, as Crone and Ridderinkhof (2011) have
SES and poverty can have profound effects on the brain and body, recently addressed “little headway has been made toward under-
and thus influence both mental and physical health. standing how brain growth maps onto mental growth during child
Policies and interventions can affect neuroplasticity. Emerg- development.” In their review, these authors have aim at bridg-
ing translational animal and human research link poverty to ing and integrating recent human brain maturation findings with
neurobiological pathways through changes in gray and white mat- the conceptual thinking of theorist in the behavioral tradition of
ter (McEwen and Gianaros, 2010). We are at the very start of studying cognitive development. In such a context, developmental
developing interventions that may aid in improving this situa- research in the area of self-regulation could serve as a reference
tion. In the fields of attention, literacy and numeracy we have point for understanding the relation between brain and mental
reviewed interventions using classroom and individual computer development.
exercises that have proven useful in some low-SES and poor It is our hope that this paper may help to enhance our cur-
populations. rent knowledge. Research could encourage both parents and
Much remains to be done to establish the efficacy and improve those responsible for public education to put more emphasis on
these interventions. Neuroscience studies and intervention need to preschool and early elementary education and to foster their task
consider the complexity of SES as the social sciences have proposed of ensuring the educational future of the world’s children.
(Evans and Kim, 2010; Mathews and Gallo, 2011).
Another issue that neuroscience should take in consideration ACKNOWLEDGMENTS
is a comprehensive approach to development in terms of theories This manuscript is being prepared for a special Issue of Frontiers
from other disciplines. Recently, Rao et al. (2010) have made such on “The social emotional developmental and cognitive neu-
an effort by using a longitudinal data set including ecologically roscience of socioeconomic gradients: Laboratory, population,
valid in-home measures of early experience during childhood, and cross-cultural and community developmental approaches.” It is
structural brain imaging during adolescence, and have found that supported in part by Ministerio de Ciencia y Tecnología (MIN-
parental nurturance at age 4 predicted the volume of the left hip- CyT) and Consejo Nacional de Investigaciones Científicas y Técni-
pocampus in adolescence. In addition, the association between cas (CONICET), Argentina (Sebastián J. Lipina); and The National
hippocampal volume and parental nurturance disappeared at age Institute of Child Health and Human Development Grant HD
8, suggesting the existence of a sensitive developmental period for 060563 to Georgia State University, USA (Michael I. Posner).

REFERENCES val66met allele is associated with auditory selective attention vary with cognitive functioning: moving from
Ansari, D., García, N., Lucas, E., Hamon, reduced hippocampal volume in their socioeconomic status. Neu- correlation to causation. Wiley Inter-
K., and Dhital, B. (2005). Neu- healthy subjects. Biol. Psychiatry 59, ropsychology 22, 293–300. discip. Rev. Cogn. Sci. 3, 377–386.
ral correlates of number process in 812–815. Dehaene, S. (1996). The organization Duncan, G. J., Morris, P. A., and
children and adults. Neuroreport 16, Bush, G., Luu, P., and Posner, M. of brain activations in number com- Rodrigues, C. (2011). Does money
1769–1773. I. (2000). Cognitive and emotional parison: event-related potentials and really matter? Estimating impacts of
Bavelier, D., and Neville, H. J. (2002). influences in the anterior cingu- the additive-factors method. J. Cogn. family income on young children’s
Cross-modal plasticity: where and late cortex. Trends Cogn. Sci. 4/6, Neurosci. 8, 47–68. achievement with data from random-
how? Nat. Rev. Neurosci. 3, 443–452. 215–222. Dehaene, S., Posner, M. I., and Tucker, assignment experiments. Dev. Psy-
Ben-Shachar, M., Dougherty, R. F., Cantlon, J. F., Brannon, E. N., Carter, E. D. M. (1994). Localization of a neural chol. 47, 1263–1279.
Deutsch, G. K., and Wandell, B. A. J., and Pelphrey, K. A. (2006). Func- system for error detection and com- Eimas, P. D., Siqueland, E. R., Jusczyk,
(2011). The development of cortical tional imaging of numerical process- pensation. Psychol. Sci. 5, 303–305. P., and Vigorito, J. (1971). Speech
sensitivity to visual word forms. J. ing in adults and 4-y-old children. Dehaene, S., Spelke, E., Pinel, P., perception in infants. Science 171,
Cogn. Neurosci. 23, 2387–2399. PLoS Biol. 4, e125. doi: 10.1371/jour- Stanescu, R., and Tsivkin, S. (1999). 303.
Berger, A., Tzur, G., and Posner, M. nal.pbio.0040125. Sources of mathematical thinking: Evans, G. W. Gonnella, C., Mar-
I. (2006). Infant babies detect arith- Cohen, L., Henry, C., Dehaene, S., Mar- behavioral and brain-imaging evi- cynyszyn, L. A., Gentile, L., and
metic error. Proc. Natl. Acad. Sci. tinaud, O., Lehericy, S., Lemer, C., dence. Science 284, 970–974. Salpekar, N. (2005). The role of chaos
U.S.A. 103, 12649–12553. and Ferrieux, S. (2004). The patho- Dehaene-Lambertz, G., Hertz-Pannier, in poverty and children’s socioemo-
Bialystok, E., and Martin, M. M. (2004). physiology of letter-by-letter reading. L., Dubois, J., Meriaux, S., Roche, A., tional adjustment. Psychol. Sci. 16,
Attention and inhibition in bilingual Neuropsychologia 42, 1768–1780. Sigman, M., and Dehaene, S. (2006). 560–565.
children: evidence from the dimen- Conger, R. D., and Donnellan, M. B. Functional organization of perisyl- Evans, G. W., and Kim, P. (2010).
sional change card task. Psychol. Sci. (2007). An interactionist perspec- vian activation during presentation Multiple risk exposure as a poten-
7, 325–339. tive on the socioeconomic context of sentences in preverbal infants. tial explanatory mechanism for the
Blair, C., and Razza, R. P. (2007). Relat- of human development. Annu. Rev. Proc. Natl. Acad. Sci. U.S.A. 103, socioeconomic status-health gradi-
ing effortful control, executive func- Psychol. 58, 175–199. 14240–14245. ent. Ann. N. Y. Acad. Sci. 1186,
tion, and false belief understanding Crane, J. (1996). Effects of home Diamond, A., Barnett, S., Thomas, 174–189.
to emerging math and literacy abil- environment, SES and maternal test J., and Munro, S. (2007). Preschool Fair, D., Cohen, A. L., Power, J. D.,
ity in kindergarten. Child Dev. 78, scores on mathematics achievement. improves cognitive control. Science Dosenbach, N. U. F., Church, J.
647–663. J. Educ. Res. 89, 305–314. 318, 1387–1388. A., Meizin, F. M., Schlaggar, B. L.,
Bradley, R. H., and Corwyn, R. F. Crone, E. A., and Ridderinkhof, K. R. Diamond, A., and Lee, K. (2011). and Petersen, S. E. (2009). Func-
(2002). Socioeconomic status and (2011). The developing brain: from Interventions shown to aid execu- tional brain networks develop from
child development. Annu. Rev. Psy- theory to neuroimaging and back. tive function development in chil- a “local to distributed” organization.
chol. 53, 371–399. Dev. Cogn. Neurosci. 1, 101–109. dren 4 to 12 years old. Science 333, PLoS Comput. Biol. 5, e1000381. doi:
Bueller, J. A., Aftab, M., Sen, S., D’Angiulli, A., Herdman, A., Stapells, 959–964. 10.1371/journal.pcbi.1000381
Gomez-Hassan, D., Burmeister, M., D., and Hertzman, C. (2008). Chil- Duncan, G. J., and Magnuson, K. Fair, D. A., Dosenbach, N. U. F., Church,
and Zubieta, J. K. (2006). BDNF dren’s event-related potentials of (2012). Socioeconomic status and J. A., Cohen, A. L., Brahmbhatt,

Frontiers in Human Neuroscience www.frontiersin.org August 2012 | Volume 6 | Article 238 | 9

Lipina and Posner Poverty and brain networks

S., Miezin, F. M., Barch, D. M., Guttorm, T. K., Leppanen, P. H. T., M. I. Posner (New York: Guilford), McAllister, A. K., Lo, D. C., and
Raichle, M. E., Petersen, S. E., and Poikkeus, A. M., Eklund, K. M., Lyyti- 475–486. Katz, L. C. (1995). Neurotrophins
Schlaggar, B. L. (2007). Development nen, P., and Lyytinen, H. (2005). Korenman, S., Miller, J. E., and regulate dendritic growth in devel-
of distinct control networks through Brain event-related potentials (ERPs) Sjaastad, J. E. (1995). Long-term oping visual cortex. Neuron 15,
segregation and integration. Proc. measured at birth predicts later lan- poverty and child development in 791–803.
Natl. Acad. Sci. U.S.A. 104, 13507– guage development in children with the United States: results from McCandliss, B. D., Beck, I. L., Sandak,
13512. and without familial risk for dyslexia. the National Longitudinal Survey of R., and Perfetti, C. (2003a). Focusing
Fair, D. A., Dosenbach, N. U. F., Cortex 41, 291–303. Youth. Child. Youth Serv. Rev. 17, attention on decoding for children
Petersen, S. E., and Schlaggar, B. Hackman, D. A., and Farah, M. J. 127–151. with poor reading skills: design and
L. (2011). “Resting state studies on (2009). Socioeconomic status and the Kuhl, P. K. (1994). Learning and rep- preliminary tests of the Word Build-
the development of control systems,” developing brain. Trends Cogn. Sci. resentation in speech and language. ing Intervention. Sci. Stud. Read. 7,
in Cognitive Neuroscience of Atten- 13, 65–73. Curr. Opin. Neurobiol. 4, 812–822. 75–104.
tion, ed. M. I. Posner (New York: Hackman, D. A., Farah, M. J., and Kuhl, P. K., Stevens, E., Hayashi, A., McCandliss, B. D., Cohen, L., and
Guilford), 291–311. Meaney, M. J. (2010). Socioeconomic Deguchi, T., Kiritani, S., and Iver- Dehaene, S. (2003b). The visual word
Fan, J., McCandliss, B. D., Sommer, T., status and the brain: mechanistic son, P. (2006). Source: infants show a form area: expertise for reading in the
Raz, M., and Posner, M. I. (2002). insights from human and animal facilitation effect for native language fusiform gyrus. Trends Cogn. Sci. 7,
Testing the efficiency and indepen- research. Nat. Rev. Neurosci. 11, phonetic perception between 6 and 293–299.
dence of attentional networks. J. 651–659. 12 months. Dev. Sci. 9, F13–F21. McClelland, J. L., Fiez, J. A., and
Cogn. Neurosci. 3, 340–347. Hebb, D. (1949). The Organization of Kuhl, P. K., Tsao, F. M., and Liu, H. McCandliss, B. D. (2002). Teaching
Farah, M. J., Shera, D. M., Savage, J. Behavior. New York: John Wiley & M. (2003). Foreign-language experi- the/r/-/l/discrimination to Japanese
H., Betancourt, L., Giannetta, J. M., Sons. ence in infancy: effects of short-term adults: behavioral and neural aspects.
Brodsky, N. L., Malmud, E. K., and Heckman, J. J. (2006). Skill formation exposure and social interaction on Physiol. Behav. 77, 657–662.
Hurt, H. (2006). Childhood poverty: and the economics of investing in phonetic learning. Proc. Natl. Acad. McEwen, B. S., and Gianaros, P. J.
specific associations with neurocog- disadvantaged children. Science 312, Sci. U.S.A. 100, 9096–9101. (2010). Central role of the brain
nitive development. Brain Res. 1110, 1900–1902. Lipina, S. J., and Colombo, J. A. in stress and adaptation: links to
166–174. Heckman, J. J. (2008). Schools, (2009). Poverty and Brain Develop- socioeconomic status, health, and
Galván, A. (2010). Neural plasticity skills, and synapses. Econ. Inq. 46, ment During Childhood: An Approach disease. Ann. N. Y. Acad. Sci. 1186,
of development and learning. Hum. 289–324. from Cognitive Psychology and Neuro- 190–222.
Brain Mapp. 31, 879–890. Hillman, C. H., Eriksen, K. I., and science. Washington, DC: American McGowan, P. O., Sasaki, A., D’Alessio, A.
Gao, W., Zhu, H., Giovanello, K. S., Kramer, A. F. (2008). Be smart, exer- Psychological Association. C., Dymov, S., Labonte, B., Szyf, M.,
Smith, J. K., Shen, D., Gilmore, cise your heart: exercise effects on Lipina, S. J., Martelli, M. I., Vuelta, B., Turecki, G., and Meaney, M. J. (2009).
J. H., and Lin, W. (2009). Evi- brain and cognition. Nat. Rev. Neu- and Colombo, J. A. (2005). Perfor- Epigenetic regulation of the gluco-
dence on the emergence of the brain’s rosci. 9, 58–65. mance on the A-not-B task of Argen- corticoid receptor in human brain
default network from 2 week-old to Hoff, E. (2003). The specificity of envi- tinean infants from unsatisfied and associates with childhood abuse. Nat.
2-year old healthy pediatric subjects. ronmental influence: socioeconomic satisfied basic needs homes. Interam. Neurosci. 12, 342–348.
Proc. Natl. Acad. Sci. U.S.A. 106, status affects early vocabulary devel- J. Psychol. 39, 49–60. Mezzacappa, E. (2004). Alerting, orient-
6790–6795. opment via maternal speech. Child Lipina, S. J., Martelli, M. I., Vuelta, B., ing, and executive attention: devel-
Gaser, C., and Schlaug, G. (2003). Gray Dev. 74, 1368–1378. Injoque Ricle, I., and Colombo, J. A. opmental properties and sociodemo-
matter differences between musicians Imfeld, A., Oechslin, M. S., Meyer, (2004). Pobreza y desempeño ejec- graphic correlates in an epidemi-
and non-musicians. Ann. N. Y. Acad. M., Loenneker, T., and Jancke, L. utivo en alumnos preescolares de la ological sample of young, urban
Sci. 999, 514–517. (2009). White matter plasticity in the ciudad de Buenos Aires (Argentina) children. Child Dev. 75, 1373–
Gould, E., McEwen, B. S., Tanapat, P., corticospinal tract of musicians: a [Poverty and executive performance 1386.
Galea, L. A., and Fuchs, E. (1997). diffusion tensor imaging study. Neu- in preschooler from the City of Miller, C. A., and Sweatt, J. D. (2007).
Neurogenesis in the dentate gyrus roimage 1, 600–607. Buenos Aires (Argentina)]. Interdis- Covalent modification of DNA regu-
of the adult tree shrew is regulated Iverson, P., Hazan, V., and Bannister, K. ciplinaria 21, 153–193. lates memory formation. Neuron 53,
by psychosocial stress and NMDA (2005). Phonetic training with acous- Lipina, S. J., Simonds, J., and Segretin, 857–869.
receptor activation. J. Neurosci. 1, tic cue manipulations: a comparison M. S. (2011). Recognizing the child in Moffitt, T. E., Arseneault, L., Belsky,
2492–2498. of methods for teaching English r–l child poverty. Vulnerable Child. Youth D., Dickson, N., Hancox, R. J., Har-
Griffin, S. A., Case, R., and Siegler, R. to Japanese adults. J. Acoust. Soc. Am. Stud. 6, 8–17. rington, H., Houts, R., Poulton, R.,
S. (1995). “Rightstart: providing the 118, 3267–3278. Lupien, S. J., McEwen, B. S., Gunnar, M. Roberts, B. W., Ross, S., Sears, M.
central conceptual prerequisites for Jones, L., Rothbart, M. K., and Pos- R., and Heim, C. (2009). Effects of R., Thomson, W. M., and Caspi, A.
first formal learning of arithmetic to ner, M. I. (2003). Development of stress throughout the lifespan on the (2011). A gradient of childhood self-
students at risk for school failure,” inhibitory control in preschool chil- brain, behavior and cognition. Nat. control predicts health, wealth, and
in Classroom Lessons: Integrating dren. Dev. Sci. 6, 498–504. Rev. Neurosci. 10, 434–445. public safety. Proc. Natl. Acad. Sci.
Cognitive Theory, ed. K. McGilly Kishiyama, M. M., Boyce, W. T., Markham, J. A., and Greenough, W. U.S.A. 108, 2693–2698.
(Cambridge MA: MIT Press), Jimenez, A. M., Perry, L. M., and T. (2004). Experience-driven brain Mohammed, A. H., Zhu, S. W., Dar-
25–50. Knight, R. T. (2009). Socioeconomic plasticity: beyond the synapse. Neu- mopil, S., Hjerling-Leffler, J., Ernfors,
Grossman, A. W., Churchill, J. D., McK- disparities affect prefrontal function ron Glia Biol. 1, 351–363. P., Winblad, B., Diamond, M. C.,
inney, B. C., Kodish, I. M., Otte, in children. J. Cogn. Neurosci. 21, Marmot, M., Allen, J., Bell, R., and Eriksson, P. S., and Bogdanovic, N.
S. L., and Greenough, W. T. (2003). 1106–1115. Goldblatt, P. (2012). Building of the (2002). Environmental enrichment
Experience effects on brain devel- Kleim, J. A., Jones, T. A., and Schallert, global movement for health equity: and the brain. Prog. Brain Res. 138,
opment: possible contributions to T. (2003). Motor enrichment and the from Santiago to Rio and beyond. 109–133.
psychopathology. J. Child Psychol. induction of plasticity before or after Lancet 14, 181–188. Molfese, D. L. (2000). Predicting
Psychiatry 44, 33–63. brain injury. Neurochem. Res. 28, Mathews, K. A., and Gallo, L. C. dyslexia at eight years of age using
Guo, G. (1998). The timing of the influ- 1757–1769. (2011). Psychological perspectives on neonatal brain responses. Brain Lang.
ences of cumulative poverty on chil- Klingberg, T. (2012). “Training work- pathways linking socioeconomic sta- 72, 238–245.
dren’s cognitive ability and achieve- ing memory and attention,” in Cog- tus and physical health. Annu. Rev. Monzalvo, K., Fluss, J., Billard, C.,
ment. Soc. Forces 77, 257–288. nitive Neuroscience of Attention, ed. Psychol. 62, 501–530. Dehaene, S., and Dehaene-Lambertz,

Frontiers in Human Neuroscience www.frontiersin.org August 2012 | Volume 6 | Article 238 | 10

Lipina and Posner Poverty and brain networks

G. (2012). Cortical networks for word form area”. Neuroimage 22, Posner, M. I. (2005). Training, mat- Tang, Y., Yang, L., Leve, L. D., and
vision and language in dyslexic and 477–479. uration, and genetic influences on Harold, G. T. (2012). Improving exec-
normal children of variable socio- Raichle, M. E. (2009). A paradigm shift the development of executive atten- utive function and its neurobiological
economic status. Neuroimage 61, in functional imaging. J. Neurosci. 29, tion. Proc. Natl. Acad. Sci. U.S.A. 102, mechanisms through a mindfulness-
258–274. 12729–12734. 14931–14936. based intervention: advances within
Neville, H. J., Stevens, C., Klein, S., Raizada, R. D. S., and Kishiyama, M. Saffran, J. R. (2002). Constraints on the field of Developmental Neuro-
Fanning, J., Bell, T., Isbell, E., M. (2010). Effects of socioeconomic statistical language learning. J. Mem. science. Child Dev. Perspect. doi:
and Pakulak, E. (2011). Improving status on brain development, and Lang. 47, 172–196. 10.1111/j.1750-8606.2012.00250.x
behavior, cognition and neural mech- how cognitive neuroscience may con- Sale, A., Berardi, N., and Maffei, L. [Epub ahead of print].
anisms of attention in lower SES tribute to levelling the playing field. (2008). Enrich the environment to Tang, Y. Y., Ma, Y., Wang, J., Fan, Y.,
children. Abstr. Soc. Neurosci. 41, Front. Hum. Neurosci. 4:3. doi: empower the brain. Trends Neurosci. Feng, S., Lu, Q., Yu, Q., Sui, D., Roth-
BIOSIS: PREV201200053638. 10.3389/neuro.09.003.2010. 32, 233–239. bart, M. K., Fan, M., and Posner,
Noble, K. G., McCandliss, B. D., and Raizada, R. D. S., Richards, T. L., Melt- Schlaggar, B. L., and McCandliss, B. D. M. I. (2007). Short-term meditation
Farah, M. J. (2007). Socioeconomic zoff, A., and Kuhl, P. K. (2008). (2007). Development of neural sys- training improves attention and self-
gradients predict individual differ- Socioeconomic status predicts hemi- tems for reading. Annu. Rev. Neurosci. regulation. Proc. Natl. Acad. Sci.
ences in neurocognitive abilities. Dev. spheric specialization of the left infe- 30, 475–503. U.S.A. 104, 17152–17156.
Sci. 10, 464–480. rior frontal gyrus in young children. Segretin, M. S., Goldin, A., Hermida, M. Tang, Y. Y., and Posner, M. I. (2009).
Noble, K. G., Norman, M. F., and Neuroimage 40, 1392–1401. J., Elías Costa, M., Sigman, M., and Attention training and attention state
Farah, M. J. (2005). Neurocognitive Ramani, G. B., and Siegler, R. S. Lipina, S. J. (2012). “Diseño e imple- training. Trends Cogn. Sci. 13,
correlates of socioeconomic status in (2008). Promoting broad and stable mentación de un programa computa- 222–227.
kindergarten children. Dev. Sci. 8, improvements in low-income chil- rizado de entrenamiento de procesos Tang, Y. Y., Zhang, W. T., Chen, K. W.,
74–87. dren’s numerical knowledge through cognitivos básicos en niños de edad Feng, S. H., Ji, Y., Shen, J., Reiman,
Oberlander, T., Weinberg, J., Paps- playing number board games. Child escolar (design and implementation E. M., and Liu, Y. (2006). Arithmetic
dorf, M., Grunau, R. Misri, S., and Dev. 79, 375–394. of a computerized program for train- processing in the brain shaped by cul-
Devlin, M. D. (2008). Prenatal expo- Rao, H., Betancourt, L., Giannetta, J. ing of basic cognitive processes in tures. Proc. Natl. Acad. Sci. U.S.A.
sure to maternal depression, neonatal M., Brodsky, N. L., Korczykowski, school-aged children),” in La pizarra 103, 10775–10780.
methylation of human glucocorti- M., Avants, B. B., Gee, J. C., Wang, de Babel. Puentes entre neurociencia, Temple, E., Deutsch, G. K., Pol-
coid receptor gene (Nr3c1) and infant J., Hurt, H., Detre, J. A., and Farah, psicología y educación (The Board of drack, R. A., Miller, S. L., Tallal, P.,
cortisol stress responses. Epigenetics M. J. (2010). Early parental care Babel. Bridges Between Neuroscience, Merzenich, M. M., and Gabrieli, J.
3, 97–106. is important for hippocampal mat- Psychology and Education), eds S. J. D. (2003). Neural deficits in children
Pappas, S., Ginsburg, H. P., and Jiang, uration: evidence from brain mor- Lipina and M. Sigman (Buenos Aires: with dyslexia ameliorated by behav-
M. (2003). SES differences in young phology in humans. Neuroimage 49, Del Zorzal), 265–278. ioral remediation: evidence from
children’s metacognition in the con- 1144–1150. Shaywitz, B. A., Skudlarski, P., Holahan, functional MRI. Proc. Natl. Acad. Sci.
text of mathematical problem solv- Rosenzweig, M. R. (2003). Effects of J. M., Marchione, R. N., Constable, U.S.A. 100, 2860–2865.
ing. Cogn. Dev. 18, 431–450. differential experience on the brain R. T., Fullbright, R. K., Zelterman, Temple, E., and Posner, M. I. (1998).
Petersen, S. E., and Posner, M. I. (2012). and behavior. Dev. Neuropsychol. 24, D., Lacadie, C., and Shaywitz, S. E. Brain mechanisms of quantity are
The attention system of the human 523–540. (2007). Age-related changes in read- similar in 5-year-olds and adults.
brain: 20 years after. Annu. Rev. Rosenzweig, M. R., and Bennet, E. L. ing systems of dyslexic children. Ann. Proc. Natl. Acad. Sci. U.S.A. 95,
Neurosci. 35, 73–89. (1996). Psychobiology of plasticity: Neurol. 61, 363–370. 7836–7841.
Posner, M. I., and McCandliss, B. effects of training and experience on Shors, T. J., Miesegaes, G., Beylin, A., Vicari, S., Albertoni, A., Chilosi A. M.,
D. (1999). “Brain circuitry during brain and behavior. Behav. Brain Res. Zhao, M., Rydel, T., and Gould, E. Cipriani, P., Cioni, G., and Bates, E.
reading,” in Converging Methods for 78, 57–65. (2001). Neurogenesis in the adult is (2000). Plasticity and reorganization
Understanding Reading and Dyslexia, Roth, T. L., and Sweatt, J. D. (2011). involved in the formation of trace during language development in chil-
eds R. Klein and P. McMullen (Cam- Epigenetic mechanisms and environ- memories. Nature 410, 372–376. dren with early brain injury. Cortex
bridge MA: MIT Press), 305–337. mental shaping of the brain dur- Siegler, R. S. (2009). Improving the 36, 31–46.
Posner, M. I., and Petersen, S. E. (1990). ing sensitive periods of develop- numerical understanding of children Walker, S. P., Wachs, T. D., Grantham-
The attention system of the human ment. J. Child Psychol. Psychiatry 52, from low-income families. Child Dev. McGregor, S., Black, M. M., Nel-
brain. Annu. Rev. Neurosci. 13, 25–42. 398–408. Perspect. 3, 118–124. son, C. A., Huffman, S. L., Baker-
Posner, M. I., and Raichle, M. E. (1994). Roth, T. L., Lubin, F. D., Funk, A. J., and Stevens, C., Fanning, J., Coch, D., Henningham, H., Chang, S. M.,
Images of Mind. New York: Scientific Sweatt, J. D. (2009). Lasting epige- Sanders, L., and Neville, H. (2008). Hamadani, J. D., Lozoff, B., Gardner,
American Library. netic influence of early-life adversity Neural mechanisms of selective audi- J. M., Powell, C. A., Rahman, A., and
Posner, M. I., and Rothbart, M. K. on the BDNF gene. Biol. Psych. 1, tory attention are enhanced by com- Richter, L. (2011). Inequality in early
(2007a). Educating the Human Brain. 760–769. puterized training: electrophysiologi- childhood: risk and protective factors
Washington DC: American Psycho- Rueda, M. R., Checa, P., and Com- cal evidence from language-impaired for early child development. Lancet
logical Association. bita, L. M. (2012). Enhanced effi- and typically developing children. 6736, 1–14.
Posner, M. I., and Rothbart, M. K. ciency of the executive attention Brain Res. 1205, 55–69. Werker, J. F., Gilbert, J. H. V., Humphrey,
(2007b). Attention as a model sys- network after training in preschool Stevens, C., Lauinger, B., and Neville, K., and Tees, R. C. (1981). Devel-
tem for the integration of cogni- children: immediate and after two H. (2009). Differences in the neu- opmental aspects of cross-language
tive science. Annu. Rev. Psychol. 58, month effects. Dev. Cogn. Neurosci. 2, ral mechanisms of selective attention speech-perception: Child Dev. 52,
1–23. S192–S204. in children from different socioeco- 349–355.
Posner, M. I., Rothbart, M. K., Sheese, Rueda, M. R., Fan, J., McCandliss, B. nomic backgrounds: an event-related Wilson, A. J., Dehaene, S., Dubois,
B. E., and Voelker, P. (2012). Con- D., Halparin, J. D., Gruber, D. B., Ler- brain potential study. Dev. Sci. 12, O., and Fayol, M. (2009). Effects of
trol networks and neuromodulators cari, L. P., and Posner, M. I. (2004). 634–646. an adaptive game intervention on
of early development. Dev. Psychol. Development of attentional networks Streeter, L. A. (1976). Language per- accessing number sense in low socio-
48, 827–835. in childhood. Neuropsychologia 42, ception of two-month-old infants economic status kindergarten chil-
Price, C. J., and Devlin, J. T. (2004). 1029–1040. shows effects of both innate mech- dren. Mind Brain Educ. 3, 224–234.
The pro and cons of labeling a left Rueda, M. R., Rothbart, M. K., McCan- anisms and experience. Nature 259, Woollett, K., and Maguire, E. A.
occipitotemporal region: “the visual dliss, B. D., Saccomanno, L., and 39–41. (2011). Acquiring “the Knowledge”

Frontiers in Human Neuroscience www.frontiersin.org August 2012 | Volume 6 | Article 238 | 11

Lipina and Posner Poverty and brain networks

of London’s layout drives structural function. Annu. Rev. Psychol. 61, Received: 22 March 2012; accepted: 26 Copyright © 2012 Lipina and Posner.
brain changes. Curr. Biol. 20, 2109– 439–466. July 2012; published online: 17 August This is an open-access article distributed
2114. 2012. under the terms of the Creative
Wynn, K. (1992). Addition and subtrac- Conflict of Interest Statement: The Commons Attribution License, which
tion by human infants. Nature 358, authors declare that the research was Citation: Lipina SJ and Posner MI (2012) permits use, distribution and reproduc-
749–750. conducted in the absence of any The impact of poverty on the development tion in other forums, provided the origi-
Zhang, T. Y., and Meaney, M. J. (2010). commercial or financial relationships of brain networks. Front. Hum. Neu- nal authors and source are credited and
Epigenetics and the environmental that could be construed as a potential rosci. 6:238. doi: 10.3389/fnhum.2012. subject to any copyright notices concern-
regulation of the genome and its conflict of interest. 00238 ing any third-party graphics etc.

Frontiers in Human Neuroscience www.frontiersin.org August 2012 | Volume 6 | Article 238 | 12