5Perspectives from Developmental Neuroscience

Publication Details

Chapter 4 described the multiple risk and protective factors that can play a role in mental, emotional, and behavioral (MEB) disorders and that can inform the design of prevention interventions, placing these contributing factors in the framework of developmental processes. This chapter illustrates research advances in the framework of developmental neuroscience, including the anatomical and functional development of the brain, molecular and behavioral genetics, molecular and cellular neurobiology, and systems-level neuroscience, that relate to the prevention of MEB disorders. Perspectives from developmental neuroscience provide a foundation for understanding the development of cognitive abilities, emotions, and behaviors during childhood and adolescence, and they thereby reveal valuable opportunities for novel advances in future prevention research.

Reducing Risks for Mental Disorders: Frontiers for Preventive Intervention Research, the 1994 Institute of Medicine (IOM) report, emphasized the importance of the relationship between prevention research and a knowledge base that includes both basic and applied research in neurobiology and genetics. This knowledge base contributes to the understanding of the causes, course, and outcomes of MEB disorders, and it continues to be increasingly important for informing how prevention efforts may intervene in causal pathways that lead to disorders.

In the years since the 1994 IOM report, understanding of the biological processes that underlie brain development has grown at an unprecedented rate, and the past several decades have witnessed much greater interest in the neurobiological underpinnings of MEB disorders. These disorders are increasingly being understood as dynamic disruptions in key developmental processes that exert their effects throughout the life span. Unraveling the causes and consequences of complex MEB disorders remains an enormous challenge. However, major advances have been made not only in identifying genetic and environmental factors that play causal roles in the genesis of disorders, but also in understanding more fully the interaction between genetic and environmental influences in causing or protecting against specific diseases. In addition, advances in the emerging field of epigenetics have begun to provide information about the complex ways in which genetic traits are expressed as disease and the possible mechanisms through which environment and experience can influence gene expression.

This chapter begins with the role of genetics and the interplay of genetic and environmental factors in MEB disorders. This is followed by a discussion of brain development and its relationship to MEB disorders. Next is an examination of neural systems and their role in complex processes that underlie the cognitive and social competence that is essential to healthy emotional and behavioral development. The third section addresses the relationship between developmental neuroscience and prevention science. The final section presents conclusions and recommendations.


The importance of understanding genetic influences in brain development goes well beyond simply explaining the hereditary components of disorders. Genes are the basic component from which the brain’s structure and function are determined and regulated. Genes encode proteins, and proteins are the building blocks of cells, interacting with the molecular and physical features of their surroundings to determine cellular structure and function. Individual cells interact functionally with other cells within the neural circuits that make up the structure of the brain, which in turn interact with other neural circuits to determine behaviors. Behaving organisms interact with their environments, which can cause adaptive changes in neural systems, circuits, and cells and ultimately in the expression of genes—which in turn modifies brain structure and function. The complexity of the pathways connecting the genes and the environments of organisms to their behaviors has frustrated most attempts to correlate genes directly with behaviors and with specific diagnostic syndromes in the field of psychiatric genetics (Inoue and Lupski, 2003; Joober, Sengupta, and Boksa, 2005; Sanders, Duan, and Gejman, 2004; van den Bree and Owen, 2003).

Inherited or sporadic genetic mutations can profoundly affect the production, structure, or function of the protein that a gene encodes. This can have a dramatic and highly consistent effect in producing disease. However, more subtle variations in the genetic sequence can also affect protein structure and function, producing much more subtle effects. For example, many of the genetic variants that have been associated with MEB disorders are single nucleotide polymorphisms, that is, substitutions of single nucleotides, the structural components of the genetic sequence (van Belzen and Heutnik, 2006; Sanders, Duan, and Gejman, 2004). Variability in the number of copies of a specific gene sequence (known as copy number variants), which can be caused by rearrangements, microdeletions, or microduplications of the sequence, has also emerged as an important contributor to MEB disorders (Lee and Lupski, 2006), such as schizophrenia (Walsh, McClellan, et al., 2008; Xu, Roos, et al., 2008; International Schizophrenia Consortium, 2008; Stefansson, Rujescu, et al., 2008) and autism (Sebat, Lakshmi, et al., 2007; Marshall, Noor, et al., 2008). These kinds of gene variations can have a more graded influence on molecular and cellular functions than do large deletions or rearrangements of genes. The influences of these gene variants on the structural and functional features of cells, neural circuits, and the behaviors they subserve are correspondingly graded as well.

Variations in the genetic sequences that encode proteins are only one level of influence on the expression of those genes in the production of cellular proteins. Variations in the sequence of the nonencoding, regulatory portions of a gene also have important influences on its expression, as can variations in other genes that encode regulatory proteins. In addition, microRNAs (small sequences of RNA, an intermediate genetic component in the process of making proteins from DNA) can influence the expression of genes and their protein products by altering how the proteins are generated from a gene sequence (Boyd, 2008; Stefani and Slack, 2008). These additional levels of regulation can determine when in the course of development, where in the brain, and to what degree a gene is expressed—all without changing the DNA sequence of the gene.

Many studies, including family studies and gene association studies, have demonstrated a genetic component to MEB disorders (Thapar and Stergiakouli, 2008; van Belzen and Heutnik, 2006). However, genetic studies have not yet found an association of single genes with most MEB disorders. Instead, sequence variants in multiple genes have been shown to be associated with an elevated risk or susceptibility for developing many diseases, such as autism (Muhle, Trentacoste, and Rapin, 2004), depression (Levinson, 2006; Lopez-Leon, Janssens, et al., 2008), schizophrenia (Owen, O’Donovan, and Harrison, 2005), addiction (Goldman, Oroszi, and Ducci, 2005), and bipolar disorder (Serretti and Mandelli, 2008). A review of these many reported associations of specific genes with individual disorders is beyond the scope of this report.

In nearly all instances of these reported associations, the influence of individual genes on the risk for developing a disorder is small (Kendler, 2005; Thapar and Stergiakouli, 2008), usually less than the influence of family history and less than that of other nongenetic risk factors. The association is also often nonspecific (Kendler, 2005), with single gene variants being associated with multiple disorders. Moreover, genetic profiles vary greatly among affected individuals. Not everyone with the susceptibility variant in any one of the associated genes will develop the disorder, and not everyone with a particular disorder will have the susceptibility variant of any associated gene. Therefore, a single genetic variant will rarely be necessary or sufficient to produce a disorder, a point similar to findings on the association of environmental risk factors with MEB disorders (described later in this chapter and in Chapter 4). One strategy that has emerged to address the complexity of linking genes to disorders is to identify more narrowly defined behaviors, characteristics, or biological markers, termed “endophenotypes,” that correlate with specific disorders or that are common to more than one disorder. These endophenotypes can serve as a simpler, more readily identifiable focus of genetic studies (Caspi and Moffitt, 2006; Gottesman and Gould, 2003; van Belzen and Heutink, 2006).

Beyond finding associations between genetic variants and MEB disorders or endophenotypes, identifying the effects that specific genes have on molecular pathways, cellular organization, functioning of neural networks, and behavior is crucially important to developing effective intervention approaches based on the modifiable components of the pathways from genes to behavior. This level of genetic research requires experimental manipulations in animal models. Most commonly this involves modification of the genome of mice by inserting, deleting, or mutating specific genes and, in some cases, controlling where in the brain, in what cell types, and when during the course of development a gene is turned off or on. This extraordinary degree of spatial and temporal control over gene expression makes animal models invaluable in identifying the molecular processes of normal and pathological brain development. The disadvantage of animal models, however, is the difficulty of representing the complex cognitive, behavioral, and emotional symptoms experienced by humans. Although the effects of experimental manipulation on certain aspects of cognition and memory can be assessed through the ability of animals to learn and repeat standardized tasks, analogues of emotional experience and thought can be inferred only through behavior that must be correlated with subjective human experience (Cryan and Holmes, 2005; Joel, 2006; McKinney, 2001; Murcia, Gulden, and Herrup, 2005; Powell and Miyakawa, 2006; Sousa, Almeida, and Wotjak, 2006).

Animal models are proving to be of central importance in identifying the likely disturbances in molecular and cellular pathways caused by single gene mutations in some neurodevelopmental disorders, including the fragile X, Prader-Willi, Angelman, and Rett syndromes. Knowledge of those molecular pathways already has led to promising treatment approaches in animal models (Chang, Bray, et al., 2008; Bear, Dolen, et al., 2008; Chahrour and Zoghbi, 2007; Dolen, Osterweil, et al., 2007; Giacometti, Luikenhuis, et al., 2007; Guy, Gan, et al., 2007). Animal models have also successfully linked risk genes with disturbances in particular molecular pathways that may predispose to the development of more complex, polygenic disorders, such as depression (Cryan and Holmes, 2005; Urani, Chourbaji, and Gass, 2005), anxiety disorders (Cryan and Holmes, 2005), obsessive compulsive disorder (Joel, 2006), autism (Moy and Nadler, 2008), schizophrenia (O’Tuathaigh, Babovic, et al., 2007), and substance abuse (Kalivas, Peters, and Knackstedt, 2006).

Despite the challenge of studying the role of genes in the etiology of MEB disorders, advances in technology continue to make large-scale genotyping more feasible and affordable, and the combination of human genetics studies and approaches using animal models has proven to be informative in identifying genes of risk in multifactorial, complex non-psychiatric disorders, such as asthma (Moffatt, 2008) and diabetes (Florez, 2008); they will undoubtedly make important contributions in psychiatric genetics in coming years.

Gene–Environment Interactions and Correlations

Most complex behaviors and the most common forms of MEB disorders are likely to arise from a combination of multiple interacting genetic and environmental influences (Caspi and Moffitt, 2006; Rutter, Moffitt, and Caspi, 2006). The effect of a common genetic variant in altering the risk for a disorder, for example, is likely to be conditioned heavily by the experiences of a developing child, just as the effects of experience in producing a disorder are likely to be conditioned by the genetic background that the child inherits from his or her parents (Rutter, Moffitt, and Caspi, 2006; Thapar, Harold, et al., 2007). These so-called gene–environment (GxE) interactions can confer both risk and protective effects on the child relative to the effects of either the genetic or environmental influences in isolation.

A number of interactions between specific identified genes and specific environmental risk factors have been demonstrated in MEB disorders (Rutter, Moffitt, and Caspi, 2006). For example, a landmark prospective epidemiological study found that the number of copies an individual carries of the short variant of a region of the serotonin transporter gene (5-HTTLPR) significantly increases, in a dose-dependent fashion, the risk for developing depressive symptoms, major depressive disorder, and suicidality—but only in the context of adverse or stressful early life experiences (Caspi, Sugden, et al., 2003) (see Figure 5-1). Similarly, a polymorphism in the gene that encodes monoamine oxidase A (MAOA), an enzyme that metabolizes neurotransmitters, moderates the effect of maltreatment on developing antisocial problems later in life (Kim-Cohen, Caspi, et al., 2006; Caspi, McClay, et al., 2002): Maltreated children who have the genotype that confers high levels of MAOA expression are less likely to develop conduct disorder, antisocial personality, or adult violent crime. In another domain, a common polymorphism of the dopamine transporter gene has been reported to interact with the risk conferred by prenatal exposure to tobacco smoke, leading to increased hyperactive-impulsive and oppositional behaviors in later childhood (Kahn, Khoury, et al., 2003).

FIGURE 5-1. Gene–environment interaction between effects of prior maltreatment and genotype for the 5-HTTLPR allele on developing depression later in life.


Gene–environment interaction between effects of prior maltreatment and genotype for the 5-HTTLPR allele on developing depression later in life. Maltreatment has the biggest effect for two copies of the short (s/s) allele and the smallest effect (more...)

In contrast to GxE interactions, gene–environment correlations are genetic influences on variations in the likelihood that an individual will experience specific environmental circumstances (Jaffee and Price, 2007; Rutter and Silberg, 2002; Rutter, Moffitt, and Caspi, 2006). Gene–environment correlations can confound cause and effect and hinder measurement of GxE interactions because a genetically determined behavioral trait can produce a systematic variation in environmental exposure, and that environmental variation can be deemed erroneously to be a cause of a behavioral trait under study (Jaffee and Price, 2007; Lau and Eley, 2008). Children with autism, for example, are chronically and consistently withdrawn from their caregivers. This chronic withdrawal might induce in the caregiver a sense of hopelessness about ever making a deep interpersonal connection with the child, prompting a secondary withdrawal on the part of the caregiver. An unsuspecting researcher might inadvertently and erroneously attribute the child’s impoverished social relatedness to the caregiver’s withdrawal, when in fact it was caused by a particular genetic variant.

Epigenetic Effects

Epigenetic effects are potentially heritable alterations of gene expression that do not involve actual modification of the DNA sequence. Instead, alterations in the level of gene expression are induced by changes in the three-dimensional packaging of DNA that in turn make a gene either more or less amenable to production of a protein product. All known mechanisms that produce epigenetic changes in gene expression involve enzymatic processes that add or remove substrates either from the DNA or from histone proteins that are physically associated with DNA and that determine its three-dimensional packing structure (Tsankova, Renthal, et al., 2007). Epigenetic modifications of gene expression are in continual flux, as competing factors modify and unmodify DNA and its associated proteins, as well as their related behavioral phenotypes.

Epigenetic determinants are increasingly invoked as possible explanations for a multitude of “complex genetic” phenotypes, in which multiple genes are each thought to account for a small amount of variance in the clinical phenotype. Moreover, recent research has shown that epigenetic mechanisms can produce short-term adaptation of the phenotype to a changing environment. For example, abundant naturalistic and experimental evidence in humans and animal models has shown that early experience influences reactivity to stress later in life, even into adulthood, and that epigenetic modification of genes that encode components of the stress response can contribute to these enduring effects (Kaffman and Meaney, 2007; Weaver, 2007).

Perhaps most remarkably, a changing environment has been shown to trigger epigenetic effects that can be transmitted across generations, in species as diverse as yeast and humans (Rakyan and Beck, 2006; Richards, 2006; Whitelaw and Whitelaw, 2006). The quality of maternal care given to rat pups, for example, produces epigenetic modifications of gene expression in the brains of the pups that influence the quality of maternal care they provide as adults to their own offspring. This cross-generation transmission has been shown to account for variability in maternal behavior toward offspring that is either nurturing or neglectful (Champagne, 2008).

Several examples suggest that epigenetic mechanisms are important in understanding the causes and in improving the prevention and treatment of MEB disorders (Tsankova, Renthal, et al., 2007). One well-known example is the Prader-Willi and the Angelman syndromes, disorders with highly distinct phenotypes that are nevertheless both caused by a mutation in the same chromosomal region. Although the locus of the mutation is the same, its effects on the behavioral phenotype of the child differ depending on which parent is the origin of the mutation (Goldstone, 2004; Lalande and Calciano, 2007; Nicholls and Knepper, 2001).

Another example of the importance of epigenetic influences in the cause of a disorder is Rett syndrome, a progressive neurodevelopmental disorder characterized by motor, speech, and social behavioral abnormalities (Chahrour and Zoghbi, 2007). Mutations in the MeCP2 gene cause Rett syndrome and, less commonly, other neurodevelopmental disorders, including classic autism, mental retardation, early-onset bipolar disorder, and early-onset schizophrenia. This gene encodes a protein that epigenetically alters the expression of other genes (Chahrour and Zoghbi, 2007; Zlatanova, 2005). In other words, this specific genetic mutation causes disease through epigenetic mechanisms, underscoring how complex, intimate, and interactive genetic and epigenetic factors are in influencing the development of disorders.

Epigenetic modifications of the genome are also necessary for various learning and memory processes in the brain (Levenson and Sweatt, 2005, 2006; Levenson, Roth, et al., 2006; Reul and Chandramohan, 2007; Fischer, Sananbenesi, et al., 2007), suggesting that these processes may be important in the etiology of various mental retardation syndromes. Epigenetic influences play a prominent role as well in changes in the brain and in behavior related to establishing preferences for drugs of abuse in animal models of addiction (Kumar, Choi, et al., 2005). Finally, epigenetic modifications of the genome have been shown to be necessary to produce the behavioral response to antidepressant medications in a mouse model of depression (Newton and Duman, 2006; Tsankova, Berton, et al., 2006).


MEB disorders in children involve disturbances in the most complex, highly integrated functions of the human brain. Understanding from a biological perspective how these functional capacities develop and how they are disrupted is an immense challenge. This section offers a brief overview of current knowledge about the complex processes that contribute to the normal development of the human brain, along with examples of their relationship to the causes of MEB disorders.

Sources of Knowledge of Human Brain Development

Knowledge of normal human brain development and of the abnormalities that produce disorders is limited by the difficulty of studying the human brain at the level of molecules and cells. The human data on brain development thus far come from a small number of postmortem studies and a larger number of in vivo, or live, brain imaging studies. The scientific value of postmortem studies is limited by the quality and number of tissue samples that are usually available and by the capability to study only a small number of brain regions (Lewis, 2002). In contrast, in vivo imaging has proved to be an important tool for studying postnatal brain development in humans across the life span (Marsh, Gerber, and Peterson, 2008), although thus far it has provided information about brain structure and function mainly at a macroscopic level of brain organization, revealing little molecular or cellular information (Peterson, 2003b).

Understanding of the molecular and cellular development of the human brain is therefore gleaned largely from studies of animal models, extrapolated to the maturational timeline of humans. Although a great deal has been learned from those animal models across a wide range of species, how well those findings relate to the development and function of the human brain is not fully known. Moreover, as noted earlier, the molecular bases of the highest-order functions of the human brain cannot be studied easily in animals.

Despite limited data from human and nonhuman primates, the consistency in findings across species suggests that the general features of brain development in animal models are likely to apply to humans as well. Those findings indicate that the wiring of neural architecture is neither fixed nor static. Instead, it is a dynamic entity that is shaped and reshaped continually throughout development by processes that have their own maturational timetables within and across brain regions. These processes are described briefly here and summarized in Figure 5-2.

FIGURE 5-2. Timeline of major events in brain development.


Timeline of major events in brain development.

Overview: Complexities of Brain Development

At the visible anatomical level, the human brain develops during gestation into a complex structure having distinct anatomical regions and a highly convoluted surface. Similarly, at the level of cellular architecture, the human brain is a highly complex, layered structure made up of many distinct kinds of cells that have highly specific interconnections. During fetal brain development, undifferentiated precursor cells need to divide and multiply. The resulting cells must then differentiate into the correct cell types, migrate to the correct place in the brain, and connect properly with other cells. These links among cells must then be organized into functional circuits that support sensation, perception, cognition, emotion, learning, and behavior. In a healthy intrauterine environment, this series of complex and interrelated neurodevelopmental events is initially under the predetermined control of regulatory genes (Rhinn, Picker, and Brand, 2006). In contrast, much of the fine detail of brain organization—how the brain is “wired”—develops through a combination of genetic influences, experience and other external influences, and the interaction of genes and experience.

Setting Up the Nervous System

The nervous system begins to develop in the human fetus 2 to 3 weeks after conception in a process called neurulation, starting as a layer of undifferentiated precursor cells called the neural plate. These cells eventually give rise to all components of the nervous system. As the initial cells divide to create more cells, the neural plate expands, folds, and fuses to form the neural tube (Detrait, George, et al., 2005; Kibar, Capra, and Gros, 2007). The neural tube continues to enlarge while cells in different parts of the tube become specialized, following a spatial pattern established by predetermined molecular mechanisms. From front to back, the neural tube becomes the forebrain (the cerebral cortices), the midbrain (containing neural pathways to and from the forebrain), the hindbrain (the brainstem and cerebellum), and the spinal cord and peripheral nervous system (Rhinn, Picker, and Brand, 2006).

Various physiological and environmental factors can affect prenatal brain development in ways that are either lethal or seriously debilitating (Detrait, George, et al., 2005; Kibar, Capra, and Gros, 2007). Low levels of the vitamin folic acid, for example, produce anencephaly and spina bifida, disorders of formation of the neural tube. Other prenatal environmental exposures can predispose a developing fetus to the development of MEB disorders later in life. For example, common prenatal infections, such as influenza, and less common ones, such as rubella, toxoplasmosis, and cytomegalovirus, can increase the risk of developing mental retardation, schizophrenia, and autism (Fruntes and Limosin, 2008; Jones, Lopez, and Wilson, 2003; Meyer, Yee, and Feldon, 2007; Pearce, 2001; Penner and Brown, 2007). Prenatal exposure to various environmental toxins, including certain insecticides used in homes and for agricultural purposes (Rauh, Garfinkel, et al., 2006), tobacco smoke (Herrmann, King, and Weitzman, 2008), and alcohol (Alcohol Research and Health, 2000), can impair behavior and cognition later in childhood (Williams and Ross, 2007). Premature birth and low birth weight can also predispose to a wide variety of disorders (Peterson, 2003a), including schizophrenia (Kunugi, Nanko, and Murray, 2001), autism (Kolevzon, Gross, and Reichenberg, 2007), and learning disabilities and educational difficulties (Peterson, 2003a).

The Right Cells in the Right Place

Between weeks 5 and 25 of human fetal gestation, undifferentiated precursor cells divide repeatedly, rapidly giving rise to large numbers of cells that will become neurons. Glial cells, the supporting cells of the nervous system, are also generated, but somewhat later than neurons, between weeks 20 and 40 (de Graaf-Peters and Hadders-Algra, 2006). Once cells are generated, two different processes overlap in time. First, the identity or “fate” of these cells becomes progressively more restricted, until the cells are fully differentiated into a specific type of neuron or glial cell. Second, neurons must travel from the site of their origin to their appropriate final location in the brain to provide the function they will ultimately serve, a process called neuronal migration (de Graaf-Peters and Hadders-Algra, 2006; Levitt, 2003; Rakic, 2003). The precise path of neuronal migration is determined by the timing and position of a cell when it is generated, together with a molecular “map” composed of a variety of molecular signals from neighboring cells that guide the migrating cell to its proper final position in a precise and reproducible manner (de Graaf-Peters and Hadders-Algra, 2006; Levitt, 2003; Rakic, 2003). The number of migrating neurons in the human fetus peaks by about week 20 of gestation, and migration stops by about week 30 (de Graaf-Peters and Hadders-Algra, 2006).

Disturbances in neuronal migration have emerged as a key area of interest in understanding the developmental basis of MEB disorders. Failures in neuronal migration produce an accumulation of neurons in the wrong areas of the brain and, consequently, can lead to disorganized brain structure and function. This can be seen in major malformations of the brain, such as lissencephaly (a brain that lacks the usual, complex folded surface) (Guerrini and Filippi, 2005). More subtle disturbances of neuronal migration can create isolated islands of neurons or disruptions of normal circuit function, leading to seizures (Guerrini and Filippi, 2005). Genetic and environmental influences on neuronal migration can produce even more subtle disturbances in the locations of cells that may not be visible at the gross anatomical level but may nevertheless affect functional circuits. In cortical areas involved in higher-level cognitive functions, these effects potentially can produce subtle changes in the brain’s behavioral, emotional, and cognitive capacities that may not manifest until later in life (Rakic, 2002, 2003).

Establishing Connections

Once cells are properly differentiated and as they are migrating to their final locations in the brain, they grow extensions, called axons and dendrites, that allow them to connect to and communicate with other neurons. Axons are primarily responsible for sending signals to other cells, and dendrites are processes that primarily receive signals from other cells. Axons use the guidance of external molecular signals to find their way to the right target cells with which they will connect and communicate. A combination of growth-promoting and growth-inhibiting signals provides the growing tip of the axon with a map of connectivity to get to the right location and connect with the right target cell (Chilton, 2006; Tessier-Lavigne and Goodman, 1996).

Dendritic growth and branching begins early in development, initially proceeding slowly but then accelerating rapidly starting in the third trimester (de Graaf-Peters and Hadders-Algra, 2006), producing a thickening of the cortex (the complex, multilayered collection of cells composing the entire outer surface of the brain) (Huisman, Martin, et al., 2002). The timing of dendritic growth differs by brain region and by layer of the cortex. For example, dendritic elaboration is slower in frontal than in visual cortex, and it begins in the deeper layers earlier than in more superficial ones (Becker, Armstrong, et al., 1984; de Graaf-Peters and Hadders-Algra, 2006; Huttenlocher, 1990; Michel and Garey, 1984; Mrzljak, Uylings, et al., 1992). Overall, dendritic development is highly active from the third trimester of gestation through the first postnatal year, continuing at lower rates through age 5 years (de Graaf-Peters and Hadders-Algra, 2006).

Differing neuronal cell types have diverse shapes and sizes. Some have relatively simple shapes. Others have many axonal branches, allowing them to innervate and influence more target cells. Some have complex dendritic trees that provide a greater range of input from other cells. This diversity of form and structure provides for a range of computational functions across different kinds of neurons, from a limited signal input and response to a complex integration of multiple signals. Connections are established with cells that are nearby and cells that are much more distant, eventually linking and integrating information from different regions of the brain.

The basis of this communication between neurons is their physical connection, called a synapse. The formation of synapses requires the development of specialized cellular machinery on both the presynaptic side of the synapse (where neurotransmitters are prepared and released from the terminals of axons) and at the postsynaptic target (where receptors for those neurotransmitters receive and process the signal) (Waites, Craig, and Garner, 2005). The rate of synapse formation increases rapidly after about weeks 24–28 and peaks, at the rate of almost 40,000 new synapses per second, between 3 and 15 months after birth (in the primary sensory and prefrontal cortices, respectively) (de Graaf-Peters and Hadders-Algra, 2006; Levitt, 2003).

The synapse is the primary site of information transfer in the nervous system, and it is also likely to be the primary site of learning and memory. Several disorders that begin early in life and are associated with profound intellectual and emotional disability can be considered disturbances of learning and memory. These include fragile X and other causes of mental retardation, Rett syndrome, and autistic spectrum disorders. Genes that have been identified as either causing or increasing the risk for developing these disorders can be conceived as having in common the disruption of normal development and function of synapses (Chao, Zoghbi, and Rosenmund, 2007; Dierssen and Ramakers, 2006; Willemsen, Oostra, et al., 2004; Zoghbi, 2003).

Refining the Nervous System: Use It or Lose It

Neurons and the connections between them are produced in an over-abundance during fetal life relative to their levels at birth and in adulthood. The number of neurons in the human brain, for example, peaks around midgestation. Thereafter, overproduction is reduced through a process of molecularly programmed cell death, called apoptosis (de Graaf-Peters and Hadders-Algra, 2006; Levitt, 2003). For continued survival, neurons require a successful interaction with a target cell, and neurons that do not achieve this interaction will die. Neuronal survival is mediated in part by the limited availability of neurotrophic factors, a class of molecules that are derived from the target cells (Monk, Webb, and Nelson, 2001).

The process of brain development also produces an initial surplus of connections between neurons. Early in postnatal life, the density of synapses in the brain increases dramatically, reaching its peak during infancy (de Graaf-Peters and Hadders-Algra, 2006; Huttenlocher, 1984; Huttenlocher and Dabholkar, 1997; Levitt, 2003). The process of forming synapses, or synaptogenesis, is paired with the complementary process of synaptic pruning, in which some synaptic connections are eliminated. Primates are widely believed to have evolved synaptic pruning as a means for removing synaptic connections that are unused and therefore not needed in the environmental context in which the animal finds itself, while conserving and increasing the efficiency of connections that are useful in that context. Thus, survival of most of the synaptic connections that subserve human behavior is influenced by patterns of neural activity, which in turn are the product of environmental influences and experience (Kandel, Schwartz, and Jessell, 2000).

Studies in humans during childhood are limited but, in combination with data from studies in monkeys, indicate that after the peak of synaptogenesis in infancy, synapse formation and synaptic pruning plateau during childhood and then reach a regressive phase between puberty and adulthood. At that point, a massive, activity-dependent pruning eliminates more than 40 percent of synapses (de Graaf-Peters and Hadders-Algra, 2006; Huttenlocher and Dabholkar, 1997; Levitt, 2003; Rakic, 2002; Rakic, Bourgeois, and Goldman-Rakic, 1994).

Another important process in developing and refining appropriate connectivity in the brain is the wrapping of neuronal axons in an insulating sheath of myelin, which promotes the rapid and efficient conduction of electrical impulses. In humans, myelination progresses rapidly from 1 to 2 months prior to birth through the first 1 to 2 years of life, but it also continues through adolescence and into adulthood (Levitt, 2003; Paus, Collins, et al., 2001; Yakovlev and Lecours, 1967). This timing is similar to the developmental timing of dendritic elaboration and synapse formation.

The survival of cells and synapses requires their ongoing neural activity, suggesting that external stimuli and environmental conditions, including relative deprivation, can have important long-term influences on brain development. These influences have been demonstrated in animal models, from rodents to nonhuman primates (Sanchez, Ladd, and Plotsky, 2001). Their demonstration in humans has been more indirect. It includes evidence that differences in cognitive and psychosocial stimulation are associated with modest differences in cognitive development (Gottlieb and Blair, 2004; Santos, Assis, et al., 2008; Walker, Wachs, et al., 2007), and that the more severe environmental deprivation that occurs with institutionalized infants reduces head size and overall physical growth and impairs emotional and social responsiveness, attentional abilities, and cognitive development (Smyke, Koga, et al., 2007).

Pathological synaptic pruning in particular may contribute to the genesis of at least some MEB disorders, although in the absence of direct longitudinal data, this hypothesis has not yet been confirmed (Levitt, 2003; Rakic, 2002). Disturbances in synaptic pruning that occur during adolescence are hypothesized to underlie many of the anatomical and functional disturbances seen in brain imaging of persons with schizophrenia (Lewis and Levitt, 2002; McGlashan and Hoffman, 2000). Longitudinal studies have reported exaggerated rates of cortical thinning in the dorsal prefrontal, parietal, and temporal cortices compared with healthy developing controls (Mathalon, Sullivan, et al., 2001; Thompson, Vidal, et al., 2001). Nevertheless, the cellular bases for this cortical thinning, as well as the mechanism whereby exaggerated cortical thinning would produce psychotic symptoms, are unknown.

Continuing Development and Mechanisms of Change

As noted, many developmental processes in the brain continue into childhood, adolescence, and young adulthood. This appears to be true of the frontal lobe in particular. In fact, several large human imaging studies have reported a progressive reduction in the thickness or volume of gray matter (regions containing neuronal cell bodies) in the cerebral cortex that begins in childhood and continues through young adulthood, particularly in areas of the frontal and parietal cortices (Giedd, Blumenthal, et al., 1999; Sowell, Peterson, et al., 2003). These are higher cortical areas that contribute to attentional processes and the regulation of thought and behavior. The decline in cortical gray matter may represent a synaptic pruning in adolescence and young adulthood that could produce more efficient processing in the neural pathways that support improvements in these cognitive processes, which constitute a vitally important feature of adolescent development.

The brain is subject to continual change even after its fundamental architecture and functional circuitry have been established, as evidenced by the capacity to learn new skills and establish new memories throughout life. Changes in brain structure in response to experience, learning, various physiological processes, and pharmacological or environmental agents are known as neural plasticity. Although the molecular mechanisms underlying neural plasticity are not fully understood, experience is known to induce anatomical changes across all levels of the nervous system, from molecular and cellular processes to entire neural pathways.

Such changes in brain structure begin with changes in the architecture of the synapse. Experience in the short term produces transient changes in the strength of communication across synaptic connections primarily by changing the availability of neurotransmitters and other signaling molecules. Experience in the longer term produces changes in synaptic activity, which can influence signaling pathways to regulate the function of receptors and other proteins or to change the number of receptors at the synapse. In addition, ongoing synaptic activity induces changes in gene expression that alter the production of proteins either to build up new synapses or to break down existing ones (Purves, Augustine, et al., 2000). The molecular pathways that alter gene expression and modify synaptic architecture have been studied most extensively in brain regions that subserve learning and memory, especially the hippocampus and the cerebellum. Whether and how these molecular pathways produce changes in the strength of synapses that encode other complex behaviors are not yet known.

In addition to these neuroplastic changes at the level of individual synapses, the brain is plastic at the level of cortical organization. Studies in monkeys have demonstrated that when a digit is amputated, the amount of tissue in the brain that controls movement and sensation changes over a period of weeks, so that the areas representing the remaining digits, which continue to receive sensory input, expand to take over the regions previously occupied by the missing digit (Merzenich, Nelson, et al., 1984; Purves, Augustine, et al., 2000). Similarly, if a monkey is trained to use a digit disproportionately to accomplish a task, the representation of that digit in the motor cortex expands to take over areas previously mapped to neighboring digits (Jenkins, Merzenich, et al., 1990; Purves, Augustine, et al., 2000). In addition, new connections in the cortex are generated when monkeys learn a new skill, such as using a tool, or after localized brain damage (Dancause, Barbay, et al., 2005; Hihara, Notoya, et al., 2006; Johansen-Berg, 2007). Similarly, the learning of new skills in humans leads to changes in the cortical regions that subserve that task (Doyon and Benali, 2005; Ungerleider, Doyon, and Karni, 2002).

One emerging question in the study of neural plasticity is the role that newly generated neurons may have in the postnatal brain. Mature, differentiated neurons have generally lost the capacity to divide to produce new cells, and a central dogma in neuroscience for most of the past century has been that all proliferation of new neurons ends during fetal life. However, many studies have recently provided indisputable evidence that postnatal production of new neurons, or neurogenesis, does in fact occur, even in adult life, in a small number of brain regions and in a large range of species (Gould, 2007). These neurons are generated from a population of neural stem cells that are retained in the brain. Although the full range of triggers for neurogenesis has yet to be identified, it appears to include a broad array of stimuli from experience and the environment, including physical activity and even antidepressant medications (Lledo, Alonso, and Grubb, 2006). The birth of new neurons in postnatal life is one of many means through which experience can modify anatomical circuitry and functional activity in the brain. The number of new neurons generated is small, however, and whether and to what extent these neurons are able to integrate into synaptic circuits and exert a significant functional influence in the brain are at present unclear (Ghashghaei, Lai, and Anton, 2007; Gould, 2007; Lledo, Alonso, and Grubb, 2006).

The ongoing capacity for change in the brain underlies potential mechanisms through which brain function can compensate for, or even recover from, a disorder, whether that disorder derives primarily from adverse genetic or environmental influences or a combination of both. In a broad sense, then, virtually all responses in the brain that help compensate for the presence of a disorder can be considered neuroplastic responses, and they are likely to have their structural basis in the remodeling of synaptic connections and neural systems in the brain. Moreover, the causes of certain MEB disorders are thought to involve the exaggeration or “hijacking” of certain learning and memory processes. This is thought to be a prominent feature of the pathogenesis of addictive disorders, for example, in which substances of abuse pharmacologically induce plasticity in brain circuits that are involved in reward and associative learning. This exaggerated plasticity helps establish new, abnormal stimulus–response associations among the substance, the cues that accompany it, and the behavioral responses to those cues that define disorders of addiction (Kalivas and O’Brien, 2008; Kauer and Malenka, 2007).

Sensitive Periods in Brain Development

Environmental influences that affect specific developmental processes have maximal effects during the developmental stages when those processes are under way. These developmental time periods, referred to as either “critical” or “sensitive” periods, thus constitute a window of influence for experience that is crucially important for proper brain development or for vulnerability of the developing brain to pathogenic influences from the environment. Perhaps the paradigmatic example of this point is the effect of monocular occlusion, in which one eye is sutured closed and prevented from receiving any sensory input. In adult animals, monocular occlusion produces no effect on vision or on brain structure and function. When imposed early in development, however, it permanently alters both: It impairs vision in that eye, it reduces cortical representation of the sutured eye, and it expands cortical representation of the open eye. Binocular occlusion produces perhaps even more extraordinary reorganization of the brain during an early critical period, as neurons in the would-be visual area respond not to light or visual stimuli, but to auditory and somatosensory stimuli instead (Purves, Augustine, et al., 2000; Wiesel, 1982).

Sensitive periods in humans are most clearly identified for disturbances in development of gross sensory and motor functions. For example, problems that create an imbalance in the activity of the two eyes early in life can have a permanent effect on the function of the cortical visual system. Failure to correct congenital cataracts by about age 4 months in human infants produces irreversible impairments in the visual system (Purves, Augustine, et al., 2000). Similarly, correction of strabismus, a misalignment of eye orientation, by age 7 produces optimal prevention of permanent visual impairment (Flynn, Schiffman, et al., 1998), possibly because synaptic elimination in the visual cortex is complete by that time.

Evidence in humans for the existence of sensitive periods when exposure to specific environmental and experiential influences confers enhanced vulnerability to the development of MEB disorders is thus far modest and largely circumstantial. The effects on cognitive development of environmental deprivation and separation from human caregivers may be more severe during early development (Nelson, Zeanah, et al., 2007), an observation consistent with the effects of early separation that have been documented in nonhuman primates (O’Connor and Cameron, 2006; Sabatini, Ebert, et al., 2007). Furthermore, traumatic experiences in childhood and adolescence appear to predispose to the development of severe character pathologies in adulthood; these effects are distinct from the effects of trauma experienced later in life (Bierer, Yehuda, et al., 2003; Golier, Yehuda, et al., 2003; Goodman, New, and Siever, 2004). These effects of childhood maltreatment in humans are consistent with animal models of child abuse and neglect that suggest that early maltreatment alters emotional responses and behaviors in adulthood while supporting learned preferences that are necessary for attachment to abusive caregivers (Moriceau and Sullivan, 2006; Roth and Sullivan, 2005; Sevelinges, Moriceau, et al., 2007). Additional evidence for sensitive periods in humans comes from studies reporting that prenatal but not postnatal exposure to tobacco smoke increases the risk of attention disorders in school-age children (Braun, Kahn, et al., 2006). The neural bases for the effects of early experience on higher-order neurodevelopmental outcomes in humans and in animal models are thus far largely unknown.


Developmental processes early in brain development establish fundamental brain structure and circuitry. To achieve the complex functions of the brain, signaling circuits that serve similar functions are grouped and integrated in networks both within the cortex and between the cortex and other regions of the brain. These neural systems subserve complex processes, such as learning and memory, attachment, social relatedness, and self-regulatory control. These behaviors underlie the cognitive and social competence that is an essential part of healthy emotional and behavioral development, and deficits in these systems play a role in many MEB disorders.

Learning and Memory

Multiple systems for learning and memory exist in the brain. Working memory, for example, is the “scratch pad” where information is retained for conscious manipulation (D’Esposito, 2007). Declarative memory, in contrast, is the conscious recall of facts, prior experiences, and semantic knowledge that is rapidly acquired and then consolidated for storage as long-term memory (Kandel, 2001; Purves, Augustine, et al., 2000). The hippocampus, working within networks with cortical regions, is important for remembering spatial and temporal relationships and for associative learning processes. It is centrally important for conscious learning and memory, contributing significantly to overall intellectual capacity (Amat, Bansal, et al., 2008; Atallah, Frank, and O’Reilly, 2004; Eichenbaum, 2000; Moser and Moser, 1998). A form of memory that often stands in starkest contrast to declarative memory is the incremental learning and memory of motor skills, procedures, and habits, which collectively is termed “procedural,” “habit,” or “stimulus-response” (S-R) learning. S-R learning relies on a neural system that is distinct anatomically and functionally from the hippocampus-based declarative memory system and includes the striatum, a portion of the basal ganglia deep within the brain (Packard and Knowlton, 2002). Changes in activity of dopaminergic neurons within the striatum also support learning in response to reward. Reward is an essential component of many learning processes, and it is thought to be involved in both declarative and S-R learning (Adcock, Thangavel, et al., 2006; Shohamy, Myers, et al., 2008).

Emotional experiences have powerful influences on memory, particularly on the accuracy and emotional tone of recalled memories in the declarative memory system. Emotional learning depends heavily on the interactions of the amygdala with the physically adjacent hippocampus, as well as with more remote structures that include the striatum and the frontal cortex. The interaction of the amygdala with memory systems imbues memories with the emotional tone experienced during and following the recalled event (McGaugh, 2004). Experimental emulation and manipulation of various emotions in animal models have shown that the interactions between the amygdala and the hippocampus are influenced heavily by the actions of various neurotransmitters and hormones that mediate the effects of emotional experience on the recall of arousing, rewarding, and stressful life events (McGaugh, 2004; Roozendaal, Okuda, et al., 2006).

In addition to declarative, S-R, and working memory systems, the brain supports associative or conditioned learning, as originally described by Pavlov. This form of learning involves the pairing of a stimulus that does not produce an innate behavioral response (the to-be “conditioned stimulus” or “CS,” such as a tone) with a stimulus that does produce an innate behavioral response (the “unconditioned stimulus” or “US,” such as a food odor that produces salivation). After repeated pairings of the CS and the US, the CS alone will elicit the unconditioned response (salivation). Conditioned learning involves numerous brain regions, including the hippocampus and the cerebellum (Thompson, 2005; Daum, Schugens, et al., 1993; Logan and Grafton, 1995).

The obverse of conditioned learning is extinction, in which the unconditioned response to the CS is modulated downward over time. Extinction involves exposing an animal repeatedly to a stimulus that has been previously conditioned to elicit fear, but now in the absence of any aversive event. This will extinguish the fearful, conditioned response. Extinction is therefore an active process and not simply a passive, dissipating process of forgetting (Myers and Davis, 2007; Quirk and Mueller, 2008). Extinction is cue-specific, in that extinction to one CS does not induce or accompany extinction to another CS (Myers and Davis, 2007). When extinction fails, as it can during times of stress, the conditioned behavior can reappear (Akirav and Maroun, 2007). The neural basis of fear extinction is thought to include the amygdala, the hippocampus, and the medial prefrontal cortex (Myers and Davis, 2007; Quirk and Mueller, 2008).

Disturbances in one or more of these various learning and memory systems have been implicated in the pathogenesis of a wide range of disorders. This may not be surprising if the brain is viewed as having been constructed quintessentially for the processes of learning and remembering in order to enhance adaptation and survival efficacy. The diverse and spatially distributed neural systems subserving a great variety of learning and memory systems can give rise to equally numerous and diverse illnesses.

For example, attention deficit hyperactivity disorder (ADHD) has been conceptualized as a disturbance in emotional and reward-based learning, given the difficulty that children with ADHD have learning from prior mistakes, as well as their poor performance on delay aversion tasks, their preferences for smaller immediate rewards over larger delayed ones, and their more frequent risk-taking behaviors (Farmer and Peterson, 1995; Oosterlaan and Sergeant, 1998; Sonuga-Barke, Taylor, et al., 1992). Localized reductions in volumes of the amygdala have been reported in ADHD, primarily over the basolateral nuclear complex (Plessen et al., 2006). Structural disturbances in the basolateral complex may disrupt emotional learning and the affective drive to sustain attention to otherwise mundane sensory stimuli (Cardinal, Parkinson, et al., 2002; Holland and Gallagher, 1999). The basolateral complex is densely connected with the inferior pre-frontal cortex (Baxter and Murray, 2002), another region in which reduced volumes have been reported in youth with ADHD (Sowell, Thompson, et al., 2003). Limbic-prefrontal circuits support the ability to tolerate delayed rewards and to suppress unwanted behaviors (Elliott, Dolan, and Frith, 2000), areas of difficulty that are defining hallmarks of ADHD (Barkley, Cook, et al., 2002; Rowland, Lesesne, and Abramowitz, 2002).

Disturbances in the extinction of conditioned fear responses have been postulated in the pathogenesis of a wide range of anxiety disorders. For example, fear is a normative response following exposure to trauma, and in most individuals it soon extinguishes completely. In a minority of individuals, however, fear will fail to extinguish, and they subsequently manifest symptoms of posttraumatic stress disorder (PTSD) (Yehuda, Flory, et al., 2006). Consequently, PTSD has been conceptualized as a disturbance of insufficient inhibitory control over conditioned fear responses (Liberzon and Sripada, 2008; Yehuda et al., 2006). Human imaging studies of PTSD patients have reported (1) exaggerated amygdala responses to a variety of emotional stimuli, presumably representing exaggerated fear responses; (2) deficient activation of frontal cortices, which is thought to mediate disordered fear extinction and impaired suppression of attention to trauma-related stimuli; and (3) reduced volumes and deficient activation of the hippocampus, which may mediate deficits in recognizing safe contexts (Bremner, Elzinga, et al., 2008; Rauch, Shin, and Phelps, 2006). Similar circuit-based disturbances have been postulated in other pediatric anxiety disorders, and they are thought to account for the minority of children whose anxiety disorders do not remit by adulthood (Pine, 2007). Preclinical and clinical studies have suggested that cognition-enhancing medications and repetitive exposure-based interventions, either alone or in combination, may offer a paradigm shift in anxiety disorders. Instead of treating the symptoms of anxiety pharmacologically, this strategy attempts to improve the extinction learning that occurs during cognitive-behavioral therapy (Myers and Davis, 2007; Quirk and Mueller, 2008).


Early bonding to a primary caregiver is an innate predisposition for children. It is an important feature of infant development that contributes to social and emotional learning, as well as to resilience and risk for psychopathology (Bakermans-Kranenburg and van Ijzendoorn, 2007; Corbin, 2007; Swain, Lorberbaum, et al., 2007). The classic model for early attachment is visual imprinting in newly hatched chicks. During a specific sensitive period, they develop an enduring selectivity for following either their mother or a replacement object. This imprinting consists of three independent behavioral processes: approaching the mother, learning and remembering her identity, and avoiding others while maintaining an affiliation with her. Specific cortical brain regions and synaptic changes are involved in the memory of and response to the imprinted object in chicks (Insel and Young, 2001).

Mammalian animal models of the attachment of an infant to a care-giver, as well as the behavioral and neuroendocrine responses to separation from that caregiver, have revealed physiological mediators of attachment and separation responses that have specific and long-term regulatory effects on the hormonal, physiological, and behavioral reactivity of the infant (Hofer, 1994). The interactions of the parent and child that are involved in attachment and separation responses include tactile sensation, motor activity, the warmth and temperature of the mother’s body, and nutritional factors (Hofer, 1994, 1996). The cry of the infant upon separation, for example, is released by loss of the warmth, specific odors, and passive tactile cues of the mother (Shair, Brunelli, et al., 2003). Nutritional and tactile factors also regulate hormone release and thereby cause abnormal levels of stress-response hormones during separation. Loss of the maternal nutrient supply affects hormone production by the adrenal gland, whereas loss of the tactile interaction between mother and infant affects hormone release by the pituitary gland (Hofer, 1996). These physiological regulators constitute the building blocks from which attachment develops.

Infants attach regardless of the quality of care provided by the object of attachment. During the imprinting-sensitive period, for example, chicks will follow their mother even while being shocked. Similarly, rat pups attach strongly even to a handler providing a shock or rough treatment, and infant monkeys will attach to abusive mothers (Moriceau and Sullivan, 2005). Indeed, human children develop strong attachment to a primary caregiver even when that individual subjects them to extreme abuse and neglect. Attachment studies of infant development have revealed that pathological caregiving manifests not as an absence of attachment, but instead as a disordered pattern of attachment that can be either of an anxious, insecure, or disorganized type, standing in contrast to the secure type of attachment that is the product of sensitive and protective caregiving and provides a necessary foundation for healthy emotional development (Bakermans-Kranenburg and van Ijzendoorn, 2007; Swain, Lorberbaum, et al., 2007).

Nonhuman primate models have demonstrated the importance of early attachment experiences in the development of subsequent attachment behaviors, social relatedness, and emotional regulation (O’Connor and Cameron, 2006; Pryce, Dettling, et al., 2004; Sabatini, Ebert, et al., 2007). Early, but not late, separation from a maternal caregiver, for example, has been shown to impair behaviors that promote effective socialization and to increase anxiety-related behaviors in social situations in adulthood (O’Connor and Cameron, 2006). Human evidence likewise suggests that the disruption of caregiving and social bonding early in life can exert dramatic, lifelong disruptive effects on the social competence and mental health of children. Dramatic reductions in the interactions of infants with caregivers, as can occur in extreme examples of institutionalized and socially deprived infants, can produce long-term impairments in emotional and social responsiveness and in attentional and intellectual capacities (Gunnar, 2001; Gunnar, Morison, et al., 2001; O’Connor, Marvin, et al., 2003; Rutter, Kreppner, and O’Connor, 2001; Smyke, Koga, et al., 2007). That the levels of disturbance in social behavioral and emotional regulation are dramatically greater following an earlier disruption of social bonds suggests that attachment to caregivers may be subject to a sensitive period early in postnatal development and that early deprivation may lead to subsequent social and emotional disturbances in a dose-dependent manner (Nelson, Zeanah, et al., 2007; O’Connor, Marvin, et al., 2003; Smyke, Dumitrescu, and Zeanah, 2002). In this context, it may be noted that the pairing of a separated infant with a very attentive adult can reverse the behavioral effects of early disruption in social bonds, but only when instituted early in life (O’Connor and Cameron, 2006; Cameron, 2007). This finding suggests that, for human infants, appropriate surrogate parents and foster care may have the potential to attenuate significantly the long-term effects of seriously deficient early parenting.

Although most MEB disorders involve the ability to develop and maintain healthy relationships, several disorders appear to arise from a primary disturbance of attachment. An example is borderline personality disorder, whose pathogenesis is thought to be closely linked to disturbances in early relationships, often involving either abuse and neglect or an inconsistency in parental nurturance (Fruzzetti, Shenk, and Hoffman, 2005; Johnson, Cohen, et al., 2006; Lieb, Zanarini, et al., 2004).

Perhaps the human condition that most obviously represents a disturbance in the formation of interpersonal attachments is reactive attachment disorder, which typically is manifested as an excessively inhibited or hypervigilant response to social interaction or, at the other extreme, as an excessively diffuse and indiscriminate sociability. Although its neurobiological underpinnings are not well understood, it is thought to be caused by a persistent disregard of the child’s basic emotional or physical needs or by repeated changes in the primary caregiver, which prevent formation of stable attachments during early development (Corbin, 2007).

Social Relatedness

Social relatedness is a complex construct that includes, among other components, the processing of sensory aspects of social stimuli, imitation and perspective taking, emotions induced by social interactions, and awareness of self and others. Distinct neural systems are likely to subserve each of these components.

Extensive evidence from human imaging studies suggests that the neural systems responsible for processing social stimuli are based primarily in the superior temporal cortex (Zilbovicius, Meresse, et al., 2006; Zahn, Moll, et al., 2007). A large body of recent work suggests that a “mirror neuron” system subserves knowledge of imitation, thought to be a precursor skill for the acquisition of knowledge of the intentional states that underlie the actions of others, although this evidence is not conclusive (Agnew, Bhakoo, and Puri, 2007; Iacoboni and Dapretto, 2006; Iriki, 2006; Lyons, Santos, and Keil, 2006; Rizzolatti and Craighero, 2004). Processing the sensory and conceptual aspects of social stimuli in the superior temporal cortex and understanding the actions of others through activity in the mirror neuron system are likely to work in concert with the medial prefrontal cortex to gain an understanding of one’s own and others’ intentional states. This understanding is referred to as having a “theory of mind” or the ability to “mentalize”—the knowledge that others have perspectives, beliefs, desires, and motivations that are different from one’s own.

Social relationships are an essential component of human mental health. Almost all forms of psychopathology involve difficulties in developing and maintaining healthy relationships. A primary example is autism, which is defined by the presence of qualitative deficits in social interaction and affiliation. Each of the systems that subserve the various aspects of social relatedness has been implicated in the pathogenesis of the socialization deficits in autistic children. For example, reductions in gray matter volume, reduced activation during the presentation of social stimuli, and reduced resting blood flow have all been reported in the superior temporal sulcus in individuals with autism (Gendry Meresse, Zilbovicius, et al., 2005; Gervais, Belin, et al., 2004; Zilbovicius, Boddaert, et al., 2000; Zilbovicius, Meresse, et al., 2006). In addition, several functional magnetic resonance imaging studies have implicated dysfunction of the mirror neuron system in persons with autism (Dapretto et al., 2006; Williams, Waiter, et al., 2006).

Self-Regulatory Control

Self-regulatory control is the capacity to weigh prospects for short-term gain from an action against its potential, more remote adverse consequences and to monitor and update the action plan as it unfolds. Broad expanses of the cortex and subcortex subserve the functions of self-regulatory control (Leung, Skudlarski, et al., 2000; Peterson, Skudlarski, et al., 1999; Peterson, Staib, et al., 2001). Both children and adults engage frontostriatal circuits to perform tasks that require self-regulatory control, but they do so progressively more with increasing age. Thus increasing activity of these systems during development is likely to be responsible for the superior performance of adolescents and adults compared with children on tasks that require self-regulatory control (Marsh, Zhu, et al., 2006).

Regulatory control involves control not only of actions, but also of emotions. Reassigning emotional labels to emotion-provoking stimuli, such as emotional faces and scenes, can alter the perceived pleasantness and arousal that the stimuli produce. Known as cognitive reappraisal, this reassignment produces activation of the lateral prefrontal, dorsomedial prefrontal, anterior cingulate, and occipital cortices. Activation of the ventral prefrontal cortex correlates inversely with activity in the amygdala, suggesting that cognitive reappraisal activates the frontal cortex and that the frontal cortex in turn modulates emotion-processing activity in the amygdala (Ochsner, Bunge, et al., 2002). Successful voluntary suppression of the unpleasant emotions activates similar circuits in direct proportion to the intensity of those emotions (Phan, Fitzgerald, et al., 2005). The circuits that cognitive reappraisal and emotional regulation engage are remarkably similar to the circuits activated by other, more purely cognitive, tasks that require self-regulatory control (Ochsner and Gross, 2005).

Maturation of self-regulatory functions largely defines human development, and the self-regulatory circuits that have been identified in normal individuals have been implicated in the pathogenesis of a wide range of neuropsychiatric illnesses. In fact, the capacity for self-regulatory control is one of the strongest predictors of outcome in longitudinal studies of psychopathology in children (Masten, 2004, 2007). Disturbances in these circuits are unlikely to cause disorders in and of themselves. Instead, they are likely to act in concert with underlying disturbances in other neural circuits that subserve important neuropsychiatric functions, such as motor planning and execution, mood and affect, or attention. The combination of disturbances in these latter circuits with dysfunction in self-regulatory systems may then transform a vulnerability or predisposition for developing an illness into the manifestation of symptoms and functional impairments that constitute an overt disorder. Age-specific vulnerabilities in the maturation of varying components of the neural circuits that mediate these self-regulatory functions are likely to contribute to the differences in age-specific prevalence and characteristic ages of onset of the various disorders described in Chapter 4.

ADHD is a prototypical example of a disorder of self-regulatory control. The largest anatomical studies have suggested that overall brain size is approximately 3 percent smaller in children with ADHD than in healthy children (Castellanos et al., 2002), an abnormality that probably derives from a disproportionate reduction in volume of the inferior prefrontal and anterior temporal cortices bilaterally (Sowell, Thompson, et al., 2003). These anatomical disturbances are consistent with the self-regulatory deficits that manifest as the hyperactivity, distractibility, and impulsivity of children with ADHD. Additional anatomical and functional disturbances involve the basal ganglia, the subcortical portions of the frontostriatal circuits that subserve self-regulatory control (Plessen and Peterson, 2008; Shafritz, Marchione, et al., 2004; Vaidya et al., 1998). Anatomical and functional disturbances in these regulatory control systems, though in different portions and subsystems than in ADHD, have also been reported in bipolar disorder (Blumberg, Leung, et al., 2003;Blumberg, Martin, et al., 2003), Tourette syndrome (Marsh, Zhu, et al., 2007; Peterson et al., 2001), obsessive compulsive disorder (Rosenberg and Keshavan, 1998), and eating disorders (Marsh, Gerber, et al., 2009).

Cognitive reappraisal already is a prominent component of the cognitive-behavioral therapies commonly used in the treatment of depression and anxiety disorders. Self-regulatory control tasks are being developed to treat various forms of psychopathology, including tic disorders and ADHD (Posner, 2005; Rueda, Rothbart, et al., 2005; Tang, Ma, et al., 2007; Woods, Himle, et al., 2008). Whether these interventions hold promise as prevention strategies is unknown.

Compensatory and Neuromodulatory Systems

Compensatory responses are attempts to correct for disturbances elsewhere in a biological system and to reestablish a biological balance, known as homeostasis. The quintessential purpose of the brain is to strive to achieve and to maintain homeostasis, both in its internal operations and in the external environment. The brain is likely to attempt to achieve homeostasis in the presence of a mental, emotional, or behavioral disorder by engaging neural systems that help compensate for the functional impairment due to the disorder.

Indeed, findings from human brain imaging studies have increasingly suggested that many differences previously documented in disorders may not represent a primary dysfunction but compensatory responses to the presence of neural dysfunction elsewhere. For example, although longitudinal studies suggest that most cortical abnormalities in children with ADHD represent a maturational delay, some of the differences compared with healthy control children appear to represent a compensatory response. In one study, the right parietal cortex was initially thinner in children with ADHD, similar to most other cortical regions, but then normalized over time only in those with favorable clinical outcomes. These findings suggest that the relative thickening of the right parietal cortex represents a compensatory response (Shaw, Lerch, et al., 2006). In addition, in a different sample of youth with ADHD, the head of the hippocampus was found to be enlarged, with the degree of enlargement being inversely proportional to the severity of the ADHD symptoms, suggesting that the relative hypertrophy of this structure also represents a compensatory response (Plessen, Bansal, et al., 2006). This interpretation has added plausibility in light of the connections of the hippocampus with frontal and parietal cortices and the fact that neurons and synapses in the head of the hippocampus increase in number and size in response to experiential demand (Bruel-Jungerman, Davis, et al., 2006; Cameron and McKay, 2001; Christie and Cameron, 2006; Eriksson, Perfilieva, et al., 1998; Kempermann, Kuhn, and Gage, 1997; van Praag, Shubert, et al., 2005).

Evidence for brain-based compensatory responses is perhaps strongest in children with Tourette syndrome (TS) (Spessot, Plessen, and Peterson, 2004). The dorsal prefrontal and parietal cortices of children with TS have larger volume in inverse proportion to the severity of their tic symptoms, suggesting that the hypertrophy is a compensatory response to the presence of tics (Peterson, Staib, et al., 2001). This hypertrophy is likely to be a consequence of the need to suppress tic symptoms frequently in social settings, which has been shown to produce massive activation of the pre-frontal, anterior temporal, and parietal cortices (Peterson, Skudlarski, et al., 1998). The hypertrophy increases inhibitory reserve for the self-regulatory functions that these regions subserve, so that children with TS perform normally and activate frontal tissues similarly to healthy controls. Adults with TS appear to fail to generate this compensatory frontal hypertrophy; as a result, they have more severe symptoms and require greater activation of frontal cortices to maintain adequate performance on tasks that require self-regulatory control (Marsh, Zhu, et al., 2007).

Hormonal Influences on Brain Development and Behavior

Differences between the sexes have been observed across multiple domains of cognitive, emotional, and behavioral development. Boys, for example, appear on average to be predisposed to more physical activity; less tolerance for frustration; and more aggression, impulsivity, and dys-regulated emotions (Eaton and Enns, 1986; Else-Quest, Hyde, et al., 2006; Zahn-Waxler, Shirtcliff, and Marceau, 2008). Girls on average exhibit more rapid language acquisition, greater empathy and social skills, and more fearfulness and anxiety (Else-Quest, Hyde, et al., 2006; Zahn-Waxler, Shirtcliff, and Marceau, 2008).

Several processes, ranging from differences in environmental exposures to innate differences in the biological processes that underlie either emotion and behavior or responses to the environment, could produce these gender differences (Zahn-Waxler, Shirtcliff, and Marceau, 2008). The differences are thought to have their basis at least in part in differences in brain structure and function, which are determined largely by the effects on brain development of both sex hormones and genes encoded on sex chromosomes (Arnold, 2004; Davies and Wilkinson, 2006; Hines, 2003). Hormone-dependent sexual differentiation of the brain is thought to be driven primarily by differences in androgen levels in fetal and early post-natal life. Production of testicular androgen in the human male fetus begins during the sixth week of gestation, producing higher testosterone levels in males than in females between weeks 8 and 24 of gestation (Knickmeyer and Baron-Cohen, 2006; Warne and Zajac, 1998). Studies in animal models demonstrate that differences between the sexes in the levels of various steroid hormones in the brain during fetal life produce sex-specific differences in neuronal proliferation, cell migration, apoptosis, dendritic branching, and the density of dendritic spines (Cooke, Hegstrom, et al., 1998). These differences between the sexes in fetal brain development in turn produce gender differences in brain form and structure that endure throughout postnatal life (Knickmeyer and Baron-Cohen, 2006; Hines, 2003). Changes in levels of steroidal hormones during puberty are then thought to lead to further modification of brain structure and function across both sexes (Romeo, 2003).

Differences between the sexes in brain structure and function are thought to underlie the well-documented gender differences in the diagnostic and age specificity of MEB disorders. For example, females overall are more likely than males to develop major depression and anxiety disorders (Pigott, 1999; Rutter, Caspi, and Moffitt, 2003; Zahn-Waxler, Shirtcliff, and Marceau, 2008), while males are more likely to develop ADHD, conduct disorder, substance abuse, tic disorders, and learning disorders (Rutter, Caspi, and Moffitt, 2003; Zahn-Waxler, Shirtcliff, and Marceau, 2008; Apter, Pauls, et al., 1993; Tallal, 1991). The age of onset of MEB disorders is generally earlier in boys than in girls, producing a male predominance of these disorders in prepubertal children. This sex-specific difference in rates of illness reverses following puberty, when the prevalence of disorders is higher in girls.


Relationship to Prevention Interventions

Developmental neuroscience provides a great deal of knowledge that will increasingly support preventive intervention approaches for MEB disorders. Knowledge is growing about the determinants of mental health in the prenatal and early postnatal periods of brain development; the importance of consistent and nurturing parental care on development of the brain; and the neural systems that support healthy attachment, socialization, adaptive learning, and self-regulation throughout infancy, childhood, and adolescence. All of this knowledge has important implications for interventions that can not only prevent MEB disorders but also actively promote positive, adaptive, prosocial behaviors and well-being. Specific opportunities to support healthy brain and behavioral development and to protect against environmental factors present themselves at distinct developmental stages, when they are most likely to have a beneficial effect.

During the prenatal period and the early years of a child’s life, neurobiological processes establish the potential for healthy development or, in the presence of various risk factors, the potential for the development of significant cognitive, emotional, and behavioral difficulties. Knowledge of these processes informs preventive approaches in a number of ways.

First, as discussed throughout this report, mental and physical health are inseparable, as are brain and physical development. Programs and interventions that support healthy pregnancy are therefore crucial. These can include efforts to ensure adequate and proper nutrition, such as requiring the fortification of foods with folic acid, a universal preventive intervention that has reduced the rates of neural tube defects in the United States by 25–30 percent (Pitkin, 2007). Similarly, reducing exposure to environmental toxins and infections during pregnancy and minimizing obstetrical complications during childbirth can have powerful effects on preventing MEB disorders (see Chapter 6).

Second, this chapter has emphasized the importance of nurturing care for healthy brain development and the lifelong adverse effects that disruptions in this care and exposure to harmful experiences early in life can have on both the development and functioning of the brain. Considerable evidence now suggests that these effects can be prevented or reduced by appropriately designed interventions if they are delivered at the proper time. Thus, for example, interventions focused on fostering the bonding and attachment of caregiver and child should begin at birth and be supported for the first several years of a child’s life. This is the aim of such approaches as home visitation and high-quality preschool, which are discussed in the following chapters. On the other hand, the brain continues to develop and retains a large capacity for plasticity throughout infancy, childhood, adolescence, and early adulthood as the neural systems that support such behaviors as attachment, socialization, learning, and self-regulation are refined to achieve healthy cognitive, emotional, and behavioral functioning. Evidence from both traditional models of learned behavior and more novel fields of investigation, such as epigenetics, suggests that environmental improvements can produce long-term changes in brain structure and function, and thus interventions applied even after the optimal sensitive periods of development can attenuate the effects of early adverse experiences.

Later developmental stages also bring developmentally specific opportunities to promote protective factors related to more mature behaviors—for example, building social relationships. Difficulties in developing and maintaining healthy relationships are an important aspect of many MEB disorders. Therefore, influencing social relationships positively and building networks of support in families, schools, and communities are among the primary aims of a wide array of prevention programs, as described in Chapters 6 and 7.

The development of the neural systems that support self-regulatory functions is important for acquiring developmentally appropriate neurocognitive skills that affect mental health and risk for MEB disorders (Blair, 2002; Fishbein, 2000; Greenberg, 2006; Pennington and Ozonoff, 1996; Rothbart and Posner, 2006; Riggs and Greenberg, 2004). Numerous studies have shown that appropriately designed and implemented interventions can improve self-regulatory control of thoughts, emotions, and behavior in people of all ages, even young children (Dowsett and Livesey, 2000; Rueda, Posner, and Rothbart, 2005), and several curricula and training programs have been designed to promote self-regulation in prevention frameworks. For example, the Promoting Alternative Thinking Strategies (PATHS) program (described in Box 6-7 in Chapter 6) has been shown to increase inhibitory control and working memory (Greenberg, 2006). Likewise, the preschool curriculum Tools of the Mind, designed to build inhibitory control, working memory, and cognitive flexibility, has been shown to improve these functions in an at-risk population (Diamond, Barnett, et al., 2007).

Another area in which research in developmental neuroscience has implications for prevention of MEB disorders is targeting the appropriate individuals for the delivery of interventions. The identification of children who are at either increased or diminished risk for developing an MEB disorder based on phenotypic characteristics, genotype, or other biological markers (such as physiological or brain imaging measures), or who have a history of environmental exposure offers the prospect for applying indicated prevention strategies. The possibility of targeting interventions based on evidence from developmental neuroscience is genuine and valid if the following criteria are met: (1) the evidence for the association between a marker or exposure and a disorder is sufficient to identify children at risk reliably, (2) a sufficiently powerful strategy for preventive intervention is identified that is relevant for the disorder and the risk factors in question, and (3) the magnitude of the risk or protection that the marker or exposure confers is sufficiently large to justify screening for the marker or exposure.

The potential use of individually identified biological information to determine risk raises important ethical concerns (Institute of Medicine, 2006a; Evans, 2007). These concerns frequently arise in the context of acquiring genetic information, and the rapid increase in genetic research related to MEB disorders has coincided with an increase in public interest and also in private-sector endeavors to provide commercially available access to individual genetic information (Couzin, 2008; Hill and Sahhar, 2006). One concern is appropriate interpretation of the available evidence to determine whether the above criteria have been met before a marker is implemented as a basis for determining individual risk. Genetic and other biological markers are often perceived to be more deterministic than other risk factors in their potential to predict future disease (Austin and Honer, 2005; Hill and Sahhar, 2006; Kendler, 2005; Institute of Medicine, 2006a). However, given the complex, multifactorial etiology of MEB disorders, single genetic variants have very limited predictive power. This is also likely to be true for physiological or brain imaging measures that are being studied in relationship to MEB disorders. Clearly and accurately communicating research findings, including both their promise and limitations, to the public, policy makers, practitioners, and researchers in related disciplines is of paramount importance.

If the evidence does support gathering individual genetic and other biological information for research studies, and especially if testing for MEB disorders becomes available outside the research environment as it has for other health conditions, important decisions must be made. These include who determines whether to test an individual, who can gain access to the test results, who counsels the individual about those results, and who can act on the information (Institute of Medicine, 2006a). On the one hand, limiting access to information about individual risk raises concerns about withholding health information. On the other hand, the availability of individualized information leads to concerns about privacy, stigmatization, and bias and could potentially have negative effects on employment and the ability to obtain adequate health, life, and disability insurance (Institute of Medicine, 2006a). To address these concerns, a broad array of social, ethical, and legal factors should ultimately contribute to decisions about how research findings are applied and how tests to gather information about individual risk are implemented. Such decisions need to incorporate a research-informed, evidence-based understanding of how practitioners, policy makers, and the public will interpret the information and how systems and individuals will make use of the information. These concerns are also important in considering how to use individually identified psychological, social, and other environmental risk factors to screen for the risk of developing MEB disorders (see also Chapter 8).

Relationship to Prevention Research

Defining the neural substrates of healthy cognitive, behavioral, and emotional development and, in particular, understanding the plasticity of such substrates in the face of environmental interventions can provide an important basis for prevention research and for identifying many promising avenues for future study.

Rich theories of the pathogenesis of MEB disorders in young people can be developed using animal models and other methods of basic science research, as well as neurobiological studies in humans. Accordingly, theories derived from developmental neuroscience should have a prominent role in informing the design of such interventions. Research that further identifies how environmental factors affect basic neurodevelopmental processes, such as neuronal migration, synaptogenesis, synaptic pruning, and myelination, may reveal potential new targets for preventive interventions. These targets might range from more specific reduction of exposures to potential pharmacological approaches that can enhance neurobiological processes or attenuate some of the deleterious effects of adverse environmental exposures. Similarly, a greater understanding of the functional activity in neural systems that subserve emotion and behavior might aid in developing improved cognitive training strategies that can protect against the development of disorder by enhancing regulatory or compensatory systems capable of reducing the risk for psychopathology.

Strategies to alter the genome are not a near-term prospect. However, identifying genetic variants that are associated with disorders and understanding the underlying molecular mechanisms may lead to prevention strategies based on correcting molecular disturbances in the pathways that lead from genes to behavior, including the molecular pathways that underlie the effects of known risk factors for disorders. Identifying gene–environment interactions can also suggest ways of correcting pathogenic mechanisms that can be used in new prevention strategies designed to target molecular mechanisms and bolster resilience to the effects of adverse environmental exposures.

In addition to uncovering causal mechanisms, an improved understanding of the genetic determinants of MEB disorders can provide a powerful tool for the study of environmental influences on the development of disorders. Accounting experimentally or statistically for genetic determinants allows for a much more powerful and experimentally controllable assessment of environmental determinants. Thus, genetic approaches should ultimately help to clarify which are the most potent environmental influences in the development of disorders and to prioritize possible biological targets for prevention interventions.

Epigenetics research not only provides support for preventive intervention approaches, as described in this chapter, but also can lead to novel ways of thinking about the design of new and more effective prevention strategies. For example, although the epigenetic causes of disorders are difficult to disentangle from the more traditional effects of learned behavior, growing knowledge of the epigenetically based, transgenerational transmission of maternal care and other behavioral adaptations to the environment raises the possibility that future prevention approaches targeting epigenetic mechanisms may be able to help break cross-generational cycles of such behaviors as violence and substance abuse.

In designing these new interventions, however, it is important to remember that epigenetically transmitted behavioral and emotional dispositions, including stress responsivity, are adaptive for different environmental circumstances (Fish, Shahrokh, et al., 2004). One must therefore take care to ensure that the interventions do not unwittingly produce a mismatch between the newly modified environment and the epigenetically transmitted behavior that was optimized for enhanced survival in the previous, unmodified environment. Such a mismatch could conceivably serve as a risk for pathology, adding a level of complexity to the optimal design of preventive interventions despite the best of intentions in the design and implementation of an intervention.

While theories from developmental neuroscience can inform prevention approaches, findings from prevention trials that suggest causal mechanisms should generate hypotheses that can be tested and further elaborated by basic and clinical neuroscientists using animal models and other neuroscience-based approaches. Therapeutic interventions for already-established human disorders generally offer little insight into the causes of disorders. The fact that penicillin treats pneumonia, for example, does not indicate that the pneumonia is caused by a deficiency of penicillin. As discussed in Chapters 4 and 10, prevention trials permit rigorous testing of causal mechanisms, as well as mediating and moderating effects. If designed in partnership with developmental neuroscientists, such trials therefore offer an unprecedented opportunity to evaluate the neurobiological correlates of preventive interventions by identifying and measuring the anatomical, functional, and neural systems–level effects of those interventions. Because longitudinal studies can identify environmental influences on intervention outcomes and phenotypes over the course of disorders, preventive trials also offer a context for evaluating the hypothesized mechanisms and effects of genetic factors by examining how genetic predispositions may inhibit or enhance the effects of an intervention (an example is the study described in Chapter 4 on serotonin transporter genotype in a prevention intervention trial by Brody, Kogan, et al., 2008). Because the effect sizes of interventions are often small, this kind of information should help in tailoring an intervention to specific individuals, thereby enhancing the magnitude of its beneficial effects.


Advances in neuroscience since 1994 have contributed to the growing knowledge of the determinants of mental health, the pathogenesis of disorders, and the ways in which the determinants of those disorders can be influenced through intervention strategies. Much evidence points to the central importance of brain development during the prenatal and early postnatal periods and of nurturing care for the development of the neural systems that support healthy attachment, socialization, adaptive learning, and self-regulation throughout infancy, childhood, and adolescence. The growing knowledge base in these areas has important implications in support of strategies to promote healthy cognitive, emotional, and behavioral development and to prevent MEB disorders.

Conclusion: Environment and experience have powerful effects on modifying brain structure and function at all stages of development in young people. Intervention strategies that modify environment and experience have great potential to promote healthy development of the brain and to prevent MEB disorders.

The growth of knowledge in developmental neuroscience has been particularly rapid in the defining of the roles of genetic, epigenetics, and gene–environment interactions on brain development. First, in the field of genetics, a great deal has been learned about the specific genes and molecular pathways that cause specific but fairly rare neurodevelopmental disorders. These advances have made realistic the previously remote hope that these devastating conditions might one day be treated or prevented. These advances have helped to point the way toward similar progress in understanding more common MEB disorders in children. Technological advances in large-scale, rapid-throughput genotyping have made feasible the study of the genetic vulnerabilities and underpinnings of more common disorders.

Second, advances in understanding and identifying gene–environment interactions have illuminated the ways in which specific genetic variants and life experiences both confer risk for and protect against developing MEB disorders. Third, much has been learned about the mechanisms of epigenetic modification of the genome that can confer enduring changes in gene expression and behavior. These epigenetic modifications have provided a much greater appreciation of the importance of biological adaptation of the developing organism to its environment. Bringing together knowledge in these three areas has important implications for the prospects of influencing causal biological pathways through modifications of the environment in new prevention intervention strategies (see Figure 5-3).

FIGURE 5-3. Intervention opportunities.


Intervention opportunities.

Conclusion: Genetic and other neurobiological factors contribute to the development of MEB disorders in young people, but their relative contribution is influenced by environmental factors. Similarly, the effects of environmental manipulations are constrained by genetic and other neurobiological factors.

Thus, efforts to understand the neurological basis of cognitive, emotional, and behavioral development, and especially to understand how these neural substrates can be modified through environmental intervention, are clearly an important basis for prevention research. Although research efforts are justified for intervention strategies at all stages of development in young people, developmental neuroscience has provided overwhelming evidence for the particular importance of fetal and early postnatal development for establishing the fundamental anatomical and functional architecture of the human brain that will endure throughout life, as well as evidence for the existence of sensitive periods for environmental influences in infancy. Therefore, the prenatal period and early infancy warrant a relatively high level of focus in research efforts.

Recommendation 5-1: Research funders, led by the National Institutes of Health, should dedicate more resources to formulating and testing hypotheses of the effects of genetic, environmental, and epigenetic influences on brain development across the developmental span of childhood, with a special focus on pregnancy, infancy, and early childhood.

Greater collaboration between prevention researchers and developmental neuroscientists could provide a powerful scientific synergy. Theories of pathogenesis derived from developmental neuroscience should inform the design of preventive interventions, and prevention trials should be used to inform and evaluate hypotheses of causal mechanisms derived from developmental neuroscience. Likewise, prevention trials should be designed to identify, measure, and evaluate neurobiological effects as possible mediators in preventive interventions. Hypotheses about causal mechanisms generated from prevention research should be tested and expanded using basic and clinical neuroscience approaches.

Conclusion: Collaborations among prevention scientists and basic and clinical developmental neuroscientists could strengthen understanding of disease mechanisms and improve preventive interventions by mutually informing and testing hypotheses of causal mechanisms and theories of pathogenesis.

In order to take greatest advantage of the potential for progress through collaboration, more detailed strategies to link prevention science with clinical and basic neuroscience are needed. This link needs to be supported both at the level of funding for individual investigators and also at the level of institutional infrastructure and support through funding for multi-disciplinary research consortia.

Recommendation 5-2: Research funders, led by the National Institutes of Health, should dedicate resources to support collaborations between prevention scientists and basic and clinical developmental neuroscientists. Such collaborations should include both basic science approaches and evaluations of the effects of prevention trials on neurobiological outcomes, as well as the use of animal models to identify and test causal mechanisms and theories of pathogenesis.

Recommendation 5-3: Research funders, led by the National Institutes of Health, should fund research consortia to develop multidisciplinary teams with expertise in developmental neuroscience, developmental psychopathology, and preventive intervention science to foster translational research studies leading to more effective prevention efforts.

A well-supported collaborative research approach of this kind would provide an opportunity to investigate the potential use of genotyping and other biological markers as a basis for indicated prevention strategies. This opportunity needs to be approached with appropriate attention to social, ethical, and legal issues related to the use of individually identified biological information.

Conclusion: The prospect of using genetic and other neurobiological markers to identify young people at risk of MEB disorders raises important concerns, such as potential stigma, bias, and denial of insurance coverage. However, knowingly withholding scientific knowledge from populations who can benefit from them also raises ethical issues.

Recommendation 5-4: The National Institutes of Health should lead efforts to study the feasibility and ethics of using individually identified genetic and other neurobiological risk factors to target preventive interventions for MEB disorders.