The Adolescent Brain

B.J. CASEY,a REBECCA M. JONES,a AND TODD A. HAREb

a
Sackler Institute, Weill Medical College of Cornell University, New York, New York, USA
b
California Institute of Technology, Pasadena, California, USA

Adolescence is a developmental period characterized by suboptimal decisions and actions that are
associated with an increased incidence of unintentional injuries, violence, substance abuse, un-
intended pregnancy, and sexually transmitted diseases. Traditional neurobiological and cognitive
explanations for adolescent behavior have failed to account for the nonlinear changes in behav-
ior observed during adolescence, relative to both childhood and adulthood. This review provides
a biologically plausible model of the neural mechanisms underlying these nonlinear changes in
behavior. We provide evidence from recent human brain imaging and animal studies that there
is a heightened responsiveness to incentives and socioemotional contexts during this time, when
impulse control is still relatively immature. These findings suggest differential development of
bottom-up limbic systems, implicated in incentive and emotional processing, to top-down control
systems during adolescence as compared to childhood and adulthood. This developmental pattern
may be exacerbated in those adolescents prone to emotional reactivity, increasing the likelihood
of poor outcomes.

Key words: adolescence; prefrontal cortex; nucleus accumbens; amygdala; limbic; impulsivity;
reward; development; risk taking; emotion

Introduction Adolescence is also a time of increased emotional

reactivity. During this period, the social environment
Adolescence is the period between childhood and is changing such that more time is spent with peers
adulthood encompassed by changes in physical, psy- versus adults, and more conflicts arise between the
chological, and social development (Ernst et al. 2006). adolescent and his/her parents (Csikszentmihalyi et al.
These alterations make this period a time of vulnerabil- 1977; Steinberg 1989). These changes in social interac-
ity and adjustment (Steinberg 2005). According to the tions may influence the rise of emotional reactivity. In
National Center for Health Statistics, there are over addition, given the increase in risky choices and behav-
13,000 adolescent deaths in the United States each ior during adolescence, it appears the value of positive
year. Approximately 70% of these deaths result from and negative information may be exaggerated. Greater
motor vehicle crashes, unintentional injuries, homi- emotional reactivity and sensitivity during adolescence
cide, and suicide (Eaton et al. 2006). Results from the may play a role in the higher incidence of affective dis-
2005 National Youth Risk Behavior Survey (YRBS) order onset and addiction during this developmental
show that adolescents engage in behaviors that increase period (Pine et al. 2001; Silveri et al. 2004; Steinberg
their likelihood of death or illness by driving a vehicle 2005).
after drinking or without a seat belt, carrying weapons, A number of cognitive and neurobiological hypothe-
using illegal substances, and engaging in unprotected ses have been postulated to explain why adolescents
sex resulting in unintended pregnancies and STDs, in- engage in suboptimal choice behavior. In a recent re-
cluding HIV infection (Eaton et al. 2006). These statis- view of the literature on human adolescent brain de-
tics underscore the importance of understanding risky velopment, Yurgelun-Todd (2007) suggests that cog-
choices and behavior in adolescents. nitive development during adolescence is associated
with progressively greater efficiency of cognitive con-
trol and affective modulation. An increase in activity
in the prefrontal regions as an indication of matu-
Address for correspondence: BJ Casey, Sackler Institute, Weill Cornell
ration (Rubia et al. 2000; Rubia et al. 2006; Tamm
Medical College, 1300 York Avenue, Box 140, New York, NY 10021,
Voice: +212-746-5832 fax: 212-746-5755 USA. et al. 2002) and diminished activity in irrelevant brain
[email protected] regions (Brown et al. 2005; Durston et al. 2006; Monk
Ann. N.Y. Acad. Sci. 1124: 111–126 (2008).

C 2008 New York Academy of Sciences.
doi: 10.1196/annals.1440.010 111
112 Annals of the New York Academy of Sciences

et al. 2003) are described as the neurobiological ex- opment is due to increases in processing speed and
planation for the behavioral changes associated with efficiency and not due to an increase in mental capac-
adolescence. This general pattern, of improved cogni- ity. Other theorists have included the construct of “in-
tive control and emotion regulation with maturation hibitory” processes in their account of cognitive devel-
of the prefrontal cortex, suggests a linear increase in opment (Harnishfeger & Bjorkland 1993). According
development from childhood to adulthood. to this account, immature cognition is characterized by
As evidenced by the National Center for Health susceptibility to interference from competing sources
Statistics on adolescent behavior and mortality, subop- that must be suppressed (e.g., Brainerd & Reyna 1993;
timal choices and actions observed during adolescence Dempster 1993) (Casey et al. 2002; Diamond 1985;
represent a nonlinear change in behavior, distinct from Munakata & Yerys 2001). . Thus goal-directed behav-
childhood and adulthood. If immaturity of prefrontal ior requires the control of impulses or delay of grati-
cortex were the basis for suboptimal choice behavior fication for optimization of outcomes, and this ability
and heightened emotional reactivity in adolescence, appears to mature across childhood and adolescence.
then children who have less developed prefrontal cor- On a cognitive or behavioral level, the immature
tex and cognitive abilities should look remarkably sim- cognition of adolescence is characterized as impulsive
ilar or even worse than adolescents. Thus, immature (i.e., lacking cognitive control) and risk taking, with
prefrontal function alone cannot account for adoles- these constructs used synonymously and without ap-
cent behavior. preciation for distinct developmental trajectories for
This review will provide evidence from develop- each. Human imaging and animal studies suggest dis-
mental animal and human neuroimaging studies that tinct neurobiological and developmental trajectories
may account for nonlinear changes in behavior and for the neural systems that underlie these separate con-
development during adolescence. A model of adoles- structs of impulse control and risky decisions. Specifi-
cent brain development is presented in the context of cally, a review of the literature suggests that impulsiv-
risk factors including suboptimal decision making and ity diminishes with age across childhood and adoles-
heightened emotional reactivity. cence (Casey et al. 2005; Casey et al. 2002; Galvan
et al. 2007) and is associated with protracted devel-
opment of the prefrontal cortex (Casey et al. 2005;
Development of Goal-directed Casey et al. 2002; Galvan et al. 2007) and is asso-
Behavior: Risk versus Impulse ciated with protracted development of the prefrontal
cortex (Casey et al. 2005). However, there are individ-
An accurate conceptualization of cognitive and neu- ual differences in the degree of impulsivity, regardless of
robiological changes during adolescence must treat age.
adolescence as a transitional developmental period In contrast to the linear increase with age associated
(Spear 2000), rather than a single snapshot in time with impulse control, risk taking appears greater dur-
(Casey et al. 2005). In other words, to understand ing adolescence relative to childhood and adulthood
this developmental period, transitions into and out of and is associated with subcortical systems known to be
adolescence are necessary for distinguishing distinct involved in evaluation of incentives and affective infor-
attributes of this stage of development. Adolescent mation. Human imaging studies that are reviewed here
behavior has been described as impulsive and risky, suggest an increase in subcortical activation (accum-
almost synonymously, yet these behaviors rely on dif- bens and amygdala) when making risky choices and
ferent cognitive and neural processes (Casey et al. in processing emotional information (Ernst et al. 2005;
press), which suggest distinct constructs with different Monk et al. 2003; Montague & Berns 2002) (Kuhnen
developmental trajectories. & Knutson 2005; Matthews et al. 2004) that is exag-
A cornerstone of cognitive development is the abil- gerated in adolescents, relative to children and adults
ity to suppress inappropriate thoughts and actions in (Ernst et al. 2005; Galvan et al. 2006).
favor of goal-directed ones, especially in the presence These findings suggest distinct neurobiological tra-
of compelling incentives (Casey et al. 2005; Casey et al. jectories for impulse versus risk taking behavior. The
2000; Casey et al. 2002). A number of classic develop- limbic subcortical systems appear to be developed by
mental studies have shown that this ability develops adolescence in contrast to control systems that show a
throughout childhood and adolescence (Case 1972; protracted and linear developmental course into young
Flavell et al. 1966; Keating & Bobbitt 1978; Pascual- adulthood. The prefrontal cortical control systems are
Leone 1970). Several theorists (e.g., Bjorkland 1985, necessary for overriding inappropriate choices and ac-
1987; Case 1985) have argued that cognitive devel- tions in favor of goal-directed ones.
Casey et al.: The Adolescent Brain 113

FIGURE 1. Illustrations of the most common magnetic resonance methods used in the study of human
development. (A) Structural magnetic resonance imaging (MRI) to produce structural images of the brain
useful for anatomical and morphometric studies, (B) diffusion tensor imaging (DTI) measures myelination
and directionality of fiber tracts between anatomical structures, and (C) functional MRI (fMRI) measures
patterns of brain activity within those structures (from Casey et al. 2005).

Animal Studies of Adolescent cortical and cortical areas. Animal studies have shown
Brain Development that dopamine is crucial for communication between
the accumbens, amygdala, and prefrontal cortex and
Until recently, much of our understanding of the that signaling between these regions relies upon the fine
adolescent brain has come from animal studies. These balance between excitatory and inhibitory dopamine
experiments have been critical for obtaining informa- transmission (Floresco & Tse 2007; Grace et al. 2007;
tion about the neurochemical and cellular changes that Jackson et al. 2001). There are significant peaks in
occur as a function of age. The validity of animal mod- dopamine expression during adolescence. Dopamine
els to study adolescence has been questioned, since it projections to the prefrontal cortex continue to develop
is argued that only humans undergo the psychologi- into early adulthood, with dopamine levels peaking in
cal stress of adolescence (e.g., Bogin 1994). However, the prefrontal cortex during adolescence versus earlier
animals including rodents and nonhuman primates ex- or later in life in nonhuman primates (Rosenberg &
hibit increased social interactions during adolescence Lewis 1994, 1995) and in rats (Kalsbeek et al. 1988).
(Primus & Kellogg 1989) as well as novelty-seeking and Dopamine receptor expression is highest in the ac-
risk-taking behaviors (Adriani et al. 1998; Spear 2000). cumbens during early adolescence (Tarazi et al. 1998).
These behavioral findings suggest that animal models These findings in rodents suggest that there are specific
are appropriate for studying neurobiological changes regions undergoing structural changes, and therefore,
during adolescence. connections and communication between subcortical
Studies in rodents have shown at the cellular level and cortical regions are in transition and in flux dur-
that there are distinct changes in limbic and pre- ing adolescence. Significant evidence suggests that the
frontal regions during adolescence. During early pu- neuroanatomical changes described above are also oc-
berty, there is an overproduction of axons and synapses, curring during adolescence in humans, but our meth-
followed by rapid pruning in later adolescence (Crews ods for studying humans only provide an approximate
et al. 2007). Specifically, there is dendritic pruning in index of such changes.
the amygdala (Zehr et al. 2006), nucleus accumbens
(Teicher et al. 1995), and prefrontal cortex (Andersen
& Teicher 2004; Andersen et al. 2000) and contin- Neuroimaging Studies of Human
ual growth in the density of the fibers connecting the Brain Development
amygdala and prefrontal cortex into early adulthood
(Cunningham et al. 2002). There is more prolonged Our current understanding of the human adoles-
pruning throughout adolescence in the prefrontal cor- cent brain has come from advances in neuroimaging
tex versus the accumbens (Andersen et al. 2000; Te- methodologies that can be used with developing hu-
icher et al. 1995). These differences in pruning in ro- man populations. These methods depend on magnetic
dents are consistent with our model suggesting that the resonance imaging (MRI) methods (see FIG. 1) and in-
accumbens matures earlier than the prefrontal cortex. clude structural MRI, which is used to measure the
Consistent with the cellular changes in animals, size and shape of structures; diffusion tensor imaging
there are alterations in neurotransmission in these sub- (DTI), which is used to index connectivity of white mat-
ter fiber tracts; and functional MRI which is used to
114 Annals of the New York Academy of Sciences

FIGURE 2. Illustration of gray matter volume maturation over the cortical surface from 5 to 20 years
of age (from Lenroot & Giedd 2006).

measure patterns of brain activity. These methods have 2004) (see FIG. 2). This pattern of change is consistent
furthered our understanding of the neurobiological ba- with nonhuman primate (Bourgeois et al. 1994) and
sis and development of reward or incentive behavior human postmortem studies (Huttenlocher 1979) indi-
relative to goal-directed behavior. cating that the prefrontal cortex is one of the last brain
regions to mature. In contrast to gray matter, white
MRI Studies of Human Brain Development matter volume increases in a roughly linear pattern
Several studies have used structural MRI to map the throughout development and into adulthood (Gogtay
developmental course of the normal brain (for review, et al. 2004). These changes most likely reflect ongoing
see Durston et al. 2001). Although the brain reaches myelination of axons by oligodendrocytes enhancing
approximately 90% of its adult size by age six, the gray neuronal conduction and communication.
and white matter subcomponents of the brain continue When examining neuroanatomical changes across
to undergo dynamic changes throughout adolescence. development, the subcortical regions are often over-
Data from recent longitudinal MRI studies indicate looked, however, it is important to note that these areas
that the change in gray matter volume over time has have some of the largest changes during development
an inverted U-shape pattern and has greater regional in the brain, particularly in the basal ganglia (Sowell et
variation than white matter (Giedd 2004; Gogtay et al. al. 1999) and specifically in males (Caviness et al. 1996;
2004, Sowell et al. 2003, 2004). In general, regions that Giedd et al. 1996; Reiss et al. 1996). Developmental
involve primary functions, such as motor and sensory changes in structural volume within basal ganglia and
systems, mature earliest compared to the higher-order prefrontal regions are interesting in light of the previ-
association areas that integrate these primary functions ously mentioned animal work showing pruning in these
(Gogtay et al. 2004; Sowell et al. 2004). MRI studies regions during adolescence. These processes allow for
show loss of cortical gray matter first in primary sen- the fine tuning and strengthening of connections be-
sorimotor areas, followed by that in the dorsolateral tween prefrontal and subcortical regions during devel-
prefrontal and lateral temporal cortices (Gogtay et al. opment and learning that may correspond to greater
Casey et al.: The Adolescent Brain 115

cognitive control. and basal ganglia are correlated with impulse control,
How do these changes in structure relate to dif- as measured by performance on a go/no-go task. The
ferences in cognition? A number of studies have re- prefrontal fiber tracts were defined by regions of in-
lated frontal lobe structural maturation and cogni- terests, which were identified in an fMRI study using
tive function using neuropsychological and cognitive the go/no-go task. In developmental DTI studies, fiber
measures (e.g., Sowell et al. 2003). Specifically, these tract measures were correlated with age, but specificity
studies showed associations between MRI-based re- of particular fiber tracts with cognitive performance
gional volumes of the prefrontal cortex and basal gan- were shown by dissociating the particular tract (Lis-
glia with measures of cognitive control (i.e., ability to ton et al. 2006) or cognitive ability (Nagy et al. 2004).
override an inappropriate response in favor of another These findings highlight the importance of examining
or to suppress attention toward irrelevant stimulus at- not only regional, but also related circuitry changes,
tribute in favor of relevant stimulus attribute) (Casey when making inferences about neural changes in cog-
et al. 1997, 1997). These findings suggest that cogni- nition across development.
tive changes are reflected in structural brain changes
and underscore the importance of subcortical (basal
ganglia) as well as cortical (e.g., prefrontal cortex) de- Functional MRI Studies of Human Brain
velopment. While these findings showed associations Development
between structure and function, a more in-depth dis- Compared to MRI and DTI, fMRI is a more direct
cussion of functional imaging evidence for changes approach for examining behavior changes during de-
in activity that more directly coincide with behavior velopment and for establishing structure–function re-
across development is presented in the fMRI section. lationships. Using fMRI to measure functional changes
in the developing brain has significant potential for the
DTI Studies of Human Brain Development field of developmental science and provides a means
The MRI-based morphometry studies previously for constraining interpretations of adolescent behavior.
reviewed suggest that during development, cortical As stated previously, the development of the pre-
connections are fine tuned via elimination of an over- frontal cortex is believed to play an important role in
abundance of synapses and by strengthening of rele- the maturation of higher cognitive abilities such as de-
vant connections, although these measures do not have cision making and cognitive control (Casey et al. 2002;
the resolution to visualize or measure synapses. Recent Casey et al. 1997; Hare & Casey 2005). Many behav-
advances in MRI technology, like DTI, provide a po- ioral paradigms, together with fMRI, have assessed
tential tool for examining the role of specific white mat- the neurobiological basis of these abilities, including
ter tracts in the development of the brain and behavior flanker, Stroop, and go/no-go tasks (Casey et al. 1997;
(for review, see Cascio et al. 2007). Examining white Casey et al. 2000; Durston et al. 2003). Collectively,
matter tracts can provide knowledge about pathways these studies show that children recruit distinct but of-
of connectivity in the brain, and presumably it is via ten larger and more diffuse prefrontal regions when
these pathways that information is able to travel from performing these tasks than do adults. The patterns of
one region of the brain to another (Cascio et al. 2007). brain activity that are important for task performance,
Relevant to this paper are the neuroimaging studies such as those regions that correlate with cognitive per-
that have linked the development of white matter fiber formance, become more fine tuned with age. Regions
tracts with improvements in cognitive ability with age. that are not correlated with task performance diminish
Recently, associations have been shown between in activity with age. This pattern has been observed
DTI-based measures of prefrontal white matter de- across both cross-sectional (Brown et al. 2005) and
velopment and cognition in children. Nagy and col- longitudinal studies (Durston et al. 2006) and across
leagues showed a positive correlation between matu- a variety of paradigms. Neuroimaging studies cannot
ration of prefrontal–parietal fiber tracts and working definitively characterize the mechanism of such devel-
memory in children (Nagy et al. 2004), which is con- opmental changes as dendritic arborization or synap-
sistent with functional neuroimaging studies showing tic pruning. However, these studies suggest that change
differential recruitment of these regions in children rel- over a period of time results in both refinement within
ative to adults. Using a similar approach, Liston and brain regions as well as fine tuning of projections from
colleagues (2006) have shown that white matter tracts these regions (Brown et al. 2005; Bunge et al. 2002;
between prefrontal–basal ganglia and posterior fiber Casey et al. 1997; Casey et al. 2002; Luna et al. 2001;
tracts continue to develop across childhood into adult- Moses et al. 2002; Schlaggar et al. 2002; Tamm et al.
hood, but only tracts between the prefrontal cortex 2002; Thomas et al. 2004; Turkeltaub et al. 2003).
116 Annals of the New York Academy of Sciences

Functional MRI Studies of Behavior

during Adolescence
The question remains how can fMRI studies help
explain whether adolescents, compared to children or
adults, are 1) lacking sufficient cognitive control (im-
pulsive), 2) risky in their choices and actions, and 3)
more sensitive to affective information when required
to exert cognitive control than children or adults.
Impulse control, as measured by cognitive control
tasks like the go/no-go task, shows a linear pattern of
development across childhood and adolescence, as de-
scribed above. However, we were interested in under-
standing changes across development in top-down con-
trol regions and subcortical reward-seeking regions. It
is only recently that risk taking in adolescents has been
examined with neuroimaging techniques (Ernst et al.
2005; May et al. 2004). These studies have focused
FIGURE 3. The traditional explanation of adolescent
primarily on the region of the accumbens, a portion behavior has been that it is due to the protracted develop-
of the basal ganglia involved in predicting reward out- ment of the prefrontal cortex. Our model takes into consid-
comes. Although two recent reports showed less ven- eration the development of the prefrontal cortex together
tral prefrontal activity (Eshel et al. 2007) and posterior with subcortical limbic regions (e.g., nucleus accumbens
mesofrontal activity (Bjork et al. 2007) in adolescents and amygdala) that have been implicated in risky choices
and emotional reactivity.
versus adults on risk-taking behavior, the goal of our
studies was to characterize the development of limbic
subcortical regions involved in motivation and emo-
tional reactivity in conjunction with top-down control et al. 2005; Galvan et al. 2007; Galvan et al. 2006;
regions (prefrontal cortex). Many studies have exam- Hare & Casey, in press). FIGURE 3 depicts this model
ined the neural response in children and adolescents illustrating how bottom-up limbic and prefrontal top-
to affective information (e.g., emotional faces) ( Baird down control regions should be considered together.
et al. 1999; Killgore et al. 2001; Monk et al. 2003; The graph shows different developmental trajectories
Thomas et al. 2001b; Yurgelun-Todd & Killgore 2006) for these systems, with limbic systems developing ear-
but typically have used passive viewing or attention lier than prefrontal control regions. According to this
tasks (Monk et al. 2003) unrelated to processing of the model, the individual is biased more by functionally
affective information. Our studies examine how affect mature limbic regions during adolescence (i.e., imbal-
influences cognitive control across development and ance of limbic relative to prefrontal control), compared
characterizes the activation of the subcortical systems to children, for whom these systems are both still devel-
(amygdala) involved in affect regulation relative to the oping, and compared to adults, for whom these systems
cortical (prefrontal) regions associated with cognitive are fully mature. This perspective provides a basis for
control. nonlinear shifts in behavior across development, due
to earlier maturation of this limbic system relative to
the less mature top-down prefrontal control region.
A Neurobiological Model Furthermore, with development and experience, the
of Adolescence functional connectivity between these regions provides
a mechanism for top-down control of these regions
How do neural changes in subcortical regions (e.g., (Hare & Casey, in press). Our model reconciles the
accumbens and amygdala) associated with reward- contradiction between health statistics of risky behav-
seeking and emotion coincide with development of ior during adolescence and the astute observation by
the prefrontal regions and do they relate to impul- Reyna and Farley (2006) that adolescents are able to
sivity and risk-taking behaviors? We have developed reason and understand risks of behaviors in which they
a neurobiological model of adolescent development engage.
within this framework that builds on rodent models According to our model, in emotionally salient situ-
(Laviola et al. 1999; Spear 2000) and recent imag- ations, the more mature limbic system will win over the
ing studies of children, adolescents, and adults (Ernst prefrontal control system. In other words, when a poor
Casey et al.: The Adolescent Brain 117

decision is made in an emotional context, the adoles-

cent may know better, but the salience of the emotional
context biases his or her behavior in opposite direction
of the optimal action.
Our neurobiological model proposes that the com-
bination of heightened responsiveness to rewards and
immaturity in behavioral control areas may bias ado-
lescents to seek immediate rather than long-term
gains, perhaps explaining their increase in risky deci-
sion making and emotional reactivity. Tracking sub-
cortical (e.g., accumbens and amygdala) and corti-
cal (e.g., prefrontal) development of decision making
and emotional reactivity across childhood and through
adulthood provides additional clarification on whether
changes reported in adolescence are specific to this pe-
riod of development or, rather, reflect maturation that
is steadily occurring in a somewhat linear pattern from
childhood to adulthood.
Two recent fMRI studies spanning from childhood
to adulthood provide empirical evidence consistent
with our neurobiological model. In the first study (Gal-
van et al. 2006), we examined behavioral and neural re-
sponses to reward manipulations across development,
focusing on brain regions implicated in reward-related
learning and behavior in animal (Hikosaka & Watan-
FIGURE 4. Magnitude and extent of accumbens and
abe 2000; Pecina et al. 2003; Schultz 2006) and adult OFC activity to reward. Adolescents (13–17 years) showed
imaging studies (e.g., Knutson et al. 2001; O’Doherty greater percent signal change to large rewards than ei-
et al. 2001; Zald et al. 2004) and in studies of addiction ther children (aged 7–11 years) or adults (23–29 years)
(Hyman & Malenka 2001; Volkow & Li 2004). Based in the accumbens (A). Children had the greatest percent
on rodent models (Laviola et al. 1999; Spear 2000) and signal change in the OFC compared to adolescents and
previous imaging work (Ernst et al. 2005), we hypoth- adults (B). Children had the greatest volume of activity
in the accumbens relative to adolescents and adults (C)
esized that relative to children and adults, adolescents
Children and adolescents showed greater volume of ac-
would show exaggerated responses to reward as in- tivity in the OFC than adults (D). Adapted from Galvan
dexed by elevated accumbens activity in concert with et al. (2006).
less mature recruitment of top-down prefrontal control
regions.
Our findings were consistent with rodent models in press; Hare et al. 2005). During the experiment,
(Laviola et al. 2003) and previous imaging studies dur- participants were presented with two emotional facial
ing adolescence (Ernst et al. 2005), which show en- expressions (fearful, neutral, or happy) and were asked
hanced accumbens activity to rewards. Adolescents, as to respond to one of the emotions (e.g., fear) and sup-
compared to children and adults, showed an exagger- press their response to the other emotion (e.g., neu-
ated accumbens response in anticipation of reward. tral). In the context of negative emotional information
However, both children and adolescents showed a less (fearful faces), reaction times improved with age but
mature response in prefrontal control regions than were longer when detecting fearful faces relative to a
adults. These findings suggest that there are different neutral or happy face. This slowing in reaction time
developmental trajectories for these regions. The en- was correlated with greater amygdala activity (Hare
hancement in accumbens activity during adolescence & Casey, in press). Activity in the orbital frontal cor-
may relate to the increase in impulsive and risky behav- tex increased with age, and greater orbital frontal ac-
iors observed during this period of development (see tivity relative to amygdala was associated with more
FIG. 4). efficiency in suppressing emotional reactivity (longer
In the second study, we examined the development reaction times and greater amygdala activity). These
of behavioral and neural responses in performance findings are in accordance with animal studies (Baxter
of an emotional go/no-go paradigm (Hare & Casey, et al. 2000) which show connectivity between the
118 Annals of the New York Academy of Sciences

FIGURE 5. Bilateral amygdala activation (left). Graph depicts amygdala activity in adults, teenagers,
and children. Adapted from Hare et al. (in press).

amgydala and orbital frontal cortex are important for Why Would the Adolescent Brain
assessing changes in emotional value of an object and Be Programmed This Way?
adapting behavior accordingly.
Differential recruitment of prefrontal and subcor- Adolescence can be described as a progressive tran-
tical regions has been reported across a number of sition from childhood into adulthood with an inde-
developmental fMRI studies (Casey et al. 2002; Monk finable ontogenetic time course (Spear 2000) yet of-
et al. 2003; Thomas et al. 2004). These findings were ten co-occurring with puberty, which is defined by
typically interpreted in terms of immature prefrontal specific biological markers. The significant neuroen-
regions rather than as an imbalance between prefrontal docrinological changes associated with puberty, such
and subcortical regional development. Given evidence as increases in adrenal and gonadal hormones, are
of prefrontal regions in guiding appropriate actions correlated with the development of secondary sexual
in different contexts (Miller & Cohen 2001), immature characteristics and can influence brain function (for a
prefrontal activity might hinder appropriate estimation review see Spear 2000). The onset and hormone fluc-
of future outcomes, especially when making a decision tuations of puberty may provide an explanation for the
within an emotional context (i.e., heat of the moment). observed functional differences in subcortical activity
This interpretation is consistent with previous research between children and adolescents, versus activity in the
showing elevated subcortical, relative to cortical, ac- prefrontal region, which reflects a linear change with
tivity when decisions are biased by immediate versus age.
long-term gains (McClure et al. 2004). Further, fMRI From an evolutionary perspective, adolescence is
studies have shown limbic subcortical activity positively the period in which independence skills are acquired
correlates with suboptimal choice behaviors (Kuhnen in order to increase the success of separating from the
& Knutson 2005). protective influence of the family. It is also a period
In sum, during adolescence, relative to childhood when there is an increase in the likelihood of harm
or adulthood, an immature ventral prefrontal cor- such as injury, depression, anxiety, drug use, and addic-
tex may not provide sufficient top-down control of tion (Kelley et al. 2004). However, our neurobiological
robustly activated reward and affect processing re- model suggests that risky behavior and emotional re-
gions (e.g., accumbens and amygdala). This imbal- activity are the products of a biologically driven imbal-
ance in development of these regions and relative ance between increased novelty and positive sensation
top-down control results in less influence of prefrontal seeking in conjunction with immature “self-regulatory
systems (orbitofrontal cortex) relative to the accum- competence” (Steinberg 2004).
bens and amygdala in reward valuation and emotional As previously mentioned, during adolescence,
reactivity. independence-seeking behaviors are prevalent across
Casey et al.: The Adolescent Brain 119

species, such as increases in peer-directed social in- and does not eat a cookie, she will receive the large
teractions and intensifications in novelty seeking and reward (2 cookies). If the child cannot wait, she should
risk-taking behaviors. In other species such as rodents, ring a bell to summon the experimenter and thereby
nonhuman primates, and birds, behaviors like seeking receive the smaller reward (1 cookie). Distractions in
out same-age peers and fighting with parents are also the room are minimized, with no toys, books, or pic-
observed and may be important adaptive skills to re- tures. The experimenter returns after 15 minutes or
move the adolescent from the home territory in order after the child has rung the bell, eaten the rewards, or
to mate (Spear 2000). Relative to adults, periadoles- shown any signs of distress. Mischel (1989) showed that
cent rats show increased novelty-seeking behaviors in children typically behave in one of two ways: 1) either
a free-choice novelty paradigm (Laviola et al. 1999). they ring the bell almost immediately in order to have
Neurochemical evidence indicates that the balance in the cookie, which means they only get one; or 2) they
the adolescent brain between cortical and subcortical wait and optimize their gains to receive both cookies.
dopamine systems begins to shift toward greater corti- This observation suggests that some individuals are
cal dopamine levels during adolescence (Spear 2000). better than others in their ability to control impulses
And as previously described, in the nonhuman primate in the presence of highly salient incentives, and this
there is an increase in dopamine enervation of the pre- bias can be detected in early childhood (Mischel et al.
frontal cortex into early adulthood (Rosenberg & Lewis 1989). This differential in impulse control appears to
1995). Thus this elevated risk-taking behavior appears remain throughout adolescence and young adulthood
to occur across species and have important adaptive (Eigsti et al. 2006).
functions. What might explain individual differences in deci-
It is possible that this developmental pattern is an sion making and behavior? Some theorists have postu-
evolutionary feature. One needs to engage in high- lated that the dopaminergic mesolimbic circuitry, im-
risk behavior in order to leave the family and village plicated in reward processing, underlies risky behavior
to find a mate. This risk behavior occurs simultane- (Blum et al. 2000), and that individual differences in
ously with an increase in sexual hormones, resulting this circuitry might relate to the propensity to engage
in adolescents seeking sexual partners and is seen in in risky behavior (O’Doherty 2004). A number of stud-
other species. In conjunction with this novelty-seeking ies have shown increases in activity in the nucleus ac-
behavior, there would need to be some mechanism for cumbens immediately prior to making risky choices on
detecting cues of safety or danger. The increase in emo- monetary-risk paradigms (Kuhnen & Knutson 2005;
tional reactivity during this period may allow adoles- Matthews et al. 2004; Montague & Berns 2002), and
cents to be more vigilant and aware of threat, to ensure as described previously, adolescents show exaggerated
their survival as they move from a safe environment to a accumbens activity to rewarding outcomes relative to
novel one. In today’s society when adolescence may ex- children or adults (Ernst et al. 2005; Galvan et al.
tend indefinitely—with individuals well into their 20s 2006). Collectively, these data suggest that as a group
living with their parents, remaining financially depen- adolescents may be more likely to engage in risky
dent, and choosing mates later in life—these behaviors choices (Gardener & Steinberg 2005). However, some
may be deemed inappropriate. adolescents will be more prone than others to engage
in risky behaviors, putting them at potentially greater
risk for negative outcomes. Therefore it is important to
Biological Predispositions, consider individual variability when examining com-
Development, and Risky Behavior plex brain–behavior relationships related to risk taking
and reward processing in developmental populations.
The recognition of individual differences in impulse To explore individual differences in risk-taking be-
control and taking risks is not new in the field of psy- havior, Galvan and colleagues (2007) recently exam-
chology (Benthin et al. 1993). Perhaps one of the classic ined the association between activity in reward-related
examples of individual differences in the social, cogni- neural circuitry in anticipation of a large monetary
tive, and developmental psychology literatures is delay reward with behavioral measures of risk taking and
of gratification (Mischel et al. 1989). Delay of gratifica- impulsivity in adolescence. Specifically, Galvan and
tion is typically assessed in 3- to 4-year-old toddlers. A colleagues used functional magnetic resonance imag-
toddler is seated in a room with two cookies and a bell. ing and anonymous self-report rating scales of risky
The child is then told that the experimenter will leave behavior, risk perception, and impulsivity in individ-
the room in order to prepare for upcoming activities uals between the ages of 7 and 29 years (see FIG. 6).
and explains to the child that if she remains in her seat There was a positive association between accumbens
120 Annals of the New York Academy of Sciences

FIGURE 6. Activity in the nucleus accumbens in anticipation of reward (A). Percent change in fMRI signal in the
accumbens in anticipation of reward as a function of age (B). The association between accumbens activity to reward and
the likelihood of engaging in risky behavior in three age groups (C) (Adapted from Galvan et al. 2007).

activity and the likelihood of engaging in risky behav- disorders (e.g., schizophrenia, depression, anxiety)
ior across development. In other words, those individ- manifest. Normal adolescent development can be in-
uals, who perceived risky behaviors as leading to dire terpreted as the coordination of emotions and behavior
consequences, activated the accumbens less to reward. in the social and intellectual environment, and the de-
Impulsivity ratings were not associated with accum- velopment of psychopathology during adolescence can
bens activity, but rather with age, further dissociating be seen as resulting from a difficulty in balancing these
impulse control from incentive-based risky behaviors. factors (Steinberg 2005). We have previously described
These findings suggest that during adolescence, some enhanced bottom-up emotional processing in subcor-
individuals have a predisposition to engage in risky tical regions relative to less effective top-down mod-
behaviors due to developmental neural changes. ulation in prefrontal regions to affective information
Adolescent behavior is repeatedly characterized as during adolescence. It is possible that this imbalance
impulsive and risky, yet this review of the imaging liter- may play a role in the increased risk for affective disor-
ature suggests different neurobiological substrates and ders during adolescence (Steinberg 2005). Clearly, not
developmental trajectories for these two types of be- all adolescents develop psychopathology; there must
havior. Specifically, impulsivity is associated with im- be individual variability in emotional reactivity and
mature ventral prefrontal development and gradually the ability to modulate these behaviors. Individual dif-
diminishes from childhood to adulthood (Casey et al. ferences may predispose a person to be at greater risk
2005). The negative correlation between impulsivity for poorer outcomes.
ratings and age in the study by Galvan and colleagues The amygdala has been implicated as a key neural
(2007) further supports this notion. In contrast, risk region in emotional dysregulation in psychiatric disor-
taking is associated with an increase in accumbens ac- ders. This region is essential to learning the emotional
tivity (Kuhnen & Knutson 2005; Matthews et al. 2004; significance of cues in the environment (see Maren &
Montague & Berns 2002) that is exaggerated in ado- Quirk 2004 for review). In animal studies, amygdala
lescents, relative to both children and adults (Ernst et lesions result in a reduction of fear behavior (Anglada-
al. 2005; Galvan et al. 2006). Thus adolescent choices Figueroa & Quirk 2005; Davis & Whalen 2001; Kalin
and behavior cannot be explained by impulsivity or et al. 2004), and human neuroimaging studies have
protracted development of the prefrontal cortex alone, shown increases in activity in the amygdala to fear-
as children would then be predicted to be greater risk ful stimuli in adults (Breiter et al. 1996; Morris et al.
takers. The findings provide a neural basis for why 1998) and in children (Thomas et al. 2001b). There is
some adolescents are at greater risk than others, but evidence for dysregulation of amygdala activity in anx-
also demonstrate a basis for why adolescent risk-taking ious and depressed children (Thomas et al. 2001a) and
behavior in general is different from risk taking in chil- adults (Leppanen 2006; Rauch et al. 2003; Thomas
dren and adults. et al. 2001a).
In a recent study, we examined individual differ-
Adolescence, Individual Differences, ences in anxiety levels as measured by the Spielberger
and Affective Disorders State Trait Anxiety Index and neural responses to af-
fective information in adolescents and adults during
Adolescence is a time of greater emotional reactiv- an emotional go/no-go task (Hare et al. in press). Ado-
ity and a period when symptoms of many psychiatric lescents showed greater initial amygdala activity than
Casey et al.: The Adolescent Brain 121

FIGURE 7. Picture depicts left orbitofrontal activity. Graph illustrates correlation of activity in the OFC and in the
amygdala in both adults and adolescents (adapted from Hare et al. in press).

adults, and sustained amygdala activity was correlated allel increases in cognitive control and self-regulation
with trait anxiety. Increased activity in orbitofrontal (Casey et al. 1997; Luna & Sweeney 2004; Luna et
regions correlated with a decrease in amygdala ac- al. 2001; Rubia et al. 2000; Steinberg 2004). These
tivity over time (i.e., repeated presentation of a fear- changes appear to show a shift in activation of pre-
ful face, see FIG. 7), suggesting dampening of emo- frontal regions from diffuse to more focal recruitment
tional reactivity due to top-down control from pre- over time (Brown et al. 2005; Bunge et al. 2002; Casey
frontal regions. These findings are consistent with et al. 1997; Durston et al. 2006; Moses et al. 2002)
animal studies showing the importance of the or- and elevated recruitment of subcortical regions during
bitofrontal cortex (OFC) in extinction of fear condi- adolescence (Casey et al. 2002; Durston et al. 2006;
tioning in animals with repeated exposure to empty Luna et al. 2001). Although neuroimaging studies can-
threat (Gallagher et al. 1999) . not definitively characterize the mechanism of such
Our results are consistent with previous fMRI stud- developmental changes, these changes in volume and
ies in clinically anxious adults and children that have structure may reflect development within, and refine-
shown unregulated amygdala activity to negative emo- ment of, projections to and from these brain regions
tional information and less activity in the prefrontal during maturation suggestive of fine tuning of the sys-
cortex (McClure et al. 2007; Shin et al. 2004; Thomas tem with development.
et al. 2001b) in adolescents at risk for anxiety disor- We have discussed the importance of considering
ders (Perez-Edgar et al. 2007). Together these find- individual variability when examining complex brain–
ings suggest that prefrontal regions serve to regulate behavior relationships related to risk taking, reward
emotional reactivity and that individual differences in processing, and emotional reactivity in developmental
emotion regulation may be due to an imbalance in ac- populations. Using an approach that looks at devel-
tivity between these regions that is exacerbated during opmental trajectories, rather than snapshots in time,
adolescence. allows one to comprehensively study these behaviors
during development and examine individual differ-
ences. It is not possible to fully explain the emergence
Conclusions of affective disorders or atypical development by sim-
ply examining one time point. Longitudinal studies
Human imaging studies show structural and func- across development would be the best methodology to
tional changes in frontolimbic regions (Jernigan et al. address these issues.
1991; Giedd et al. 1999; Giedd et al. 1996; Sowell et Taken together, the findings synthesized here indi-
al. 1999; for review, Casey et al. 2005; Jernigan et al. cate that increased risk-taking behavior and greater
1991; Giedd et al. 1999; Giedd et al. 1996; Sowell et al. emotional reactivity in adolescence are associated
1999) for review, (Casey et al. 2005) that seem to par- with different developmental trajectories of subcortical
122 Annals of the New York Academy of Sciences

limbic regions relative cortical control regions. These decisions about hypothetical scenarios versus real-life
developmental changes can be exacerbated by individ- situations (Sobesky 1983). The environmental context
ual differences (e.g., genetic risk) in baseline activity of and emotional significance of the decision greatly in-
limbic systems. fluence the adolescent (Steinberg 2005).
This model of development reconciles a number Our findings and model have significant implica-
of contradictions and myths about adolescence. First, tions for heated debates on public policy and the treat-
there have been many reports that suggest that ado- ment of minors in our judicial system. Adolescents
lescent behavior is due to protracted development of show adult levels of intellectual capability earlier than
prefrontal cortex. However, if this were the case, then they show evidence of adult levels of impulse control
children would engage in similar or worse behavior (Reyna & Farley 2006). As such, adolescents may be
than adolescents. The National Center for Health capable of making informed choices about their fu-
Statistics on adolescent behavior and mortality shows ture (e.g., terminating a pregnancy) but do not yet
that suboptimal choices and actions observed during have full capacity to override impulses in emotion-
adolescence represent a nonlinear change in behavior, ally charged situations that require decisions in the
distinct from childhood and adulthood. Adolescents, heat of the moment. Unfortunately, judges, politicians,
unlike children, may be in situations (e.g., driving a advocates, and journalists are biased toward drawing
car) that may put them at greater risk for mortality, a single line between adolescence and adulthood for
but even when taking these conditions into account, different purposes under the law that is at odds with de-
there is still a significant elevation of risky behavior in velopmental cognitive neuroscience (Steinberg et al. in
adolescents in comparison to children. Furthermore, press). Our neurodevelopmental model of adolescence
experimental studies have shown that when risk is held will hopefully help to make strides in moving this sin-
constant, such as in the appraisal of risky vignettes, gle line to multiple lines that consider developmental
children perceive greater risk in hypothetical scenar- changes across both context (emotionally charged or
ios than do adolescents (reviewed in Furby & Beyth- not) and time (in the moment or in the future).
Marom 1992). Our neurodevelopmental model pro-
vides an explanation for these nonlinear changes in
Acknowledgments
behavior.
Reyna and Farley (2006) have reconciled the second This work was supported in part by grants from
myth by showing that adolescents are able to reason the National Institute of Drug Abuse R01 DA18879
and understand risks of behaviors in which they en- and the National Institute of Mental Health R01
gage and do not consider themselves invincible. Prior MH73175 and P50 MH62196 to BJC.
research has also shown that adolescents knowingly
engage in risky behavior, and this is often due to in-
fluences of feelings, emotions, and peers (Gardener & Conflict of Interests
Steinberg 2005; Steinberg 2004, 2005). The observa-
The authors declare no conflicts of interest.
tion that adolescents know that they are engaging in
risky behavior is not supported by the sole explanation
of a less developed prefrontal cortex. In this context, References
our model suggests that the adolescent is capable of
making rational decisions, but in emotionally charged Adriani, W., Chiarotti, F., & Laviola, G. (1998). Elevated novelty
situations the more mature limbic system will win over seeking and peculiar d-amphetamine sensitization in peri-
adolescent mice compared with adult mice. Behav Neurosci,
the prefrontal control system.
112(5), 1152–1166.
When faced with an immediate personal decision, Andersen, S. L., & Teicher, M. H. (2004). Delayed effects of early
adolescents will rely less on intellectual capabilities stress on hippocampal development. Neuropsychopharmacol-
and more on feelings. Nevertheless, when reasoning ogy, 29(11), 1988–1993.
about a hypothetical, moral dilemma, the adolescent Andersen, S. L., Thompson, A. T., Rutstein, M., Hostetter, J. C.,
will rely more on logical information (Steinberg 2005). & Teicher, M. H. (2000). Dopamine receptor pruning in
In other words, when a poor decision is made in the prefrontal cortex during the periadolescent period in rats.
Synapse, 37(2), 167–169.
heat of the moment, the adolescent may know better,
Anglada-Figueroa, D., & Quirk, G. J. (2005). Lesions of the basal
but the salience of the emotional context biases his amygdala block expression of conditioned fear but not
or her behavior in opposite direction of the optimal extinction. J Neurosci, 25(42), 9680–9685.
action. This work coincides with studies of social cog- Baird, A. A., Gruber, S. A., Fein, D. A., Maas, L. C., Steingard,
nition showing that adolescents make more rational R. J., Renshaw, P. F., et al. (1999). Functional magnetic
Casey et al.: The Adolescent Brain 123

resonance imaging of facial affect recognition in children Casey, B. J., Galvan, A., & Hare, T. A. (2005a). Changes in
and adolescents. J Am Acad Child Adolesc Psychiatry, 38(2), cerebral functional organization during cognitive devel-
195–199. opment. Curr Opin Neurobiol, 15(2), 239–244.
Baxter, M. G., Parker, A., Lindner, C. C., Izquierdo, A. D., Casey, B., Getz, S., & Galvan, A. (in press). Developmental Review.
& Murray, E. A. (2000). Control of response selec- Casey, B. J., Giedd, J. N., & Thomas, K. M. (2000a). Struc-
tion by reinforcer value requires interaction of amygdala tural and functional brain development and its rela-
and orbital prefrontal cortex. J Neurosci, 20(11), 4311– tion to cognitive development. Biol Psychol, 54(1–3), 241–
4319. 257.
Benthin, A., Slovic, P., & Severson, H. (1993). A psychometric Casey, B. J., Thomas, K. M., Davidson, M. C., Kunz, K., &
study of adolescent risk perception. J Adolesc, 16(2), 153– Franzen, P. L. (2002a). Dissociating striatal and hippocam-
168. pal function developmentally with a stimulus-response
Bjork, J. M., Smith, A. R., Danube, C. L., & Hommer, compatibility task. J Neurosci, 22(19), 8647–8652.
D. W. (2007). Developmental differences in posterior Casey, B. J., Thomas, K. M., Welsh, T. F., Badgaiyan, R. D., Ec-
mesofrontal cortex recruitment by risky rewards. J Neu- card, C. H., Jennings, J. R., et al. (2000b). Dissocaiation of
rosci, 27(18), 4839–4849. response conflict, attentional selection, and expectancy
Bjorkland, D. F. (1985). The role of conceptual knowledge in with functional magnetic resonance imaging. Proc
the development of organization in children’s memory. Natl Acad Sci, USA, 97(18), 8728–8733.
In C. J. Brainerd & M. Pressley (Eds.), Basic Processes in Casey, B. J., Tottenham, N., & Fossella, J. (2002b). Clinical, imag-
Memory Development: Progress in Cognitive Development Research ing, lesion, and genetic approaches toward a model of
(pp. 103–142). New York: Springer-Verlag. cognitive control. Dev Psychobiol, 40(3), 237–254.
Bjorkland, D. F. (1987). How age changes in knowledge base Casey, B. J., Tottenham, N., Liston, C., & Durston, S. (2005b).
contribute to the development of children’s memory: An Imaging the developing brain: what have we learned
interpretive review. Developmental Review, 7, 93–130. about cognitive development? Trends in Cognitive Science,
Blum, K., Braverman, E. R., Holder, J. M., Lubar, J. F., Monas- 9(3), 104–110.
tra, V. J., Miller, D., et al. (2000). Reward deficiency syn- Casey, B. J., Trainor, R. J., Orendi, J. L., Schubert, A. B., Nys-
drome: a biogenetic model for the diagnosis and treatment trom, L. E., Giedd, J. N., et al. (1997b). A developmental
of impulsive, addictive, and compulsive behaviors. J Psy- functional MRI study of prefrontal activation during per-
choactive Drugs, 32 Suppl, i-iv, 1–112. formance of a go-no-go task. Journal of Cognitive Neuroscience,
Bogin, B. (1994). Adolescence in evolutionary perspective. Acta 9, 835–847.
Paediatr Suppl, 406, 29–35; discussion 36. Caviness, V., Kennedy, D., Richelme, C., Rademacher, J., &
Bourgeois, J. P., Goldman-Rakic, P. S., & Rakic, P. (1994). Synap- Filipek, P. (1996). The human brain age 7–11 years: a
togenesis in the prefrontal cortex of rhesus monkeys. Cereb volumetric analysis based on magnetic resonance images.
Cortex, 4(1), 78–96. Cereb Cortex, 6(5), 726–736.
Brainerd, C. J., & Reyna, V. F. (1993). Memory independence Crews, F., He, J., & Hodge, C. (2007). Adolescent cortical de-
and memory interference in cognitive development. Psy- velopment: a critical period of vulnerability for addiction.
chol Rev, 100(1), 42–67. Pharmacol Biochem Behav, 86(2), 189–199.
Breiter, H. C., Etcoff, N. L., Whalen, P. J., Kennedy, W. A., Csikszentmihalyi, M., Larson, R., & Prescott, S. (1977). The
Rauch, S. L., Buckner, R. L., et al. (1996). Response and ecology of adolescent activity and experience. Journal of
habituation of the human amygdala during visual process- Youth and Adolescence, 6, 281–294.
ing of facial expression. Neuron, 17(5), 875–887. Cunningham, M. G., Bhattacharyya, S., & Benes, F. M. (2002).
Brown, T. T., Lugar, H. M., Coalson, R. S., Miezin, F. M., Amygdalo-cortical sprouting continues into early adult-
Petersen, S. E., & Schlaggar, B. L. (2005). Developmental hood: implications for the development of normal and
changes in human cerebral functional organization for abnormal function during adolescence. J Comp Neurol,
word generation. Cereb Cortex, 15(3), 275–290. 453(2), 116–130.
Bunge, S. A., Dudukovic, N. M., Thomason, M. E., Vaidya, C. Davis, M., & Whalen, P. J. (2001). The amygdala: vigilance and
J., & Gabrieli, J. D. (2002). Immature frontal lobe con- emotion. Mol Psychiatry, 6(1), 13–34.
tributions to cognitive control in children: evidence from Dempster, F. N. (Ed.) (1993). Resistance to Interference: Developmental
fMRI. Neuron, 33(2), 301–311. Changes in a Basic Processing Mechanism. (Vol. 1). New York:
Cascio, C. J., Gerig, G., & Piven, J. (2007). Diffusion tensor Springer-Verlag.
imaging: Application to the study of the developing brain. Diamond, A. (1985). Development of the ability to use recall to
J Am Acad Child Adolesc Psychiatry, 46(2), 213–223. guide action, as indicated by infants’ performance on AB.
Case, R. (1972). Balidation of a neo-Piagetian capacity construct. Child Development, 56, 868–883.
Journal of Experimental Child Psychology, 14, 287–302. Durston, S., Davidson, M. C., Thomas, K. M., Worden, M. S.,
Case, R. (1985). Intellectual Development Birth to Adulthood. New York: Tottenham, N., Martinez, A., et al. (2003). Parametric
Academic Press. manipulation of conflict and response competition using
Casey, B. J., Castellanos, F. X., Giedd, J. N., Marsh, W. L., Ham- rapid mixed-trial event-related fMRI. Neuroimage, 20(4),
burger, S. D., Schubert, A. B., et al. (1997a). Implication 2135–2141.
of right frontostriatal circuitry in response inhibition and Durston, S., Davidson, M. C., Tottenham, N., Galvan, A., Spicer,
attention-deficit/hyperactivity disorder. J Am Acad Child J., Fossella, J. A., et al. (2006). A shift from diffuse to focal
Adolesc Psychiatry, 36(3), 374–383. cortical activity with development. Dev Sci, 9(1), 1–8.
124 Annals of the New York Academy of Sciences

Durston, S., Hulshoff, H. E., Casey, B. J., Giedd, J. N., Buitelaar, Gogtay, N., Giedd, J. N., Lusk, L., Hayashi, K. M., Greenstein,
J. K., & Van Engeland, H. (2001). Anatomical MRI of the D., Vaituzis, A. C., et al. (2004). Dynamic mapping of
developing human brain: What have we learned? J Am human cortical development during childhood through
Acad Child Adolesc Psychiatry, 40(9), 1012–1020. early adulthood. Proc Natl Acad Sci U S A, 101(21), 8174–
Eaton, L. K., Kinchen, S., Ross, J., Hawkins, J., Harris, W. A., 8179.
Lowry, R., et al. (2006). Youth risk behavior surveillance- Grace, A. A., Floresco, S. B., Goto, Y., & Lodge, D. J. (2007).
United States, 2005, surveillance summaries. Morbidity and Regulation of firing of dopaminergic neurons and con-
Mortality Weekly Report, 55(SS-5), 1–108. trol of goal-directed behaviors. Trends Neurosci, 30(5), 220–
Eigsti, I. M., Zayas, V., Mischel, W., Shoda, Y., Ayduk, O., Dad- 227.
lani, M. B., et al. (2006). Predicting cognitive control from Hare, T., & Casey, B. (in press). Biological Psychiatry.
preschool to late adolescence and young adulthood. Psy- Hare, T. A., & Casey, B. J. (2005). The neurobiology and devel-
chol Sci, 17(6), 478–484. opment of cognitive and affective control. Cognition, Brain,
Ernst, M., Nelson, E. E., Jazbec, S., McClure, E. B., Monk, Behavior, 9(3), 273–286.
C. S., Leibenluft, E., et al. (2005). Amygdala and nu- Hare, T. A., Tottenham, N., Davidson, M. C., Glover, G. H., &
cleus accumbens in responses to receipt and omission of Casey, B. J. (2005). Contributions of amygdala and striatal
gains in adults and adolescents. Neuroimage, 25(4), 1279– activity in emotion regulation. Biol Psychiatry, 57(6), 624–
1291. 632.
Ernst, M., Pine, D. S., & Hardin, M. (2006). Triadic model of Hare, T. A., Tottenham, N., Voss, H. U., Glover, G. H., & Casey,
the neurobiology of motivated behavior in adolescence. B. J. (in press). The adolescent brain and potential risk for
Psychol Med, 36(3), 299–312. anxiety and depression. Biological Psychiatry.
Eshel, N., Nelson, E. E., Blair, R. J., Pine, D. S., & Ernst, M. Harnishfeger, K. K., & Bjorkland, F. (1993). The ontogeny of
(2007). Neural substrates of choice selection in adults inhibition mechanisms: A renewed approach to cognitive
and adolescents: development of the ventrolateral pre- development. In M. L. Howe & R. Pasnek (Eds.), Emgering
frontal and anterior cingulate cortices. Neuropsychologia, Themes in Cognitive Development (Vol. 1, pp. 28–49). New
45(6), 1270–1279. York: Springer-Verlag.
Flavell, J. H., Feach, D. R., & Chinsky, J. M. (1966). Spontaneous Hikosaka, K., & Watanabe, M. (2000). Delay activity of orbital
verbal rehearsal in a memory task as a function of age. and lateral prefrontal neurons of the monkey varying with
Child Development, 37, 283–299. different rewards. Cereb Cortex, 10(3), 263–271.
Floresco, S. B., & Tse, M. T. (2007). Dopaminergic regulation of Huttenlocher, P. R. (1979). Synaptic density in human frontal
inhibitory and excitatory transmission in the basolateral cortex - developmental changes and effects of aging. Brain
amygdala-prefrontal cortical pathway. J Neurosci, 27(8), Res, 163(2), 195–205.
2045–2057. Hyman, S. E., & Malenka, R. C. (2001). Addiction and the brain:
Furby, L., & Beyth-Marom, R. (1992). Risk taking in adolescence: the neurobiology of compulsion and its persistence. Nat
A decision-making perspective. Developmental Review, 12(1), Rev Neurosci, 2(10), 695–703.
1–44. Jackson, M. E., Frost, A. S., & Moghaddam, B. (2001). Stimu-
Gallagher, M., McMahan, R. W., & Schoenbaum, G. (1999). lation of prefrontal cortex at physiologically relevant fre-
Orbitofrontal cortex and representation of incentive quencies inhibits dopamine release in the nucleus accum-
value in associative learning. J Neurosci, 19(15), 6610– bens. J Neurochem, 78(4), 920–923.
6614. Jernigan, T. L., Zisook, S., Heaton, R. K., Moranville, J. T.,
Galvan, A., Hare, T., Voss, H., Glover, G., & Casey, B. J. (2007). Hesselink, J. R., & Braff, D. L. (1991). Magnetic resonance
Risk-taking and the adolescent brain: who is at risk? Dev imaging abnormalities in lenticular nuclei and cerebral
Sci, 10(2), F8–F14. cortex in schizophrenia. Arch Gen Psychiatry, 48, 881–890.
Galvan, A., Hare, T. A., Parra, C. E., Penn, J., Voss, H., Glover, Kalin, N. H., Shelton, S. E., & Davidson, R. J. (2004). The
G., et al. (2006). Earlier development of the accumbens role of the central nucleus of the amygdala in mediating
relative to orbitofrontal cortex might underlie risk-taking fear and anxiety in the primate. J Neurosci, 24(24), 5506–
behavior in adolescents. J Neurosci, 26(25), 6885–6892. 5515.
Gardener, M., & Steinberg, L. (2005). Peer influence on risk tak- Kalsbeek, A., Voorn, P., Buijs, R. M., Pool, C. W., & Uylings, H.
ing, risk preference, and risky decision making in adoles- B. (1988). Development of the dopaminergic innervation
cence and adulthood: an experimental study. Developmental in the prefrontal cortex of the rat. J Comp Neurol, 269(1),
Psychology, 41, 625–635. 58–72.
Giedd, J. N. (2004). Structural magnetic resonance imaging of Keating, D. P., & Bobbitt, B. L. (1978). Individual and develop-
the adolescent brain. Ann N Y Acad Sci, 1021, 77–85. mental differences in cognitive processing components of
Giedd, J. N., Blumenthal, J., Jeffries, N. O., Castellanos, F. X., mental ability. Child Development, 49, 155–167.
Liu, H., Zijdenbos, A., et al. (1999). Brain development Kelley, A. E., Schochet, T., & Landry, C. F. (2004). Risk taking
during childhood and adolescence: a longitudinal MRI and novelty seeking in adolescence: introduction to part
study. Nat Neurosci, 2(10), 861–863. I. Ann N Y Acad Sci, 1021, 27–32.
Giedd, J. N., Snell, J. W., Lange, N., Rajapakse, J. C., Casey, B. J., Killgore, W. D., Oki, M., & Yurgelun-Todd, D. A. (2001).
Kaysen, D., et al. (1996). Quantitative magnetic resonance Sex-specific developmental changes in amygdala re-
imaging of human brain development: ages 4–18. Cereb sponses to affective faces. Neuroreport, 12(2), 427–
Cortex, 6, 551–560. 433.
Casey et al.: The Adolescent Brain 125

Knutson, B., Adams, C. M., Fong, G. W., & Hommer, D. (2001). Morris, J. S., Friston, K. J., Buchel, C., Frith, C. D., Young, A.
Anticipation of increasing monetary reward selectively re- W., Calder, A. J., et al. (1998). A neuromodulatory role
cruits nucleus accumbens. J Neurosci, 21(16), RC159. for the human amygdala in processing emotional facial
Kuhnen, C. M., & Knutson, B. (2005). The neural basis of finan- expressions. Brain, 121(Pt 1), 47–57.
cial risk taking. Neuron, 47(5), 763–770. Moses, P., Roe, K., Buxton, R. B., Wong, E. C., Frank, L. R.,
Laviola, G., Adriani, W., Terranova, M. L., & Gerra, G. (1999). & Stiles, J. (2002). Functional MRI of global and local
Psychobiological risk factors for vulnerability to psychos- processing in children. Neuroimage, 16(2), 415–424.
timulants in human adolescents and animal models. Neu- Munakata, Y., & Yerys, B. E. (2001). All together now: when
rosci Biobehav Rev, 23(7), 993–1010. dissociations between knowledge and action disappear.
Laviola, G., Macri, S., Morley-Fletcher, S., & Adriani, W. (2003). Psychol Sci, 12(4), 335–337.
Risk-taking behavior in adolescent mice: psychobiologi- Nagy, Z., Westerberg, H., & Klingberg, T. (2004). Maturation of
cal determinants and early epigenetic influence. Neurosci white matter is associated with the development of cog-
Biobehav Rev, 27(1–2), 19–31. nitive functions during childhood. J Cogn Neurosci, 16(7),
Lenroot, R. K., & Giedd, J. N. (2006). Brain development in 1227–1233.
children and adolescents: insights from anatomical mag- O’Doherty, J., Kringelbach, M. L., Rolls, E. T., Hornak, J., & An-
netic resonance imaging. Neurosci Biobehav Rev, 30(6), 718– drews, C. (2001). Abstract reward and punishment repre-
729. sentations in the human orbitofrontal cortex. Nat Neurosci,
Leppanen, J. M. (2006). Emotional information processing in 4(1), 95–102.
mood disorders: a review of behavioral and neuroimaging O’Doherty, J. P. (2004). Reward representations and reward-
findings. Curr Opin Psychiatry, 19(1), 34–39. related learning in the human brain: insights from neu-
Liston, C., Watts, R., Tottenham, N., Davidson, M. C., Niogi, roimaging. Curr Opin Neurobiol, 14(6), 769–776.
S., Ulug, A. M., et al. (2006). Frontostriatal microstructure Pascual-Leone, J. A. (1970). A mathematical model for transition
modulates efficient recruitment of cognitive control. Cereb in Piaget’s developmental stages. Acta Psychologica, 32, 301–
Cortex, 16(4), 553–560. 345.
Luna, B., & Sweeney, J. A. (2004). The emergence of collabora- Pecina, S., Cagniard, B., Berridge, K. C., Aldridge, J. W., &
tive brain function: FMRI studies of the development of Zhuang, X. (2003). Hyperdopaminergic mutant mice
response inhibition. Ann N Y Acad Sci, 1021, 296–309. have higher “wanting” but not “liking” for sweet rewards.
Luna, B., Thulborn, K. R., Munoz, D. P., Merriam, E. P., Garver, J Neurosci, 23(28), 9395–9402.
K. E., Minshew, N. J., et al. (2001). Maturation of widely Perez-Edgar, K., Roberson-Nay, R., Hardin, M. G., Poeth, K.,
distributed brain function subserves cognitive develop- Guyer, A. E., Nelson, E. E., et al. (2007). Attention al-
ment. Neuroimage, 13(5), 786–793. ters neural responses to evocative faces in behaviorally
Maren, S., & Quirk, G. J. (2004). Neuronal signalling of fear inhibited adolescents. Neuroimage, 35(4), 1538–1546.
memory. Nat Rev Neurosci, 5(11), 844–852. Pine, D. S., Cohen, P., & Brook, J. S. (2001). Emotional reactiv-
Matthews, S. C., Simmons, A. N., Lane, S. D., & Paulus, M. P. ity and risk for psychopathology among adolescents. CNS
(2004). Selective activation of the nucleus accumbens dur- Spectr, 6(1), 27–35.
ing risk-taking decision making. Neuroreport, 15(13), 2123– Primus, R. J., & Kellogg, C. K. (1989). Pubertal-related changes
2127. influence the development of environment-related social
May, J. C., Delgado, M. R., Dahl, R. E., Stenger, V. A., Ryan, interaction in the male rat. Dev Psychobiol, 22(6), 633–
N. D., Fiez, J. A., et al. (2004). Event-related functional 643.
magnetic resonance imaging of reward-related brain cir- Rauch, S. L., Shin, L. M., & Wright, C. I. (2003). Neuroimaging
cuitry in children and adolescents. Biol Psychiatry, 55(4), studies of amygdala function in anxiety disorders. Ann N
359–366. Y Acad Sci, 985, 389–410.
McClure, E. B., Monk, C. S., Nelson, E. E., Parrish, J. M., Reiss, A., Abrams, M., Singer, H., Ross, J., & Denckla, M. (1996).
Adler, A., Blair, R. J., et al. (2007). Abnormal attention Brain development, gender and IQ in children. A volu-
modulation of fear circuit function in pediatric generalized metric imaging study. Brain, 119, 1763–1774.
anxiety disorder. Arch Gen Psychiatry, 64(1), 97–106. Reyna, V., & Farley, F. (2006). Risk and rationality in adolescent
McClure, S. M., Laibson, D. I., Loewenstein, G., & Cohen, J. decision making: implications for theory, practice, and
D. (2004). Separate neural systems value immediate and public policy. Psychological Science in the Public Interest, 7(1),
delayed monetary rewards. Science, 306(5695), 503–507. 1–44.
Miller, E. K., & Cohen, J. D. (2001). An integrative theory of Rosenberg, D. R., & Lewis, D. A. (1994). Changes in the
prefrontal cortex function. Annu Rev Neurosci, 24, 167–202. dopaminergic innervation of monkey prefrontal cortex
Mischel, W., Shoda, Y., & Rodriguez, M. I. (1989). Delay of during late postnatal development: a tyrosine hydroxylase
gratification in children. Science, 244(4907), 933–938. immunohistochemical study. Biol Psychiatry, 36(4), 272–
Monk, C. S., McClure, E. B., Nelson, E. E., Zarahn, E., Bilder, 277.
R. M., Leibenluft, E., et al. (2003). Adolescent immaturity Rosenberg, D. R., & Lewis, D. A. (1995). Postnatal maturation of
in attention-related brain engagement to emotional facial the dopaminergic innervation of monkey prefrontal and
expressions. Neuroimage, 20(1), 420–428. motor cortices: a tyrosine hydroxylase immunohistochem-
Montague, P. R., & Berns, G. S. (2002). Neural economics and ical analysis. J Comp Neurol, 358(3), 383–400.
the biological substrates of valuation. Neuron, 36(2), 265– Rubia, K., Overmeyer, S., Taylor, E., Brammer, M., Williams,
284. S. C., Simmons, A., et al. (2000). Functional frontalisation
126 Annals of the New York Academy of Sciences

with age: mapping neurodevelopmental trajectories with Steinberg, L., Cauffman, E., Woolard, J., Graham, S., & Banich,
fMRI. Neurosci Biobehav Rev, 24(1), 13–19. M. (in press). Are adolescents less mature than adults?
Rubia, K., Smith, A. B., Woolley, J., Nosarti, C., Heyman, I., Minors’ access to abortion, the juvenile death penalty,
Taylor, E., et al. (2006). Progressive increase of frontostri- and the alleged APA “Flip-Flop”.
atal brain activation from childhood to adulthood during Tamm, L., Menon, V., & Reiss, A. L. (2002). Maturation of brain
event-related tasks of cognitive control. Hum Brain Mapp, function associated with response inhibition. J Am Acad
27(12), 973–993. Child Adolesc Psychiatry, 41(10), 1231–1238.
Schlaggar, B. L., Brown, T. T., Lugar, H. M., Visscher, K. Tarazi, F. I., Tomasini, E. C., & Baldessarini, R. J. (1998). Postna-
M., Miezin, F. M., & Petersen, S. E. (2002). Functional tal development of dopamine and serotonin transporters
neuroanatomical differences between adults and school- in rat caudate-putamen and nucleus accumbens septi. Neu-
age children in the processing of single words. Science, rosci Lett, 254(1), 21–24.
296(5572), 1476–1479. Teicher, M. H., Andersen, S. L., & Hostetter, J. C., Jr. (1995).
Schultz, W. (2006). Behavioral theories and the neurophysiology Evidence for dopamine receptor pruning between adoles-
of reward. Annual Reviews of Psychology, 57, 87–115. cence and adulthood in striatum but not nucleus accum-
Shin, L. M., Orr, S. P., Carson, M. A., Rauch, S. L., Macklin, bens. Brain Res Dev Brain Res, 89(2), 167–172.
M. L., Lasko, N. B., et al. (2004). Regional cerebral blood Thomas, K. M., Drevets, W. C., Dahl, R. E., Ryan, N. D., Birma-
flow in the amygdala and medial prefrontal cortex during her, B., Eccard, C. H., et al. (2001a). Amygdala response
traumatic imagery in male and female Vietnam veterans to fearful faces in anxious and depressed children. Arch Gen
with PTSD. Arch Gen Psychiatry, 61(2), 168–176. Psychiatry, 58(11), 1057–1063.
Silveri, M. M., Tzilos, G. K., Pimentel, P. J., & Yurgelun-Todd, Thomas, K. M., Drevets, W. C., Whalen, P. J., Eccard, C. H.,
D. A. (2004). Trajectories of adolescent emotional and Dahl, R. E., Ryan, N. D., et al. (2001b). Amygdala re-
cognitive development: effects of sex and risk for drug sponse to facial expressions in children and adults. Biol
use. Ann N Y Acad Sci, 1021, 363–370. Psychiatry, 49(4), 309–316.
Sobesky, W. (1983). The effects of situational factors on moral Thomas, K. M., Hunt, R. H., Vizueta, N., Sommer, T., Durston,
judgements. Child Development, 54, 575–584. S., Yang, Y., et al. (2004). Evidence of developmental dif-
Sowell, E. R., Peterson, B. S., Thompson, P. M., Welcome, S. ferences in implicit sequence learning: an FMRI study of
E., Henkenius, A. L., & Toga, A. W. (2003). Mapping children and adults. J Cogn Neurosci, 16(8), 1339–1351.
cortical change across the human life span. Nat Neurosci, Turkeltaub, P. E., Gareau, L., Flowers, D. L., Zeffiro, T. A., &
6(3), 309–315. Eden, G. F. (2003). Development of neural mechanisms
Sowell, E. R., Thompson, P. M., Holmes, C. J., Jernigan, T. L., & for reading. Nat Neurosci, 6(7), 767–773.
Toga, A. W. (1999). In vivo evidence for post-adolescent Volkow, N. D., & Li, T. K. (2004). Drug addiction: the neuro-
brain maturation in frontal and striatal regions. Nat Neu- biology of behaviour gone awry. Nat Rev Neurosci, 5(12),
rosci, 2(10), 859–861. 963–970.
Sowell, E. R., Thompson, P. M., & Toga, A. W. (2004). Mapping Yurgelun-Todd, D. (2007). Emotional and cognitive changes dur-
changes in the human cortex throughout the span of life. ing adolescence. Curr Opin Neurobiol, 17(2), 251–257.
Neuroscientist, 10(4), 372–392. Yurgelun-Todd, D. A., & Killgore, W. D. (2006). Fear-related
Spear, L. P. (2000). The adolescent brain and age-related behav- activity in the prefrontal cortex increases with age dur-
ioral manifestations. Neurosci Biobehav Rev, 24(4), 417–463. ing adolescence: a preliminary fMRI study. Neurosci Lett,
Steinberg, L. (1989). Pubertal maturation and parent-adolescent 406(3), 194–199.
distance: an evolutionary perspective. In G. Adams, R. Zald, D. H., Boileau, I., El-Dearedy, W., Gunn, R., McGlone,
Montemayor & T. Gullotta (Eds.), Advances in Adolescent F., Dichter, G. S., et al. (2004). Dopamine transmission
Behavior and Development (pp. 71–97). Newbury Park, CA: in the human striatum during monetary reward tasks. J
Sage Publications. Neurosci, 24(17), 4105–4112.
Steinberg, L. (2004). Risk taking in adolescence: what changes, Zehr, J. L., Todd, B. J., Schulz, K. M., McCarthy, M. M., & Sisk,
and why? Ann N Y Acad Sci, 1021, 51–58. C. L. (2006). Dendritic pruning of the medial amygdala
Steinberg, L. (2005). Cognitive and affective development in ado- during pubertal development of the male Syrian hamster.
lescence. Trends Cogn Sci, 9(2), 69–74. J Neurobiol, 66(6), 578–590.