The beautiful adolescent brain: An evolutionary developmental perspective

Abstract

The adolescent brain has been characterized as a defective car, with no brakes or steering wheel—only an accelerator. This characterization has been used to explain the impulsive and risky behavior of this transient developmental period. But why do adolescents respond to the world the way they do? In this article, we consider adolescent‐specific changes in the brain and behavior from a developmental evolutionary viewpoint in how they might be adaptive. We suggest ways in which the adolescent brain has evolved to explore and learn from new and changing environments as the adolescent gains independence from the caregiver and transitions into an adult. We highlight adolescent‐specific changes in the brain and behavior in response to emotional and social cues that may facilitate learning to independently secure resources (e.g., food, water, and shelter) and to establish new social bonds beyond the family or pack for their own survival. Specifically, we focus on how rewards, social cues, and threats in the environment influence behavior and may serve an adaptive role for the adolescent.

In this article, adolescent‐specific changes in brain and behavior are considered through an evolutionary developmental lens in how they might be adaptive. We highlight examples of how changes in response to emotionally and socially relevant cues and contexts during this developmental phase are likely essential for learning to negotiate one's rapidly changing world independently and for the survival of the individual and species.

INTRODUCTION

Early characterizations of the adolescent brain suggested that it was like a defective car, with no brakes or steering wheel—just an accelerator. The lack of a steering wheel or brakes presumably reflects the lack of a mature prefrontal cortex implicated in impulse control and decision‐making, leaving behavior to be driven by basic desires and threats without consideration of future goals or consequences of actions. Yet, adolescence is a remarkable period of learning, as the individual increasingly explores their rapidly changing world without the protection of a parent or caregiver. ¹ , ² , ³ , ⁴ , ⁵ In this article, we highlight adolescent‐specific changes in brain and behavior and how they may serve an adaptive function when seen through an evolutionary developmental lens. Cross‐species examples are provided to help illustrate this viewpoint. ⁶

It is important to state at the onset that just as descriptions of the adolescent brain as defective are not accurate portrayals of adolescence, neither is the suggestion that all adolescent behaviors are simply adaptive. The same mechanisms that may facilitate adaptive behavior in one context (e.g., rewarding, nonviolent, predictable) may promote risks and vulnerabilities in another (e.g., violent, abusive, or coercive), as evidenced by an increase in mental illness, ⁷ criminal behavior, ⁸ and mortality ⁹ largely due to preventable causes during this transient development period. Moreover, any single snapshot in time of any one individual or developmental group will not wholly generalize to understanding the entire group across a variety of situations, experimental manipulations, and prior experiences. ¹ , ¹⁰ , ¹¹ , ¹² Neuroscience research demonstrates this point by showing that patterns of brain function not only change across development but also within individuals and within individuals under different contexts. ¹⁰ , ¹¹

However, there are general patterns in adolescent behavior that are not specific to humans. These tendencies such as increased reward and novelty seeking, changes in peer and parent interactions, and altered responses to threats are observed across mammalian species (e.g., monkeys, rats, and mice). ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ This conservation across species likely reflects the surprising stability in these different species’ environments. Even the rapidly changing worlds of adolescents and the associated demands on them are in many ways predictable. ⁹ , ¹⁷ It is a time when the organism must learn to independently secure resources (e.g., food and water) for their survival and establish new social bonds beyond the family or pack for their own survival and that of the species. ⁵ , ⁹ , ¹⁷ , ¹⁸ Given that social status and socialization play a significant role in securing resources (e.g., food and water), survival, and reproduction of the species, it would seem reasonable that mechanisms have evolved for the detection of rewards, social cues, and threats in the environment. Likewise, learning to predict rewards, outcomes of social interactions, and threats would likely serve as critical abilities necessary for successfully transitioning from dependence on the parent/caregiver in childhood to relative independence as an adult. ⁹

In assessing the adaptiveness of the adolescent brain and behavior, we anchor our discussion on four key questions posed by the ethologist Nikolaas Tinbergen in his tribute to the Nobel laureate Konrad Lorenz: ⁹ , ¹⁹ (1) How does behavior change across development? (2) How has it evolved across species? (3) What are the potential mechanisms underlying adolescent behavior? and (4) How may adolescent behavior be adaptive?

In addressing each of these questions, we focus on three different areas of developmental research. These include the influence of rewards, ²⁰ , ²¹ , ²² , ²³ peers, ²⁴ , ²⁵ , ²⁶ and potential threats ²⁷ , ²⁸ , ²⁹ , ³⁰ on brain and behavior. We selected these three areas because they: (1) show peak changes during adolescence; (2) have been the focus of several studies on the adolescent brain; (3) have been examined across species; ¹ , ⁶ , ³¹ and (4) allow us to address our four key questions above to understand how the adolescent brain and behavior are adaptive. Moreover, experiencing rewards, social interactions, and threats likely generate changes in the brain and behavior that influence goal‐directed behavior—a cornerstone of cognitive development ⁶ —and adaptive plasticity. ⁵ Goal‐directed behaviors involve agency in that the organism or individual is choosing an action in response to a stimulus, event, or context in anticipation of a certain outcome. Learning associations among events, actions, and consequences is required to optimize desired outcomes (e.g., gains in rewards, peer acceptance, avoiding threats, losses, and peer rejection). This type of goal‐directed learning in rapidly changing environments is thought to be an essential part of adolescent development. ⁵

HOW DOES BEHAVIOR CHANGE WITH DEVELOPMENT?

Below, we provide selected examples that illustrate specific changes in the influence of rewards, ²⁰ , ²¹ , ³² peer influence, ²⁴ , ²⁵ and potential threats ²⁵ , ²⁷ , ²⁹ , ³⁰ on behavior during adolescence.

Influence of rewards

There is a large literature demonstrating a heightened sensitivity to the immediacy, size, and uncertainty of rewards during adolescence. ²⁰ , ³² , ³³ , ³⁴ However, a key question is how does reward and the value of reinforced information guide actions and goal‐directed learning and memory in adolescents as they adapt to their rapidly changing environments?

To begin, we illustrate how previously rewarded information can subsequently drive or alter actions in adolescents. In this study, Davidow and colleagues ³⁵ examined how the value of reinforced information guides behavior in 8‐ to 25‐year‐old participants using two tasks: a variant of a monetary incentive delay task and a subsequent go/no‐go task. In the first task, participants were instructed to make rapid motor responses to cues presented on a screen. Rapid responses to one of these cues were monetarily rewarded, while no reward was given for rapid responses to the other cue. Approximately an hour later, the participants performed a go/no‐go task that included these previously rewarded and nonrewarded cues. The results showed more impulsive errors when the no‐go stimulus (nontarget) was a previously rewarded cue than when it was a nonrewarded cue. This disruption in response inhibition was evidenced by more impulsive errors to the rewarded cue with increasing age. This pattern remitted over the course of the task in adults, but the impulsive pattern persisted in adolescents. These findings suggest that not only is the value of reinforced information associated with momentary changes in behavior, but it can also persist and facilitate behavioral changes in a new task context in adolescents.

While this example simplistically illustrates the motivating effects of rewarded information on actions during adolescence, rewards also are associated with improvements in goal‐directed learning and behavior. Using cues from a reinforcement learning task, Insel and colleagues showed that cues that were reinforced during the task improved subsequent response inhibition performance to these cues throughout adolescence from 13 to 20 years. ³⁶ Relatedly, Heffer et al. ³⁷ showed improvement in response inhibition accuracy to previously rewarded compared to previously punished cues, which started around 14 years of age and increased with age. These findings illustrate that adolescents can use reward value to guide goal‐directed behavior, an ability that improves throughout adolescence.

Other studies have examined choice behavior and learning during adolescence. As an example, adolescent and adult choice behavior was measured using a monetary gambling task in which participants chose to accept or reject gambles of varying expected value (i.e., the sum of all of the possible outcomes of a particular choice multiplied by their probabilities). Increasing expected value had a stronger influence over gambling choices in adolescents (13−17 years) relative to adults (25−30 years) even when controlling for subjective value (i.e., how hard an individual might work or would pay for something) ²¹ (Figure 1A). These behavioral data suggest that the value of available options has a greater influence on adolescent versus adult choices.

FIGURE 1.

Illustration of optimal choices by adolescents relative to adults. (A) Mean choices to gamble or not based on expected reward value for adolescent and adults (adapted from Barkley‐Levenson and Galván, 2014. ²¹ ) (B) Percent optimal choices on a probabilistic reinforcement learning task by for adolescents and adults. Adapted from Davidow et al., 2016. ²²

Probabilistic reinforcement learning tasks also have demonstrated enhanced performance in adolescents relative to adults. For example, Davidow et al. ²² had adolescents (13−17 years) and adults (20−30 years) perform a probabilistic reinforcement task in which participants had to learn which of two flowers different butterflies preferred. The association between cues (flowers and butterflies) and outcomes was probabilistic, which required the continual use of reinforcement/feedback to update choices. The reinforcement was simply a feedback screen with “correct” or “incorrect.” The adolescents showed enhanced learning relative to adults with an overall higher percentage of optimal choices (preferred flower by the butterflies) by the end of the experiment (Figure 1B). These findings highlight an enhanced ability of the adolescent to learn about probabilities of events and rewards in the environment that influence choice and goal‐directed behavior.

Other reinforcement learning studies have examined how changing reward contingencies (e.g., reversal learning) impacts adolescent behavior. ³⁸ , ³⁹ The rationale for these studies is that reversal learning is likely an important ability for adolescents given it is a time when rapid learning is needed in an often‐changing environment. Using a similar probabilistic reinforcement task as described above, Eckstein et al. ³⁹ instructed participants 8−30 years of age that a coin reward was hidden in one of two boxes. One of the boxes was associated with a 75% reward probability. However, in this task, the reward contingencies switched several times after a number of successful rewarded trials. Accuracy in choosing the rewarded box during the reversal conditions was higher in teens than in children or adults. This ability to quickly alter behavior as rewards shifted was interpreted as reflecting greater goal‐relevant learning in a rapidly changing environment (i.e., exploratory behavior).

Similarly, in another reversal learning task, Van Der Schaaf et al. ³⁸ simultaneously presented two stimuli (a face and a scene) with one of the two stimuli outlined with a black border. The participants’ task was to predict whether the outlined stimulus was followed by a reward or a punishment, after which the actual outcome was presented. The solution was 100% deterministic. After several trials, the stimulus‐outcome associations were reversed, signaled by either an unexpected reward or an unexpected punishment. Adolescents relative to younger and older participants showed enhanced performance during the reversal trials (Figure 2A). This quadratic pattern was more evident with unexpected punishment (Figure 2C) than with unexpected reward (Figure 2B). A greater sensitivity to unexpected reward and especially unexpected punishment may serve to drive a change in behavior to explore new choices and outcomes in a changing environment. Together, these latter studies show enhanced learning and better adaptation to volatile changes in reward contingencies in adolescents than adults. These aspects of learning may be the very abilities adolescents need when exploring new environments on their own.

FIGURE 2.

Age effects on reversal learning. (A) Quadratic trend between proportion correct on reversal scores and age (F _(3,54) = 19.45, p < 0.001). (B) Linear and quadratic trend between accuracy on reward‐reversal trials and age (linear: F _(3,54)  =  16.05, p < 0.001, quadratic: F _(3,54)  =  9.73, p < 0.003). (C) Quadratic trend between accuracy on punishment‐reversal trials and age (F _(3,54)  =  18.4, p < 0.001). Adapted from Van Der Schaaf et al., 2011. ³⁸

Finally, there is evidence of adolescent‐specific changes in memory as a function of reward. Cohen et al. ²³ used a reward‐motivated memory association task in which participants were cued as to whether remembering a pair of images would help them win a high or low bonus reward of $15 or $1, respectively. A unique object was then presented and paired with either a face or a place stimulus. Twenty‐four hours later, adolescents, unlike children and adults, showed better subsequent general source memory for the category of stimuli paired with each object (i.e., whether the object was paired with a face or place) when there was the potential of a high reward bonus ($15). They showed worse memory for the category with which an object was paired when there was the potential of only a small reward bonus ($1) (Figure 3). This effect was driven by adolescents more selectively prioritizing specific high‐reward associations in memory and not retaining more general, low‐reward information. This study illustrates how adolescents use reward value in goal‐directed learning and memory.

FIGURE 3.

Low‐ and high‐reward associative memory by age. General source memory performance shows a nonlinear relationship with age, with a peak in high‐reward source memory in light gray and a trough in low‐reward source memory in dark gray during adolescence. Adapted from Cohen et al., 2022. ²³

Together, these studies suggest a heightened sensitivity to rewards during adolescence that influences learning, memory, and behavior. We provided examples of how the value of information and choices were associated with enhanced goal‐directed learning and memory in adolescents relative to adults. We also provided evidence of a heightened sensitivity to both gains and losses in reward during this developmental phase that was associated with enhanced learning under volatile reward contingencies—an ability that may benefit the organism when independently exploring new and changing environments. In fact, the enhanced value‐related learning, reversal learning, and response inhibition to previously reinforced cues together likely play a role in learning from and exploring changing environments for the adolescent.

Influence of peers

There is a well‐documented influence of peers on adolescent risky and cognitive behavior. ²⁵ , ⁴⁰ , ⁴¹ , ⁴² Here, rather than focusing on the many complex facets of social cognition and behavior during adolescence, we instead focus on how a social cue or the mere presence of peers can influence cognition and behavior. For example, even simple tasks that use static positive social cues (e.g., smiling faces) can alter adolescent behavior. To illustrate this point, we highlight our study ⁴³ using a go/no‐go task and positive and neutral social cues as both targets and nontargets. We found that adolescents, unlike children and adults, were significantly more impulsive with the positive than the neutral faces even though they were instructed to ignore them (Figure 4A).

FIGURE 4.

Impulsive errors to social emotional cues on a go/no‐go task by age group. (A) Adolescents show more impulsive errors to positive versus neutral social cues than do children or adults (adapted from Somerville et al., 2011 ⁴³ ). (B) Adolescents show more impulsive errors to social cues of potential threat versus neutral social cues than do children or adults. Adapted from Dreyfuss et al., 2014. ²⁹ * p < 0.05.

The above study shows how the simple use of positive social cues can impact adolescent behavior, but what about when there is a real social interaction? Peer presence, especially when the peer(s) indicate a risk preference, has been shown to influence adolescent decisions and cognitive performance. ²⁵ , ⁴⁰ , ⁴¹ , ⁴² In one of the first studies to examine in‐person peer influences on adolescent decision‐making, Gardner and Steinberg ⁴⁰ tested over 300 teens (13−16 years old), college undergraduates (18−22 years old), and adults (24 and older) who were randomly assigned to engage in a simulated driving task alone or in the presence of two friends. Adolescents and to a lesser extent the undergraduates took a greater number of risks when observed by peers—a pattern not observed in adults. This finding has been replicated in numerous studies across labs and cultures. ⁴⁴

It is important to note that peers are not the only individuals who can influence adolescent decision‐making. We and others have shown that adolescents’ decisions can be influenced by unknown adults and parents in negative and positive ways. ⁴⁵ , ⁴⁶ , ⁴⁷ Likewise, peers can influence adolescent behavior in positive ways. ²⁶ For example, Sullivan et al. ⁴⁸ found that when playing a game in which adolescents could allocate monetary rewards to self or other, adolescents were more selfish when alone, but in the presence of a peer became more altruistic. Moreover, mouse tracking of real‐time decisions suggested that the peer's presence sped the prosocial decision. As such, sensitivity to peers and others can facilitate socialization and positive responses toward others.

Influence of threats

Another significant focus of adolescent research has been on how adolescents respond to potential threats in the environment. Often, these studies have focused on responses to fearful facial expressions or aversive scenes. These studies show adolescents to be more reactive and impulsive to these cues relative to adults ²⁷ , ²⁹ and more distractible by them as evidenced by longer reaction times suggestive of interference effects. ²⁸ , ⁴⁹ We illustrate this developmental pattern in children, adolescents, and adults who completed a go/no‐go task with fearful and neutral facial expressions as stimuli. Adolescents, unlike children and adults, made more impulsive errors with nontarget fearful faces in comparison to nontarget neutral faces (Figure 4B).

A similar result was observed using negative, neutral, and positive scenes from the International Affective Picture System that served as the background ⁵⁰ for a go/no‐go task in participants 11−25 years old. ²⁷ While all ages showed longer reaction times when the background image was negative, age differences in accuracy on inhibitory trials suggested that response inhibition is more readily disrupted by negative emotional distraction in adolescents relative to younger and older participants. Together, these findings suggest a heightened sensitivity to cues and scenes of potential threat during adolescence, as evidenced by greater distraction and impulsivity to them at this age than in younger or older ages.

A problem with using facial expressions—and even aversive scenes—to examine developmental differences in threat responses is that different age groups have different experiences with faces and scenes (i.e., older individuals presumably have more experience with them than younger individuals). A solution to this potential confound is to utilize classical conditioning paradigms that involve pairing a neutral stimulus such as a tone or colored square with an aversive stimulus (e.g., shock, loud noise). After only a few pairings of these two stimuli, the neutral stimulus begins to take on qualities of the aversive stimulus in that it predicts the occurrence of it (i.e., conditioned stimulus, CS+) and results in a fear response such as freezing behavior in rodents or heightened arousal in humans as measured by galvanic skin conductance (conditioned response). With repeated presentation of the conditioned stimulus alone without the aversive stimulus, the fear memory is extinguished, resulting in a decrease of the fear response.

Using this approach, we examined the development of cued fear acquisition and extinction learning in children 5−11 years old, adolescents 12−17 years old, and adults 18−28 years old. ³⁰ A loud aversive sound was paired with either a yellow or blue colored square and arousal was measured by the galvanic skin conductance response (SCR) to the square paired with the aversive stimulus (e.g., blue square; CS+) compared to the colored square (e.g., yellow square) never paired with the aversive stimulus (CS−). Children, adolescents, and adults alike showed similar differential SCR responses to the CS+ versus the CS−, suggesting relatively equivalent fear acquisition.

Given evidence of equivalent cued fear learning across age groups, we then presented the CS+ repeatedly without the aversive stimulus to extinguish the cued fear memory. Surprisingly, adolescents, unlike children and adults, showed less fear extinction than did children and adults (Figure 5A). Other groups have replicated this finding of diminished cued fear extinction in adolescents. ⁵¹ , ⁵²

FIGURE 5.

Cued fear acquisition and extinction in humans and mice by age group. (A) Relative to children and adults, adolescent humans show diminished cued fear extinction learning as indexed by the difference in skin conductance responses (SCRs) to the CS+ relative to the CS−. (B) Adolescent mice postnatal day (P) 29 show diminished cued fear extinction learning to the CS+ as indexed by freezing behavior relative to preadolescents (P23) and adults (P70). Differential SCR is the difference in the skin conductance response to CS+ versus CS−. ***p < 0.001 Adapted from Pattwell et al., 2012. ³⁰

Together, these studies on the influence of rewards, peers, and threats on behavior show adolescent‐specific developmental changes. These changes include heightened reactivity and learning in new and changing environments that may facilitate exploratory goal‐directed behaviors.

HOW HAVE THESE BEHAVIORS EVOLVED ACROSS SPECIES?

Understanding similarities and differences across species can provide new insights about how certain behaviors may have evolved, how they may be adaptive, and what the potential mechanisms underlying the behaviors are. Many of the paradigms used in humans described above are largely based on elegant animal work, given the conservation of core circuitry across species. ²⁰ , ²² , ³⁰ The beauty of examining behavior across species when asking developmental questions, especially in rodents, is that these species develop much faster than humans—in a matter of weeks to months as opposed to years to decades. However, a potential problem with such studies is that the behavior and paradigm must be simple enough for the rodent to be able to elicit it or learn it before maturing into a different phase of development. Simple reward and aversive learning paradigms have been used for this purpose. Below, we highlight examples of such studies to illustrate the influence of rewards, peers, and threats on adolescent behavior across species.

Influence of reward

Adolescent sensitivity to rewards appears to be conserved across mammalian species (e.g., monkeys, rats, mice), which also show patterns of reward‐related behavior similar to those of humans. ³ , ³¹ For example, rats show a similar inverted U‐shaped developmental trajectory in reward and novelty‐seeking behaviors ⁵³ and risk‐taking ⁵⁴ as that observed in human adolescents. ³¹ , ⁵³ There is also a heightened sensitivity to rewards that can influence their learning and behavior.

For the sake of illustration, we highlight a reward learning study in adolescent rats. In this study, Sturman and colleagues compared the performance of adolescent (postnatal days 28–42) and adult (postnatal day 60+) rats in reinforcement learning and extinction. The rats were trained to poke their nose into a hole whenever a cue (light) appeared by the hole to receive a food‐pellet reward. Then, the animals underwent extinction sessions in which nose poking to the light was no longer reinforced. Adolescent and adult rats’ performance was similar, but adolescents appeared to acquire the reward‐related learning more quickly than adults as indexed by faster latencies to poke to the light in early training sessions and in early epochs of those training sessions (Figure 6A). This pattern of behavior is somewhat reminiscent of enhanced learning in human adolescents that exceeded that of adults ²² during a probabilistic reinforcement learning. While the rodent and human tasks are clearly different, they still illustrate heightened responses to rewards.

FIGURE 6.

Reward learning and extinction in the adolescent and adult rats. (A) Adolescent rats show faster latencies in hole‐poking to a light cue predicting reward than adults during the first 5 min of early training sessions. (B) During extinction training, adolescent rats compared to adult rats show more perseverative nose poking to a cue that no longer is predictive of reward. Adapted from Sturman et al., 2010. ⁵⁶ * p < 0.05.

Adolescent rats also showed extinction‐resistant instrumental learning ⁵⁵ , ⁵⁶ (Figure 6B). These findings are somewhat consistent with Davidow's ³⁵ findings of adolescents continuing to impulsively respond to previously reinforced cues in a response inhibition task—a pattern that remitted over the course of the task in adults. Taken together, these behavioral similarities in the influence of rewards on adolescent behavior in both humans and rodents suggest potential mechanisms for facilitating trial‐and‐error learning in a changing environment which likely facilitates exploratory behavior.

Influence of peers

Several studies have examined social behavior in adolescent rodents. They, like humans, show sensitivity in response to peers ⁵⁷ and the presence of peers increases certain risk behaviors. ³¹ To provide a surprising and telling illustration, Logue and colleagues ¹⁵ examined the influence of cage mate peers on adolescent drinking behavior. They raised different cohorts of mice in same‐sex triads and later tested their ethanol‐drinking behavior in a novel environment when they reached adolescence (between postnatal days 28 and 30) or adulthood (between postnatal days 84 and 86). Half of the animals in each age group were either tested alone or with their cage mates. Logue found that adolescent mice increased their drinking behavior when in the presence of peers (Figure 7). Adults did not show this behavioral pattern. These findings suggest that the effect of peer influence on human adolescent behavior may reflect an evolutionarily conserved process, whereby the presence of peers increases the adolescent's sensitivity to potential rewards (and their actions) in their immediate environment.

FIGURE 7.

Drinking behavior in adolescent and adult mice when alone or with cage mates (social). Adapted from Logue et al., 2014. ¹⁵ * p < 0.05.

Influence of threats

Just as threats influence human adolescent behavior, so too do rodent studies show similar adolescent‐specific effects. We highlight two studies below that examine cued and contextual fear learning and extinction. The first is part of the previously described developmental human study on cued fear learning and extinction. In that study, both developing humans and mice were tested ³⁰ (see Figure 5A for human results). In the mouse, cued fear learning was assessed in preadolescent (postnatal day 23), adolescent (postnatal day 29), and adult (postnatal day 70) mice. Following cued fear acquisition that involved repeated pairings of a tone with a foot shock in a conditioning chamber, the mice then underwent cued fear extinction during which time the tone was presented in a new context without the presentation of a foot shock. As observed in humans (Figure 5A), preadolescent and adult mice showed relatively good cued fear extinction, but adolescent mice showed less extinction learning (Figure 5B).

The finding of diminished cued fear extinction in adolescents has been replicated across several studies, especially in mice ³⁰ , ⁵⁸ , ⁵⁹ , ⁶⁰ , ⁶¹ , ⁶² , ⁶³ , ⁶⁴ but varies to some extent in rats and under different experimental conditions. ⁶⁵ Nonetheless, a greater re‐emergence of fear responses following extinction in human and mice adolescents relative to younger and older ages may help them remain vigilant to potential threats when exploring new environments independently.

The ability to predict and detect potential threats in the environment is essential for survival, but it is also important to know in which environments a cue is a threat (e.g., bear in the woods) or not a threat (e.g., bear in the zoo). ⁶⁶ Contextual fear relies on the ability to learn about environments associated with threat, whereas cued fear relies on learning about cues within an environment that are associated with threat. In another developmental study of mice, we examined contextual fear in preadolescent (postnatal day 23), adolescent (postnatal day 29), and adult mice (postnatal day 70). The mice were placed in a conditioning chamber and exposed to several foot shocks. ⁶⁷ The next day, the mice were placed back in the conditioning chamber, but no shocks were administered. In contrast to our findings of similar cued fear acquisition in preadolescents, adolescents, and adults, results of the contextual fear conditioning study show no evidence of contextual fear learning in the adolescent mice (i.e., little if any freezing), while preadolescent and adult mice showed evidence of contextual fear (Figure 8). ⁶⁷

FIGURE 8.

Contextual fear learning in mice by age. Adolescent mice (P29–P39) froze significantly less than both younger (P23–P27) and older (P49–P70) mice. * p < 0.05. P, postnatal day. From Pattwell et al., 2011. ⁶⁷

Given this surprising finding, several cohorts of mice of different ages were tested, resulting in the identification of a sensitive window between postnatal days 29−39 when adolescents showed diminished freezing behavior relative to younger and older mice. In contrast to our finding of extinction‐resistant cued fear in adolescents, their expression of contextual fear appears absent (i.e., little to no freezing behavior following conditioning in adolescent mice). Serendipitously, Pattwell later discovered that the same mice that showed no contextual fear as adolescents showed contextual fear as adults. So, it is not that the animals did not learn, but rather that they did not express contextual fear as adolescents. ⁶⁷

Together, the above studies on the influence of reward, peers, and threat highlight similarities in adolescent behavior across species suggesting that they have evolved in an analogous way to respond to the challenges of this developmental phase of life. The similarities in behavior relate to the conservation of key brain circuitry underlying these behaviors. Next, we highlight developmental changes in brain circuitry and networks during adolescence that may help to explain these behaviors.

WHAT ARE THE POTENTIAL MECHANISMS UNDERLYING ADOLESCENT BEHAVIOR?

The brain shows a remarkable ability for changes throughout the lifespan, including learning and adaptation to new environments. This potential for change is especially present during the extended period of adolescence ⁶ , ⁶⁸ , ⁶⁹ , ⁷⁰ when connections between neurons are pruned or strengthened as a function of experience. This refinement in connections facilitates more efficient communication among networks of brain cells and regions that are important for processing and integrating cognitive, emotional, and social information. ¹⁰ , ⁷¹ , ⁷² The timing of these changes varies in different brain networks, with slow or protracted developmental changes in brain circuitry involving the prefrontal cortex, which has been implicated in decision‐making and impulse control, especially in emotionally charged situations. ⁷³ , ⁷⁴ In contrast, subcortical regions implicated in learning and processing of motivational, emotional, and contextual information (e.g., amygdala, ventral striatum, and hippocampus) show early structural, functional, and connectivity changes that appear to peak during this time (e.g., Figure 9) relative to the continued changes seen especially in lateral prefrontal control circuitry. ³² , ⁴³ , ⁴⁹ , ⁶¹ , ⁷⁵ , ⁷⁶

FIGURE 9.

Adolescent‐specific increases in brain responses to rewards and potential threats. (A) Change in MR signal to large versus small rewards in children (7−11 years), adolescents (13−17 years), and adults (23−29 years) (adapted from Galvan et al., 2006 ²⁰ ). (B) MR signal to cues of potential threat (fearful faces) in children (7−12 years), adolescents (13−18 years), and adults (19−32 years). Adapted from Hare et al., 2008. ⁴⁹ MR, magnetic resonance. * p < 0.05.

These changes in brain structure, function, and connectivity within and between networks of regions parallel significant cognitive, emotional, and social psychological development during this time. ⁷⁰ , ⁷² , ⁷⁷ The ongoing development of these regions and their connections with one another also parallels a heightened tendency for what appears to be risky or short‐sighted decisions and actions in emotionally charged situations (e.g., in the presence of rewards, peers, threats) that have characterized adolescents as a group. ¹¹ , ²⁴ , ⁷⁸ Yet, adolescents are quite good at making simple cognitive decisions when not under these conditions, ⁷⁹ , ⁸⁰ and we have highlighted several examples of adolescents showing enhanced choice behavior and learning and memory relative to adults that may in part rely more heavily on motivational rather than strategic or fully integrated networks.

Figure 10 provides a simplistic illustration of proposed changes in dopamine‐rich brain circuitry implicated in reward and threat responses with age and experience throughout the extended period of adolescence. These subcortical and cortical circuits are tightly interconnected, show hierarchical changes, and are implicated in learning and memory. Accordingly, changes in connections within subcortical circuits (in red) implicated in learning are presumed to drive subsequent changes in inputs to and connections with prefrontal cortical regions (in orange) which then alter projections back to subcortical regions (in green). Integration within and between these circuits based on age and experience is then thought to help instantiate cortico‐cortical circuit refinement of the brain (in blue). These hierarchical changes in dopaminergic‐rich brain circuitry are thought to increase the efficiency of how different circuits communicate with one another across the brain with age and experience. ⁸¹ This view is consistent with Thelen's notion of development as hierarchical, ⁸² with each developmental stage being dependent on the preceding one. Accordingly, the functionality of subcortical circuits may be a necessary precursor for signaling prefrontal cortical circuitry, which is then a necessary precursor for cortico−cortical interactions for more complex interactions among cognitive and affective processes. ⁸¹ , ⁸³ Importantly, Thelen conceptualized development as a process of reorganization that emerges due to the interaction of the system/organization/person with another and with the environment. As such, behavior can be biased not only by developmental changes in brain circuitry with age and experience, but also by the specific task or environmental context (i.e., rewarding, social, and threatening). Similarly, Wilbrecht and Davidow ⁵ elegantly make this point using a heuristic ballot‐box model in which the ability of different circuits to contribute to goal‐directed learning and behavior depends on the remodeling of these circuits during development as well as their recruitment under different task demands and contexts. While significant learning occurs during development, this learning continues across the life course and reflects the complex integration of new inputs with prior experiences across different contexts and neurocognitive systems—all of which influence goal‐directed behavior.

FIGURE 10.

Simplistic illustration of hierarchical development from subcortico‐subcortical to cortico‐cortical connections in dopamine‐rich circuitry with age and experience. Red lines and shading denote subcortical connections; orange denotes subcortical to cortical connections; green denotes cortical to subcortical connections; and blue denotes cortico‐cortical connections. Amyg, amygdala; dACC (prelimbic cortex in rodent), dorsal anterior cingulate cortex; hipp, hippocampus; lPFC, lateral prefrontal cortex; Striatum, ventral striatum; vmPFC (infralimbic cortex in rodent), ventromedial prefrontal cortex. Adapted from Casey et al., 2019. ⁸¹

Evidence for hierarchical changes in the brain during development comes from postmortem human and nonhuman primate studies showing selective stabilization and regional elimination of predominantly excitatory synaptic connections in the sensorimotor cortex before the prefrontal cortex. ⁸⁴ , ⁸⁵ Human neuroimaging studies have shown corresponding patterns of regional brain changes in cortical thickness of sensorimotor regions before prefrontal and other association cortical regions across development. ⁸⁶ Studies of the functional connectome (i.e., how brain regions coordinate their function, which is a signal of more efficient and sophisticated brain organization) similarly show evidence of hierarchical cortical development along a sensorimotor‐association axis ⁸⁷ , ⁸⁸ Likewise, as stated previously, changes in subcortical regions appear to peak earlier during adolescence (Figure 9), while association cortex including the prefrontal cortex shows extended development. ³² , ⁴³ , ⁴⁹ , ⁶¹ , ⁷⁵ , ⁷⁶ In parallel, regional increases in white matter are thought to occur through the myelination of axons, influencing the conduction of electrical impulses across the brain. ⁷ Regional structural and functional changes occur together with neurochemical changes. For example, dopamine—a neurotransmitter importantly implicated in motivation, learning, and memory—shows dramatic changes in innervation of prefrontal and striatal circuitry and function during this time. ⁸⁹ , ⁹⁰ , ⁹¹ , ⁹² Together, these studies suggest that regional changes in synaptic density, morphology, and related cortical cell firing, in myelination, and availability of neurochemicals and their receptors lead to transient imbalances in communication within and between networks of regions that impact behavior during adolescence. ⁷

Some of the most robust evidence for hierarchical subcortical to cortical development comes from tract tracing studies in developing rodents. These studies show, for instance, that projections that ascend from subcortical to cortical regions (e.g., amygdala to the prefrontal cortex) emerge earlier than projections that descend from the prefrontal cortex to the amygdala. ⁹³ , ⁹⁴ , ⁹⁵ , ⁹⁶ , ⁹⁷ Below, we illustrate changes in dopamine‐rich circuitry with age and experience to provide potential mechanisms for the changes in many of the behaviors we see during adolescence in the context of rewards, peers, and threats.

Development of reward‐related circuitry

A number of human imaging studies have shown enhanced reward‐related learning and memory in adolescents compared to adults. ²² , ²³ Davidow et al. ²² provide a nice example of the role of subcortico−subcortical functional coupling in this learning. As described previously, adolescents show enhanced probabilistic reinforcement learning relative to adults on their butterfly task (see Figure 1B). However, unlike adults, adolescents activated both the striatum and hippocampus during learning. Adults only activated the striatum. Importantly, greater functional coupling between these subcortical learning systems (striatum and hippocampus) was associated with better learning in adolescents. This functional coupling was also positively associated with a memory bias for incidental items presented on the reinforcement/feedback screen (i.e., greater memory when the feedback was positive than when negative). This latter finding is consistent with evidence of better paired‐associate source memory for high‐ versus low‐reward trials and coupling between the subcortical regions of the hippocampus and ventral tegmental area. ²³ Thus, subcortico−subcortical coupling is related to enhanced trial‐and‐error reinforcement learning and a positive reward bias on episodic and source memory during adolescence.

Evidence of an interplay between cortical and subcortical regions is seen in probabilistic reinforcement and reversal learning imaging studies. Adolescents specifically show better choice performance on probabilistic reinforcement tasks than adults ²² even when reward contingencies are frequently switched. ³⁸ , ³⁹ Findings from developmental human imaging studies ⁹⁸ show increased connectivity between the medial prefrontal cortex and the striatum with age that have been associated with better probabilistic reinforcement learning. When cues from such studies are subsequently used in tasks to assess response inhibition in the same participants, improved accuracy for previously learned high reward cues relative to low reward cues has been shown to emerge between early and late adolescence. ³⁶ Whereas younger adolescents show increased coupling between the striatum and ventromedial prefrontal cortex for higher gain cues, older adolescents show increased coupling between the striatum and dorsolateral prefrontal cortex for higher gain cues. Together, these studies suggest that what might change during adolescence is how reward value and learning signals are used to guide behavior and expectations when they conflict with a current goal or violate expectations. The imaging findings highlight the potential role of changes in subcortical and cortical connectivity with this value‐based learning during development.

Cortico‐subcortical circuitry has also been implicated in reversal learning. Human imaging and nonhuman animal studies in adults have implicated the ventral lateral prefrontal cortex in reversal learning. Findings from rodent and nonhuman primate studies show deficits in this ability following damage to this region, ⁹⁹ while evidence from human functional magnetic resonance imaging (fMRI) studies likewise show the importance of the lateral ventral prefrontal cortex (PFC) as well as the ventral striatum. ¹⁰⁰ Given that human neuroimaging studies have emphasized a role of the lateral ventral prefrontal cortex and its development in response inhibition, ¹⁰¹ , ¹⁰² , ¹⁰³ changes in this region during reversals may reflect behavioral inhibition of the previously learned response, whereas ventral striatum activity may reflect the rapid learning and updating of new associations. ¹⁰⁰ Given the heightened sensitivity to rewards during adolescence in the ventral striatum, ²⁰ , ³² the enhanced reversal learning during this developmental period may be more a product of value‐related learning than behavioral inhibition. More evidence is needed to determine the exact interplay between cortical and subcortical connectivity in mediating flexible goal‐directed behavior in volatilely changing environments. However, enhanced value‐related learning, reversal learning, and response inhibition all likely play a role in learning from and exploring changing environments to gain resources, from a pull toward incentives to overriding of responses based on prior value‐learning when reward continencies change and/or are reversed. The findings from human and nonhuman animal studies suggest an important role in the development of dopamine‐rich cortico‐subcortical circuitry and functional connections in this learning and exploration.

An example of changing responses to potential rewards with the emergence of cortico‐cortical connectivity by young adulthood is illustrated by the previously described study of response inhibition to a previously rewarded cue. ³⁵ While the influence of a rewarded cue on inhibitory control during a go/no‐go task increased with age (i.e., more impulsive responses to a previously rewarded cue when it was a nontarget/no‐go stimulus), this pattern of performance remitted in young adults over the course of the task, but not in adolescence. Accurate no‐go performance to the previously rewarded cue was associated with an age‐related increase in the lateral prefrontal cortex—ventromedial prefrontal cortex connectivity. These results reflect an increase in the coordination of distinct higher‐order cortical regions within the prefrontal cortex when goal‐directed behavior conflicts with reward value with age.

Together, these findings illustrate hierarchical changes in and influences of developing dopamine‐rich subcortical and cortical systems and connectivity in goal‐directed behavior with learning and development. They also highlight the importance of task demands and contexts (i.e., rewarded versus nonrewarded cues, reversal versus nonreversal learning, predicted versus unpredicted outcomes) on brain and behavior.

Development of circuitry implicated in peer interactions

Although the developmental cognitive neuroscience literature on adolescent social cognition is rapidly growing, for the purposes of this article, we focus on how a simple social cue (facial expression) or the mere presence of a peer influences goal‐directed behavior and underlying brain circuitry.

As described previously, we showed that the simple presentation of positive or negative social cues (e.g., happy or fearful faces) during a go/no‐go task increased impulsive actions in adolescents relative to children and adults. ²⁹ , ⁴³ , ¹⁰⁴ In these same studies, we examined neural correlates of this behavior. We showed heightened activity and functional connectivity in the subcortical regions of the ventral striatum and amygdala to these social cues relative to neutral ones. ⁴³ , ⁴⁹ , ¹⁰⁴ However, increasing age and functional coupling of cortical and subcortical regions of the medial prefrontal cortex and amygdala were associated with a decrease in impulsive actions in response to these social cues relative to neutral ones. ¹⁰⁴ Together, these results suggest the relevance of subcortical systems in reactive responses to social cues during adolescence and suppression of these actions with the development of cortico‐subcortical regions and functional connections.

Important for understanding the influence of peers on goal‐directed behavior is to examine how their presence impacts behavior and decision‐making. As described earlier, adolescents engage in more risky decisions when performing a simulated driving task when a peer was watching than did adults. ⁴⁰ Chein and colleagues ²⁴ used a similar task to examine the neural correlates of this risky behavior. They found greater activity in reward‐related areas of the ventral striatum and orbitofrontal cortex of adolescents when performing the task in the presence of a peer. In contrast, adults who performed well regardless of the social condition, activated regions implicated in cognitive control (e.g., lateral prefrontal cortex) when performing the task regardless of whether a peer was present or not. Together, these findings illustrate the influences of earlier developing subcortical versus cortical control systems in value‐based choice behavior and goal‐directed behavior, but they also illustrate the importance of context (i.e., social versus nonsocial cues and conditions) on adolescent brain and behavior.

Development of threat‐related circuitry

While human functional connectivity studies are correlational and do not provide directional information (i.e., cannot differentiate between subcortico‐cortical or cortico‐subcortical projections), animal studies can. The role of developmental changes in subcortico‐cortical and cortico‐cortical connections in fear learning and extinction in mice can be measured. ³⁰ , ⁶⁷ As described previously, cued fear memories in adolescent humans and mice appear resistant to extinction (Figure 5), while contextual fear learning is not evident during adolescence (Figure 8). Figure 11 provides a simplified illustration of fear circuitry that includes a subset of regions from our original hierarchical model implicated in cued and contextual fear memory (Figure 10). At the subcortical level, fear expression involves the basolateral amygdala that detects threat and projects to the central nucleus of the amygdala, which projects to the hypothalamus and autonomic nervous system to produce the fear response (fight, flight, or freeze). Fear expression is maintained via recursive projections between subcortical and cortical regions of the basolateral amygdala and the prelimbic cortex. Fear extinction involves cortico‐subcortical projections from the infralimbic cortex to inhibit the basolateral and central nucleus of the amygdala which in turn inhibits the fear response. Finally, the hippocampus is thought to gate the fear response via projections to the basolateral amygdala (fear expression), prelimbic cortex, and infralimbic cortex (fear regulation) based on the context or situation (i.e., threat when you see a bear in the woods versus safe when you see a bear at the zoo). ⁶⁶

FIGURE 11.

Illustration of developmental changes in fear circuitry during adolescence. During adolescence, there is a surge in inputs to the prelimbic cortex (PL) from the basolateral amygdala (BLA) and ventral hippocampus (solid gray lines) in rodents, while connectivity between the infralimbic cortex (IL) to the BLA and to the central nucleus of the amygdala (CE) is weaker than in adults (dotted gray lines). Adapted from Casey et al., 2025. ¹¹¹

During adolescence in mice, there is a surge in subcortico‐cortical projections from the basolateral amygdala and ventral hippocampus to the prelimbic cortex, as measured using microprisms, to image prefrontal cortical spine maturation, and retrograde labeling across preadolescence, adolescence, and adulthood. ⁶¹ This surge in connections has been associated with the opposing behaviors observed in cued and contextual fear during adolescence. Specifically, the surge in basolateral amygdala inputs to the prelimbic cortex is associated with the maintenance of a fear response (i.e., resistance in cued fear extinction) reflecting a recurrent loop of activation. This increase in amygdala‐prelimbic connections together with weak infralimbic‐amygdala connectivity during adolescence are together thought to underlie the persistent cued fear response in adolescents following extinction. ³⁰ , ⁶¹ In contrast, the surge in inputs from the ventral hippocampus to the prelimbic cortex has been associated with gating the amygdala‐prelimbic recursive circuit, resulting in less fear expression during hippocampus‐based contextual learning during adolescence. ⁶¹ , ⁶⁷

Cortico‐subcortical projections also influence the fear response and its regulation and show developmental changes during adolescence. ⁹³ , ⁹⁷ , ¹⁰⁵ , ¹⁰⁶ The majority of human developmental imaging studies on fear regulation have focused on the human analog of the infralimbic cortex—the ventromedial prefrontal cortex. ¹⁰⁷ , ¹⁰⁸ Ventromedial prefrontal‐amygdala functional connectivity shows significant changes from childhood to adulthood with a shift from positive to negative functional coupling between these regions, with negative coupling beginning during adolescence. This shift in connectivity has been associated with decreasing subcortical activity to threat‐related cues ¹⁰⁸ , ¹⁰⁹ and is consistent with rodent studies showing earlier emergence of ascending projections (i.e., from the amygdala to infralimbic cortex) relative to descending connections (i.e., from infralimbic to the amygdala) involved in fear regulations (e.g., cued fear extinction). ⁹³ , ⁹⁷ As illustrated in Figure 11, weak infralimbic‐amygdala connectivity during adolescence in humans and mice is associated with diminished fear extinction. ³⁰ , ¹¹⁰ , ¹¹¹ Together, these studies highlight that goal‐directed learning is not only influenced by the development of subcortical and cortical frontolimbic regions and connections but also by the task conditions and demands (i.e., cued versus contextual fear; acquisition versus extinction).

Together, the findings suggest adolescent‐specific changes in the brain that are associated with altered responses to rewards, peers, and threats. The question from an evolutionary developmental perspective is why would the brain be programmed this way? Below, we speculate on how these changes are adaptive and may facilitate the organism's survival.

HOW MIGHT ADOLESCENT BEHAVIOR BE ADAPTIVE?

Adolescence is a time that requires exploration beyond the home, nest, or territory to find and secure food and water relatively independently and to establish new social bonds beyond the family to facilitate the survival and reproduction of the species. ⁵ This exploration may be daunting, and it requires travel across long distances, the need to find new or additional sources of nourishment, and has the potential for exposure to predators. Likewise, this exploration may lead to new opportunities for more abundant resources, formation of exciting new social bonds, and attraction to and/or of a potential mate. As such, this is a time of tremendous learning and opportunity for the adolescent. Here, we synthesize findings from the highlighted studies and suggest that a heightened sensitivity to socially relevant cues (rewards, peers, and threats) together with a decreased apparent fear of new and potentially threatening contexts may be ideal mechanisms for meeting some of the many developmental challenges that adolescence brings. ⁶

Influence of rewards

To begin, a mechanism for pulling the adolescent out of the safety and familiarity of the home would seem essential for their moving out and exploring new environments. As such, a heightened sensitivity to rewards as shown in the effects of the immediacy of rewards on action may serve as that pull mechanism. The enhanced and persistent reward learning ³⁵ , ⁵⁵ , ⁵⁶ even following extinction may help the adolescent maintain memory for previous locations of potential reward if/when new sources of resources fail to produce the desired outcomes and to further sustain goal‐directed learning during exploration. Finally, the ability to quickly alter behavior as expected rewards do not occur or unexpected punishments occur, as evidenced in reversal learning studies, ³⁸ , ³⁹ may in fact facilitate exploratory behavior even further. Thus, the sensitivity to rewards, enhanced value‐related learning, reversal learning, and response inhibition to previously rewarded cues observed across studies all likely facilitate dispersion and finding and acquiring basic resources.

Influence of peers

Another adaptive aspect of exploration is not only obtaining food and water but also in forming new social bonds beyond the family or niche and attraction to or of a potential mate. The sensitivity to peer influence seen across species ¹⁵ , ²⁴ suggests an evolutionarily conserved process, whereby the presence of peers increases the adolescent's sensitivity to potential rewards and outcomes. The heightened reactivity as evidenced in impulsivity to social cues (smiling) ⁴³ may facilitate agency in approaching and interacting with others. The influence of peers also appears to serve a socialization role as evidenced by Sullivan et al., ⁴⁸ who found adolescents to be more altruistic and less selfish with resources in the presence of peers versus when alone.

Influence of threats

Unfortunately, exploring new environments can bring risks and threats (e.g., predators). As such, the ability to detect potential threats in the environment to avoid or defend oneself against is critical for survival. The impulsivity observed during adolescence to cues of potential threat ²⁷ , ²⁹ implies a reactiveness that may be associated with fight, flight, or freezing that is important for survival, especially given that a caregiver is less likely to be there to help protect them as they may have been when they were younger. When exploring new and changing environments, it may be adaptive to sustain vigilance for potential cues of threats or predators as seen in the diminished cued fear extinction. ⁶⁷

The observed lack of fear in previously threatening contexts during adolescence ⁶⁷ may seem maladaptive at first glance, but it may serve an adaptive function. It may prevent the animal from stopping further exploration after a negative encounter. Thus, there may be evolutionary mechanisms related to threat as well as those discussed with regard to rewards that help to facilitate exploratory behavior in adolescents. The sensitive window of development when contextual fear is not expressed is during early adolescence (P29−39), before animals are reproductively mature. The subsequent emergence of fear for previously threatening contexts in these same animals as they begin to reach sexual maturity (P49−70) likely serves an adaptive function too. Such memories could potentially prevent these animals when they are older from entering or building a nest or home for their young or themselves in dangerous environments.

Together, these behaviors facilitate learning through exploration and social interaction that can promote the survival of the individual and species. However, exploratory behavior in unpredictable situations or those that are violent, abusive, or coercive may potentially lead to devastating and even fatal outcomes for adolescents, as evidenced by increased mental illness, ⁷ criminal behavior, ⁸ and mortality ⁹ during this time.

Throughout this paper, we have illustrated how experience and learning intersect with brain development to promote adaptive behavior during adolescence across different contexts.  Our approach was not to delve deeply into contrasting results of developmental studies but rather to use the highlighted studies as exemplars for our key themes. Other recent reviews have more closely examined the myriad of factors that contribute to the variability in results that have emerged across developmental studies (e.g., Nussenbaum and Hartley, ¹² Galván, ¹ and Baker et al. ³ ). These authors note that although mixed results arise for various reasons, the role of context seems to best encapsulate and explain variation in findings.  For example, a multitude of rearing environments and earlier‐life experiences likely contribute to individual‐level developmental processes that lead to vast variation in trajectories of brain development and maturation. ¹ , ¹¹² Likewise, as described throughout this article, different contexts involving rewards, peers, threats (or even simple experimental manipulations) can bias learning and behavior and likely reflect the integration of new inputs with prior experiences across different contexts and neurocognitive systems with age. As experiences change with age and across the life course, the brain accordingly shifts in response to experience‐dependent learning and behavior.  As discussed at the beginning of this article, neuroscience research demonstrates that patterns of brain function not only change across development, but within individuals and within individuals under different contexts. ¹⁰ , ¹¹ As such, any one snapshot in time of any one individual will not wholly be generalizable for understanding the entire group of individuals or set of studies. ¹

CONCLUSIONS

We began this article by suggesting that the adolescent brain is beautifully orchestrated to facilitate learning during a developmental period of increasing challenges and opportunities. Using an evolutionary developmental lens, we provided examples of how the adolescent brain has evolved exquisitely across species to meet these demands. We highlight several examples of adolescent‐specific changes in response to emotionally and socially relevant cues and outcomes (e.g., rewards, peers, and threats) that we suggest are essential for learning to negotiate one's own world, survive, and facilitate the thriving of the individual and species. These include not only learning to independently secure resources such as food and water for their survival but also establishing new social bonds important for the reproduction and survival of the species. ⁵ , ⁹ , ¹⁷ , ¹⁸ It seems reasonable given that social status and socialization play an important role in securing resources, survival, and reproduction of the species that neurocognitive mechanisms have evolved for the detection of rewards, social cues, and threats in the environment. Likely, these mechanisms serve a critical function in the successful transition of the organism from dependence on the parent/caregiver to relative independence as an adult. ⁹ As such, we argue that these changes represent adaptive rather than deviant development of the adolescent brain.

AUTHOR CONTRIBUTIONS

B.J.C., A.O.C., and A.G. equally contributed to the conception, writing, and editing of this article.

COMPETING INTERESTS

The authors declare no competing interests.

PEER REVIEW

The peer review history for this article is available at https://publons.com/publon/10.1111/nyas.15314.

Casey, B. J. , Cohen, A. O. , & Galvan, A. (2025). The beautiful adolescent brain: An evolutionary developmental perspective. Ann NY Acad Sci., 1546, 58–74. 10.1111/nyas.15314