Revisiting adolescence as a sensitive period for sociocultural processing

Abstract

Waves of research and public discourse have characterized adolescence as periods of developmental risk and opportunity. Underlying this discussion is the recognition that adolescence is a period of major biological and social transition when experience may have an outsized effect on development. This article updates and expands upon prior work suggesting that adolescence may be a sensitive period for sociocultural processing specifically. By integrating evidence from developmental psychology and neuroscience, we identify how trajectories of social and neurobiological development may relate to adolescents’ ability to adapt to and learn from their social environments. However, we also highlight gaps in the literature, including challenges in attributing developmental change to adolescent experiences. We discuss the importance of better understanding variability in biology (e.g., pubertal development) and cultural environments, as well as distinguishing between sensitive periods and periods of heightened sensitivity. Finally, we look toward future directions and translational implications of this research.

Introduction

Adolescence is widely recognized as a period of tremendous flux in physical, cognitive, and socioemotional development with consequences that persist into adulthood and beyond. Sensitive periods provide a theoretical framework for relating experiences during a particular developmental period to long-term developmental outcomes. Although often discussed in relation to infancy and early childhood, an influential review by Blakemore and Mills (2014) integrated findings from psychology and neuroscience to suggest that adolescence is a sensitive period for sociocultural processing. At the time of its writing, developmental cognitive neuroscience was a nascent field, with the authors themselves noting the need for further evidence to establish this theory — particularly in humans (p. 191). Along with the field’s maturation over the past decade has come greater methodological sophistication, larger sample sizes, and the rise of interdisciplinary and collaborative team science approaches. Given these advances, the present article provides a necessary update and expansion upon Blakemore and Mills’ (2014) regarding whether adolescence is a sensitive period for sociocultural processing.

More precise and critical examinations of theory in adolescent developmental cognitive neuroscience that move beyond identifying supportive, consistent findings are needed not only to advance research but also to tangibly improve the lives of young people (Pfeifer & Allen, 2015). As this requires theoretical clarity, we first outline definitions and criteria for evaluating the assertion that adolescence is a sensitive period for sociocultural processing. During this process, we highlight contemporary debates regarding how sensitive periods are defined, and also provide a much-needed definition of sociocultural processing. Next, we review more recent literature in consideration of these criteria, including domains that have historically been associated with the question of whether adolescence is a sensitive period for sociocultural processing as well as additional relevant areas of study. Finally, we synthesize recent findings across the field and discuss translational implications. As experiences during adolescence build on those from the first decade of life to shape health and well-being into adulthood (Sawyer et al., 2012), the science of sensitive periods in adolescence may inform how we can best support positive developmental outcomes for youth globally.

Defining Sensitive Periods

A sensitive period is a limited time window during development when the effect of experience on the brain’s structure or function is powerful and lasting (Hensch et al., 2005; Will et al., 2008). Sensitive periods do not preclude the possibility of further consequences after this window closes. In contrast, the related term “critical period” implies that subsequent effects are impossible or highly limited; as this is not established for sociocultural processes in adolescence, we use the term “sensitive periods” throughout.

Mechanistic versus functional evidence for sensitive periods

Some mechanistic perspectives within psychology and neuroscience define and identify sensitive periods primarily through evidence for a common suite of neurobiological processes that up- and downregulate plasticity during development. Mechanisms of developmental plasticity are described as “experience-expectant”, meaning that they are posited to support efficient learning relevant to survival in light of species-typical experiences across members of a species for (Greenough et al., 1987; Nelson & Gabard-Durnam, 2020). In contrast, “experience-dependent” mechanisms support general, ongoing learning throughout the lifespan and are thought to be more gradual and easier to reverse. Noting distinctions between experience-expectant and experience-dependent mechanisms is helpful in many contexts. For instance, knowledge of mechanisms regulating the timing of plasticity suggests that sensitive period timing is flexible rather than fixed to specific ages (Bavelier et al., 2010; Gabard-Durnam & McLaughlin, 2020). Further, knowledge of general and ongoing experience-dependent mechanisms explains how experiences in adulthood can also have lasting impacts on the brain and behavior but remain distinct from sensitive periods (Jones & Jefferson, 2011). Additionally, impacts of experience during sensitive periods might still be retrained or augmented later in life, albeit through substantially more effortful and intentional processes (e.g,. early language acquisition versus structured, intensive efforts at second language learning).

The experience-expectant vs. experience-dependent distinction also has limitations. First, the difference between the two types of mechanisms is not always clear-cut, especially across species. For example, some species exhibit multiple, highly flexible sensitive periods, while others are more strongly impacted by the environment during key windows of development, but do not have nervous systems at all (Frankenhuis, 2020). In these and other cases, the experience-expectant vs. experience-dependent distinction may create communication barriers with researchers in related disciplines, such as evolutionary biologists. Second, some suggest that a suite of plasticity-enhancing and braking neural mechanisms may not be needed to achieve all sensitive period-like effects. Instead, the cumulative impact of general, ongoing mechanisms may produce patterns where information from the environment has an outsized impact on a system at specific points in time (Arcaro et al., 2019). Third, some argue that characterizing what experiences constitute an expected, species-typical childhood environment has been subject to bias from contemporary perspectives within rich and industrialized nations (Frankenhuis & Amir, 2022).

Functional evidence for sensitive periods is common in developmental psychology, but also in fields such as psychiatric epidemiology (Ben-Shlomo & Kuh, 2002, Schaefer, Cheng, & Dunn, 2022) and evolutionary anthropology (Frankenhuis & Fraley, 2017). Seeking functional evidence for sensitive periods involves identifying the outsized and lasting impact of experience during certain times in development. Research focused on seeking functional evidence for sensitive periods does not directly identify their mechanistic underpinnings but may nonetheless be complementary. Functional evidence may facilitate interdisciplinary insights and translational impacts for sensitive periods research by helping us understand how the timing of experience relates to cognition, behavior, and health.

Sensitive periods versus periods of heightened sensitivity

In considering the evidence for adolescence as a sensitive period, we suggest a need to distinguish between sensitive periods and periods of heightened sensitivity. Social reorientation theory indicates that the focus of social engagement shifts away from caregivers and toward peers and prospective romantic partners in adolescence (Nelson et al., 2005; Nelson et al., 2016). Consistent with this theory, adolescents exhibit heightened sensitivity to peer rejection (Sebastian et al., 2010) and peer influence (e.g., O’Brien & Bierman, 1988; Gardner & Steinberg, 2005). Some of these trends provide evidence of a period of heightened sensitivity during which certain stimuli evoke greater attention and affective reactivity, but do not alone describe the long-term learning that reflects a sensitive period.

Sensitive periods and heightened sensitivity conceptualizations of adolescence may sometimes be conflated because they are certainly complementary. Substantial literature suggests that heightened sensitivity enhances learning (Makintosh, 1975). For example, sensitive periods for language acquisition likely occur in infancy, and infants who pay more attention to speech sounds subsequently exhibit greater word production and comprehension abilities (Vouloumanos & Curtin, 2014). The spontaneous, bottom-up sensitivity to speech sounds that facilitates learning during this sensitive period stands in contrast to the overt and laborious top-down attention required for secondary language acquisition later in life (White et al., 2013). Further, stimuli with heightened salience may promote learning even when sensitive period plasticity is waning (Kuhl et al., 2003). Indeed, heightened sensitivity to relevant stimuli may be a feature of human sensitive periods in general.

Despite this, our intuitions about how heightened sensitivities map onto sensitive periods may not always prove to be straightforward. For example, heightened responsiveness to rewards has been proposed to help adolescents learn from their environments. Some prior studies have identified inverted U-shaped patterns in neural responses to reward (e.g., nucleus accumbens activation to reward outcomes) that peak in mid-adolescence and are consistent with a heightened sensitivity effect (Braams et al., 2015; Galvan et al., 2006; van Leijenhorst et al., 2010; though other studies do not find such effects, e.g., van Duijvenvoorde et al., 2014). Meanwhile, developmental patterns in reinforcement learning, including learning rates from reward feedback, tend to be characterized by steady linear improvement (Master et al., 2020; Xia et al., 2021). Studies do not identify consistent evidence that an increase in reward sensitivity during adolescence benefits reinforcement learning (Rosenbaum, Grassie, & Hartley, 2022; Nussenbaum et al., 2022; van den bos et al., 2012). Instead, heightened sensitivity to reward in adolescence has been more strongly and consistently linked with associative learning rather than reinforcement learning (Towner, Chierchia, & Blakemore, 2024). This example highlights the complexity of mapping heightened sensitivities to enhanced learning and the necessity for empirical research to refine links between the two concepts.

Additionally, overtly heightened sensitivity may not be the clearest indicator of sensitive periods. For example, the heightened emotional salience and relevance of music preferences is a common trope of adolescence. Indeed, polling across generations finds that people tend to report “the best” music as coming from their teens, with a peak at age 17 (Van Dam, 2024). Despite this, one study found that music experienced during childhood (rather than preschool or adolescence) was preferred by adults during an acute stress manipulation and uniquely supported emotion regulation; this suggests a sensitive period for the role of auditory cues in safety learning that occurs prior to adolescence (Gabard-Durnam et al., preprint). As in the previous example, understanding potential sensitive periods for learning is supported by direct investigation rather than inferences based on heightened sensitivities alone.

Defining Sociocultural Processing

We highlight two complementary uses of the term “sociocultural processing.” The first emphasizes gaining the ability to navigate sociocultural environments, where culture refers to the dynamic set of values, knowledge, rules, and concepts that facilitate interpretations of experiences and guide behavior within social and socially constructed contexts (Hudelson, 2004). Life history theories describe different strategies for optimal resource allocation across the lifespan. The prolonged duration of childhood and adolescence is thought to support extended learning opportunities, reflecting humans’ slow life history strategy (Bruner, 1972). While achieving independence and moving away from the natal environment are characteristics of adolescence that are conserved across mammalian species (Piekarski et al., 2017), humans have done so in highly variable conditions both historically and globally. Adolescence has been proposed to be a sensitive period for sociocultural processing (Blakemore & Mills, 2014) and a period for the biological embedding of culture (Worthman & Trang, 2018) that provides developmental opportunities for this tremendous range of adaptation.

A second usage of “sociocultural processing” emphasizes how interactions with others guide development. With roots in developmental psychology, this view draws on Vygotsky’s (1978) sociocultural theory, in which learning is primarily embedded in relational contexts. Taking both usages together, we define the study of sociocultural processing in adolescence within developmental cognitive neuroscience as investigations of i) the acquisition of skills and knowledge for navigating sociocultural contexts and ii) the neurodevelopmental processes by which learning and development are socially facilitated.

Where’s the “Cultural” in “Sociocultural Processing”?

A significant limitation in developmental cognitive neuroscience is the paucity of studies explicitly addressing culture. A nascent area of developmental cultural neuroscience examines how culture shapes adolescent neurodevelopment (for a review, see Qu et al., 2021), and several studies of this type are integrated throughout this review to address culture and acculturation as they might relate to sensitive periods. However, approximately 99% of samples from publications in adolescent developmental neuroscience are from Western countries, reflecting a significant challenge to claims of representativeness (Qu et al., 2021). This reflects a broader issue whereby most human psychology/neuroscience research has taken place in Western, educated, industrialized, rich and democratic (WEIRD) societies, with limited acknowledgment of this cultural bias (Henrich et al., 2010; although see Clancy & Davis, 2019 for critiques of this terminology, e.g., for failing to address race/ethnicity in the context of within-population variation). As the unfolding of sensitive periods reflects the interplay between the environment and developmental mechanisms, it is scientifically rigorous and ethically critical to consider environmental variability from wider geographic, cultural, and historical lenses (Frankenhuis & Amir, 2022; Worthman, Dockray, & Marceau, 2019). This is not solely a matter of sample representation but requires interrogating and addressing how the predominance of white and Western European perspectives in academic psychology and neuroscience has shaped the framing of research questions and limited participation from other groups in scientifically accepted knowledge-building.

Situating Sensitive Periods Along Other Neurodevelopmental Models of Adolescence

The notion that adolescence is a sensitive period for sociocultural processing is one among several prominent models of adolescent neurodevelopment and behavior. Another prominent category of models explains adolescent behaviors through interactions between developing neural systems. The dual systems model, for example, attributes notably impulsivity, risk-taking, and peer susceptibility to the earlier maturity of socioemotional systems compared to cognitive control systems (Shulman et al., 2016). Other major models in this category include triadic (Ernst, Pine, & Hardin, 2006) and imbalance models (Casey, Galván, & Somerville, 2016). Much like social reorientation models (see Sensitive periods versus periods of heightened sensitivity), models emphasizing adolescence as a period of socio-affective engagement (Crone & Dahl, 2012) highlight greater social motivation, especially concerning peers. The socio-affective perspective further suggests that adolescent neurodevelopment promotes motivation and flexible, context-dependent learning (Forbes & Dahl 2010; Crone & Dahl, 2012). Some models more specifically emphasize the role of puberty, for example clarifying how puberty exerts both direct and indirect effects on adolescent mental health via neural and psychosocial changes (Pfeifer & Allen, 2021). Additional perspectives emphasize how adolescents’ emerging neurobiology may make them more vulnerable to stress, which can lead to risk taking and substance (Spear, 2000 & Spear, 2011).

Though this is not a comprehensive review of adolescent neurodevelopmental models, the models described here, like sensitive periods theories, recognize that adolescent neurodevelopment is substantial and ongoing, and relate neurodevelopment to changes in behavior across development. While dual, triadic, and imbalance models attempt to integrate differences across developing neural systems to explain such changes in behavior, sensitive period theories are more fundamentally concerned with learning and long-term outcomes. Social reorientation models and socio-affective engagement models are perhaps more strongly aligned with sensitive periods in relating social and peer experiences to learning the skills needed for adulthood. Sensitive period theories uniquely emphasize the role of experience during developmental windows via time-limited plasticity mechanisms to drive neural reorganization.

Criteria for Evaluating Evidence

Now that we have clarified our terminology, we turn to integrating these concepts to establish criteria regarding what it means for adolescence to be a sensitive period for sociocultural processing. We are inspired by mechanistic perspectives to explore the possibility that established plasticity-regulating mechanisms are invoked in adolescence to sculpt neurodevelopment in a manner that facilitates sociocultural development. However, there are relatively few studies in the human adolescent literature that measure plasticity mechanisms themselves, even indirectly. As a result, we generally review the literature from a functional lens, where establishing adolescence as a sensitive period for sociocultural processing requires evidence of lasting impacts of adolescent experience on sociocultural development (including indices of underlying neurodevelopment) that are notable in quality and extent. Specifically, we consider evidence for lasting effects i) of adolescent experiences on the development of sociocultural skills and knowledge and ii) of social experiences occurring in adolescence on learning and development more broadly. While specific criteria are described below, we note that examining studies against these criteria is not necessarily an evaluation of overall study quality, especially when identifying sensitive periods was not a central aim of the original study. Furthermore, the present article is not a systematic review.

First, finding evidence for sensitive periods requires identifying developmentally specific effects. To address developmental specificity, we prioritize studies that use large age ranges and longitudinal follow-ups (Fuhrmann et al., 2015; Woodard & Pollak, 2020). While adolescent-specific effects are most consistent with sensitive period theories, non-linear, adolescent-emergent patterns identifying rapid change during this period may also be relevant, even if their endpoints are not clearly defined (Somerville, 2013). Tying developmental effects to age or puberty can both suggest adolescent specificity. Puberty is a protracted process of physical and hormonal changes marking the onset of adolescence and leading to social changes and reproductive maturity. Further, puberty is proposed to regulate the opening and closing of different sensitive periods (Piekarski et al., 2017). When possible, we prioritize longitudinal studies because cross-sectional studies typically cannot evaluate within-person change (Kraemer et al., 2000); indeed, longitudinal and cross-sectional studies can reveal different developmental patterns (Pfefferbaum & Sullivan, 2015). To facilitate interpretation, we refer to ages 10-13 as early, 14-17 as mid, and 18-21 as late adolescence throughout and include specific ages when studies do not fit into these approximate age groupings.

Second, to address specificity to the social domain, we note when studies make explicit comparisons to non-social conditions or control variables and the degree to which studies rule out alternatives when making claims about uniquely social effects. However, we also consider that social and cognitive development are intertwined, such that development in non-social processes does not necessarily challenge sensitive period conceptualizations of adolescence.

Third, we distinguish between studies that identify evidence for enduring effects of experience on development consistent with sensitive periods versus those that focus on acute responses that provide evidence for periods of heightened sensitivity.

Review of the Literature

Structural Brain Development

Studies of human brain structure suggest that adolescence is a period of significant neural remodeling. In this section, we highlight how structural neuroimaging studies have identified changes in whole-brain volume, cortical gray/white matter volume, subcortical gray matter volume, and indirect measures of myelination during adolescence. The overall spatiotemporal pattern of structural development suggests ongoing development in many regions, including but not limited to those supporting social cognitive and socio-affective processing, and an emerging area relates neurodevelopmental changes to pubertal development (Pfeifer & Allen, 2021). While studies of brain volume and thickness do not identify underlying cellular mechanisms, emerging metrics in human neuroscience implicate mechanisms of sensitive period plasticity that converge with animal research (Sydnor & Satterthwaite, 2023). Taken together, these findings highlight the biological plausibility of sensitive periods as partly driving neural remodeling observed in adolescence. However, the degree to which experiences in adolescence and the contributions of puberty to neurodevelopmental change are not yet clear.

Trends in Global and “Social Brain” Cortical Development

Studies characterizing trajectories of macroscopic brain properties typically not directly test sensitive period theories of specific sociocultural processes, but rather help us understand if they are neurobiologically plausible. Recent investigations of structural brain development have used team and open science approaches to replicate and challenge critical findings that the brain undergoes significant structural remodeling during adolescence. One such effort involving replication across four longitudinal samples identified the following major trends in average structural brain development: (1) whole-brain volume increases until ages 10-15 and then decreases into the early 20s, (2) cortical gray matter volume tends to be highest in childhood and decreases across adolescence, and (3) cortical white matter volume increases until mid-to-late adolescence (Mills et al., 2016; see also Aubert-Broche et al., 2013; Lebel & Beaulieu, 2011; Wierenga et al. 2014). A major effort to develop brain growth charts across the lifespan identified similar trajectories of average whole-brain and gray matter volume (with peaks around age 12 and 6, respectively); however, white matter volume increased until later than previously identified (into the late 20s) (Bethlehem et al., 2022). Additionally, after reaching relative stability in childhood, cortical thickness begins to decline in late childhood/early adolescence (ibid). The average midpoint of this phase of cortical thinning is at age 14; this value may help index individual differences regarding the pace of development (Fuhrmann et al., 2022).

Do brain regions implicated in sociocultural processing undergo especially pronounced structural changes during adolescence? Social processes are diverse and recruit an extensive set of regions, sometimes called the “social brain” (Alcalá-López, 2018). Using longitudinal data from middle childhood through adulthood, Mills and colleagues (2014) identified changing trajectories of gray matter volume, cortical thickness, and surface area within a set of regions strongly implicated in mentalizing and social cognition, including the medial prefrontal cortex (mPFC; defined as medial Brodmann Area 10), temporoparietal junction (TPJ), posterior superior temporal sulcus, and the anterior temporal cortex. Recent analyses replicated gray matter volume and cortical thickness decreases in three of these four regions (the anterior temporal cortex was not examined) (Becht et al., 2020). Direct comparisons to other brain regions were not made. However, other studies indeed identify structural changes of similar magnitude in areas of the cortex beyond the “social brain” (e.g., Tamnes et al., 2017). Thus, while notable structural changes within social brain regions enhance the plausibility of adolescence as a sensitive period for sociocultural processing, evidence from structural neuroimaging does not support specificity to the social domain.

Are Changes in Structural Brain Development Tied to Developmental Plasticity?

Structural brain metrics derived from MRI do not typically directly measure specific cellular processes and likely reflect multiple underlying neurobiological processes (Paus, 2008). For example, known declines in gray matter across adolescence have been proposed to reflect a combination of synaptic pruning, a reduction of neuropil and glial cells, and the encroachment of white matter and myelination (Mills & Tamnes, 2014). However, some researchers speculate that neurodevelopmental features associated with earlier development (e.g., thicker cortex) may indicate greater remaining plasticity (Becht et al., 2020).

Non-invasive human neuroimaging measures have advanced substantially within the last few years, including newer metrics that may be more specifically related to critical period onset, maintenance, and closing (Sydnor & Satterthwaite, 2023). For example, critical periods are triggered by the maturation of GABA-associated neurotransmission and subsequent increases in inhibitory relative to excitatory activity; this shift in what is termed “E-I balance” supports the fine-tuning of neurodevelopment by suppressing spontaneous background firing and improving sensitivity to stimulus-evoked activity (Takesian & Hensch, 2013; Toyoizumi et al., 2013). Recently, the maturation of GABA-dependent inhibitory circuitry has been characterized using pharmacological fMRI in tandem with standard fMRI and machine learning approaches (Larsen et al., 2022). This study of participants ages 8-22 identified age-related increases in inhibitory relative to excitatory in association cortex, but not in unimodal or sensorimotor regions. These findings are consistent with predictions of critical period emergence during and into late adolescence, but do not support specificity to the social domain.

Meanwhile, intracortical myelination serves to consolidate neural circuitry and may signal the closure of sensitive periods (Sydnor & Satterthwaite, 2023); this process can be measured indirectly via the T1-weighted to T2-weighted ratio (Glasser & Van Essen, 2011) in developmental samples (Glasser et al., 2022). One study found that across ages 8-21, regions of association cortex exhibited more gradual and linear age-related increases in myelination compared to those of sensorimotor cortex (Baum et al., 2022). Findings from magnetization-transfer imaging similarly identify overall increases in intracortical myelination across adolescence (Whitaker et al., 2016) occurring in specific layers of cortex and primarily in association cortex (Paquola et al., 2019). As social brain regions overlap with association cortex, this pattern of spatiotemporal findings (much like those from prior structural studies) is compatible with sensitive periods for sociocultural processing but does not suggest specificity to the social domain. Additionally, they suggest a relatively late closure of sensitive periods in the association cortex. Therefore, sensitive periods for sociocultural processing, if they exist, may be one aspect of broader, domain-general changes and quite protracted, extending beyond late adolescence.

By targeting processes with established associations to neuroplasticity, indirect measures of E:I balance and intracortical myelination provide novel opportunities to advance our understanding of critical and sensitive periods. However, these measures still require additional validation and, like other non-invasive measures before them, are limited by the indirect and non-microscopic nature of the measures.

To What Degree are Changes in Structural Brain Development Driven by Experience?

This section highlights that neurodevelopment is clearly ongoing during adolescence, and that sensitive period mechanisms may be at play. While it is highly likely that neural changes are sculpted by ongoing experience to some degree, other factors may also drive observed neural changes. Structural brain metrics are significantly influenced by genetics (Winkler et al., 2010), and converging evidence from twin studies (Brouwer et al., 2021; Valk et al., 2021) and post-mortem histology (Whitaker et al., 2016) identifies a substantial genetic contribution to neurodevelopment in adolescence specifically. In addition, there is ample evidence that cascading effects of early life experiences impact adolescent neurodevelopment (e.g., Hodel et al., 2015). A few studies have attributed regional patterns of gray matter development to experiences occurring later in childhood (specifically, in the TPJ in development from 7-9 years of age; van der Meulen et al., 2020) and in early adolescence (Brouwer et al., 2015). Association regions that exhibit protracted development extending into and beyond adolescence have lower heritability, suggesting greater influence of experience during adolescence (Winkler et al., 2010).

What are the Contributions of Puberty to Structural Brain Development?

Puberty involves dramatic increases in hormone levels across the adrenal, gonadal, and growth axes, and pubertal hormones may organize neurodevelopment. Like age, pubertal stage and testosterone levels tend to be associated with reduced gray matter volume in diverse regions across the cortex, particularly across the frontal and temporal lobes (this may partly be because the prefrontal cortex is commonly the focus of studies; see Vijayakumar et al., 2018 for a systematic review). Beyond the cortex, changes in gray matter volume for some subcortical structures are accelerated in early or mid-puberty (Goddings et al., 2014), and such changes may be better explained by pubertal status as compared to age (Wierenga et al., 2018). Emerging evidence also suggests that puberty and pubertal hormones are associated with widespread sexually dimorphic changes in cortical thickness and surface area (Nguyen et al., 2013, Herting et al., 2015; Vijayakumar et al., 2021; Byrne et al., 2023), white matter microstructure (Barendse et al. 2018; Chahal et al., 2018; Herting et al., 2017;), and changes in the subcortical volume of the amygdala (Goddings et al., 2014; Herting et al., 2014; Vijayakumar et al., 2018). Challenges in disentangling the effects of age and puberty, as well as a limited number of longitudinal studies, likely contribute to inconsistencies in this literature (Herting & Sowell, 2017; Vijayakumar et al., 2018).

Social Cognition

Social cognition encompasses a vast number of subprocesses and subfields. This review selectively examines face processing, a basic and widely studied process in cognitive neuroscience with recent developments relevant to the study of sensitive periods. Overall, face-processing studies suggest the possibility of a puberty-related transition toward understanding complex socioemotional expressions in peers. Meanwhile, adolescence may also be a period in which certain types of plasticity are lost in relation to face processing. Taken together, this shift in plasticity for some types of information about faces highlights subtleties in considering adolescence as a potential sensitive period for sociocultural processing.

Opening and Closing Sensitive Periods for Face Processing in Adolescence

Face perception exhibits protracted development into adolescence (Fuhrmann et al., 2016; Rodger et al., 2015; Thomas et al., 2007), and some advancements may be tied to pubertal development. Puberty is a candidate for explaining individual differences in sensitive periods timing, as pubertal timing varies by several years in humans (Tanner & Whitehouse, 1976) and is influenced by environmental factors (Mendle et al., 2015). Puberty is associated with a dip in the recognition of adult faces (Scherf et al., 2012) and improved recognition of peer faces (Picci & Scherf, 2016). In addition to shifts in who is recognized, adolescence may be a period in which there are changes in what is recognized. For example, one study identified enhanced discrimination of facial expressions conveying complex social-emotional states (sexual interest and contempt) but not basic emotional states (happiness, anger) with pubertal development (Motta-Mena & Scherf, 2017). Studies have commonly identified changes in amygdala activation across puberty when processing emotional face stimuli, but the direction of these associations is unclear across studies (Dai & Scherf, 2019). The types of puberty measures and face stimuli vary across the neuroimaging literature, including adult and adolescent faces.

The transition into adolescence may also be associated with the loss of plasticity in earlier-developing aspects of face perception (Pascalis, 2020). The classic “other-race” effect finds that individuals with more experience with faces from their own race tend to better recognize those faces than those from other races (Kelly et al., 2005). While this effect emerges early in infancy, one study suggests that it can be reversed with extended experience up until approximately 12 years of age (McKone et al., 2019), indicating the transition into adolescence as an endpoint rather than a starting point for this type of perceptual narrowing.

Peer Processes

Adolescence involves navigating and making sense of increasingly complex peer interactions. Adolescents respond uniquely to peers in various ways, including heightened autonomic arousal in the presence of peers, reduced cognitive control in affective decision-making contexts, and heightened sensitivity to peer rejection. A few studies of peer-related effects with wide age spans suggest developmentally specific effects, although studies have mainly been cross-sectional. In this section, we highlight theories that changes in pubertal hormones may increase peer-related social motivations to inform decision-making during adolescence. Generally, this work has identified changes in sensitivity to peers and has focused on predicting behavior in the short term, especially risk-taking. However, some studies suggest that aspects of adolescent peer experiences impact socioemotional development and mental health in adulthood.

Time Spent with Peers in Person and Online

In the social reorientation model, (Nelson et al., 2005; Nelson et al., 2016), a widely cited piece of evidence for adolescents’ greater orientation toward peers is that adolescents spend more time with peers and less time with their families than they did as children. However, empirical investigations paint a more complex picture. Early and highly cited studies were small and sampled from white, suburban, working-class, and middle-class American youth (Csikszentmihalyi & Larson, 1984, Larson & Richards, 1991; Larson, Richards, Moneta, Holmbeck, & Duckett, 1996). One study with this population found that between 9-15 years of age, adolescents’ time with their parents decreased by half. Yet, this additional time was accounted for by increases in time spent with peers for female participants only (Larson & Richards, 1991), and there was no change in time spent talking to family members overall (Larson et al., 1996). Critically, research with other populations finds substantial variability in how children and adolescents spend their time. A study of urban African American 5^th-8^th graders reported no change in the amount of time spent with family across early and into mid-adolescence and reported no increase in time spent alone or with friends (Larson, Richards, Sims, & Dworkin, 2001). Further, studies of adolescents between the approximate ages of 10-17 in East Asian, South Asian, and European countries reveal that patterns in time with family and peers vary substantially around the globe (Larson & Verma, 1999). Overall, time spent with peers varies by culture and may not provide strong evidence for universal theories of adolescent development.

The last few decades have seen a rise in concerns about how much of adolescents’ time is spent in digital spaces. Despite this, a sample largely representative of 12-15-year-old adolescents in North Carolina found that daily digital technology use was not consistently associated with the amount or quality of interactions with parents (Jensen et al., 2020). This finding complicates the notion that the digital spaces are necessary sites of displacement for time spent with the family by time spent with peers. The effects of social media use on well-being and mental health are currently of major public interest (for a recent synthesis of reviews and meta-analyses in this field, see Valkenburg, 2021). Using a large sample spanning early to late adolescence, greater estimated social media use was found to be associated with longitudinal decreases in life satisfaction at specific windows of development (ages 14-15 and 19 in males; 11-13 and 19 in females; Orben et al., 2022). These findings point to a potential sensitive period of vulnerability to social media that the authors speculate is driven by puberty, as females undergo pubertal development earlier than males. As longitudinal decreases were estimated for the following year only, an understanding of whether such effects are longer-lasting would help us understand if this is a sensitive period versus a more transient vulnerability. Additionally, person-specific longitudinal approaches suggest that sizable subgroups of adolescents exhibit positive effects of social media use on self-esteem and well-being (Beyens et al., 2021; Valkenburg et al., 2021). Thus, sensitive periods research in relation to social media use might also incorporate the potential for positive impacts of social media in some adolescents.

Long-Term Impacts of Adolescent Social Connection and Disconnection

Social connectedness during adolescence may have a lasting influence on development, with close friendship quality during early and mid-adolescence being related to physical (Allen et al., 2015) and mental (Narr et al., 2019) health over a decade later. These findings suggest that adolescent social experiences have long-term effects, which is very compatible with the theory that adolescence is a sensitive period for sociocultural processing. However, inferences regarding developmental specificity are limited, as the youngest participants in these studies were adolescents, not children.

Compared to adults, early- and mid-adolescents experience stronger affective responses to negative social evaluation (Sebastian et al., 2010). Loneliness, or perceptions of social isolation, tend to increase between childhood and late adolescence/young adulthood (Qualter et al., 2015) and may mediate the relationship between social connectedness and depressive symptoms seen by early adolescence (Witvliet et al., 2010). One three-year longitudinal study found that positive social support at age 14 buffered against future depressive symptoms at age 17 for adolescents with a history of childhood adversity (van Harmelen et al., 2016). However, it is unclear whether supportive friendships lead to more resilient functioning (van Harmelen et al., 2017), or if resilient functioning and supportive friendships change together over time (van Harmelen et al., 2020). Further, developmental studies investigating the neural correlates of social exclusion and its impacts on subsequent behaviors tend to group children and adolescents together (Vijayakumar et al., 2017; Bolling, Pelphrey, & Vander Wyk, 2016) or use adolescent-only samples (Cheng et al., 2020; Peake et al., 2013; Falk et al., 2014; Wasylyshyn et al., 2018). Overall, adolescent social experiences of connection and disconnection may be relevant to a range of long-term behavioral and mental health outcomes in a manner that is consistent with sensitive period effects. However, evidence regarding the developmental specificity and dynamic unfolding of such effects is generally limited.

Peers Influence Decisions, Attitudes, and Behaviors

The mere presence of peers has been found to elicit unique effects in adolescents. Being ostensibly watched by a peer induces heightened autonomic arousal and embarrassment, an effect that was found to peak around mid-adolescence in a cross-sectional sample of children to adults (Somerville et al., 2013). The presence of peers or social stimuli can diminish cognitive control and forms of reasoning in adolescents (King et al., 2018; Breiner et al., 2018; Wolf et al., 2015). However, peer presence tends not to impact (Smith et al., 2018) or can even improve (King et al., 2018) “cold” cognitive control involving more neutral stimuli and outcomes (e.g., Barker et al., 2018; Buzzell et al., 2019). Numerous studies additionally characterize how adolescents adopt behaviors and attitudes modeled by their peers, finding that adolescents conform to peers’ perceptions of risk (Knoll et al., 2015; Knoll et al., 2017), and that peers can also drive adolescents toward prosocial or lower-risk options (e.g., Do et al., 2020; Braams et al., 2019). Further, adolescents’ propensity to engage in health-risking behaviors is partly explained by the degree to which the behaviors are aligned with their social motivations (Blakemore & Mills, 2014; Ellis et al., 2012; Pfeifer & Berkman, 2018; Andrews et al., 2020). While such peer effects are intriguing and compatible with heightened sensitivity and sensitive period theories, most studies are limited in their ability to truly confirm or falsify such theories due to the predominance of adolescent-only samples (lack of developmental specificity) and limited longitudinal follow-up (uncertainty regarding lasting effects). Below, we highlight studies of particular relevance despite some of these limitations:

In one longitudinal sample of mid-adolescents through young adults, participants’ tendency to adopt their peers’ preferences predicted the quality of their peer relationships 1.5 years later (Reiter et al., 2021). This study notably established a link between a form of sensitivity (susceptibility to peer influence) and later social functioning (quality of peer relationships). The effect was driven by younger adolescents, which is consistent with the notion that early adolescence is a period for peer susceptibility, and that heightened sensitivity to peers may support social learning. This study could not identify an adolescent-specific effect, as the youngest participants were adolescents and not children.

Hormones are another avenue for understanding heightened sensitivities, as shifts in adolescents’ social motivations may be modulated by hormonal changes (Pfeifer & Allen, 2021). Testosterone is related to the motivation to learn about and seek social status (Eisenegger et al., 2011) and is associated with greater conformity to high-status behaviors across social contexts (Rowe et al., 2004). According to the dual hormone hypothesis (Mehta & Josephs, 2010), testosterone is more strongly associated with status-seeking behaviors when cortisol levels are low, but this research has primarily been conducted in adults. Work with early adolescents has identified that those with high levels of testosterone and low levels of cortisol tended to exhibit the greatest conformity to prosocial behaviors espoused by peers (Duell et al., 2021). This study also found that, when making prosocial decisions after viewing peer behavior, individuals high in testosterone and low in cortisol exhibited greater signal in brain regions associated with prosocial decision-making (including posterior superior sulcus/TPJ, orbitofrontal cortex, insula, and caudate). While the limited age range in this study alone prevents inferences regarding heightened sensitivity and sensitive period theories, this study highlights how the relative dynamics of increasing hormone levels across puberty could explain the emergence of heightened sensitivities to peers.

How Specific are Peer and Status Effects in Adolescence?

While the studies reviewed in this section provide confirmatory evidence about the importance of peer processes during adolescence, they do not address critical, testable hypotheses emerging from social reorientation theory: that information about social status and peers becomes uniquely salient during adolescence as compared to processing in other domains and from other sources. One study sought to explicitly evaluate this claim by examining within-person neurodevelopmental trajectories of self- and other-evaluation on traits relevant to social status versus academic competence (Cosme et al., 2022). In a longitudinal sample spanning ages 9-17, status-related information did not elicit unique increases in neural signal across regions associated with social processing.

Relatedly, we note that adolescents do not overwhelmingly or exclusively favor peers across contexts. Adolescents do not indiscriminately focus on peers rather than adults when engaging in strategic decision-making (McCormick et al., 2018) and in laboratory tasks they often make decisions that benefit their parents at the expense of friends across multiple domains (Guassi Moreira et al., 2020). Compared to adults, adolescents more strongly modulate their prosocial behaviors in response to social influence, regardless of whether that influence came from an adult or a peer (but social influences were stronger than non-social/computer influences; Chierchia et al., 2020). This lack of peer- and status-specificity challenges social reorientation theories. Further, if adolescents’ heightened social sensitivities support sensitive periods, they may influence diverse aspects of social functioning inclusive of but also beyond peer and mating-related outcomes.

Calibration to Social Environments in Adolescence

Another source of evidence for sensitive periods might come from both mathematical modeling approaches and empirical studies that examine psychobiological calibration to environmental conditions experienced during adolescence. Emerging empirical studies characterize calibration across several psychobiological systems, implicating the adolescent caregiving environment and broader sociocultural context. Some studies in this area use clear and specific developmental comparisons using age or other indices of maturation. However, many studies use broad indices of environmental influence (e.g., time spent in another country), and additional investigations might clarify which aspects of the environment drive developmental change.

Mathematical Modeling Approaches

Mathematical modeling approaches can also contribute to characterizing the evolution of sensitive periods. Developmental plasticity itself is an outcome of evolutionary pressures, and maintaining plasticity across stages of development may facilitate survival when the environment is variable within an individual’s lifespan and across members of a species (Frankenhuis & Fraley, 2017). Using formal models that simulate empirical study paradigms, one study found that sensitive periods may be likely to evolve in “mid-ontogeny” (i.e., in middle childhood or adolescence, as opposed to very early in life) when information about the environment becomes more reliable over time (Walasek, Frankenhuis, & Panchanathan, 2021). Though these models are species and domain general, the authors suggest that the greater reliability of social cues indicative of social status and mate value in mid-ontogeny raise the possibility that becoming highly attuned to peer social information during adolescence may be evolutionarily adaptive. On the other hand, formal models find that sensitive periods evolve early in life when information is generally reliable from the outset, as is the case for basic sensory and language processes.

Empirical Evidence for Calibration of Psychobiological Systems

The pubertal stress recalibration hypothesis is one line of research examining how psychobiological systems calibrate to the environment. This hypothesis suggests that the hypothalamic-pituitary-adrenal (HPA) axis adapts to the environment during pubertal development (DePasquale et al., 2018). Evidence for this hypothesis comes from studies of children who were adopted following early-life institutionalization; specifically, as they progress through puberty their previously blunted cortisol stress reactivity is restored to patterns more typically seen in non-institutionalized youth (DePasquale et al., 2018; Gunnar et al., 2019). A study employing a similar design found that as previously institutionalized children undergo adrenal development, they recover post-stressor hormone-coupling patterns (positive cortisol-DHEA coupling) more comparable to those of non-adopted youth (Howland et al., 2020). Studies of never-institutionalized youth also have also examined the HPA axis in relation to the sociocultural environment. For example, parent-child synchrony in cortisol levels during everyday life (Papp et al., 2009) and during discussions of conflict (Saxbe et al., 2014) extends into mid-adolescence, with higher synchrony in some cases among dyads who spend more time together (Papp et al., 2009). Another study of bicultural Mexican American 14-year-olds found that cultural orientation to either Mexican and/or Anglo cultures shapes patterns of cortisol responses to acute stressors during adolescence (Gonzales et al., 2018). Taken together, such research suggests that sociocultural environments may sculpt the development of the HPA axis during adolescence, though studies establishing developmental specificity to puberty have primarily been conducted with previously institutionalized adolescents.

Caregiving and cultural environments may sculpt neurocognitive development beyond the HPA axis and adolescent stress responses. In one study of post-institutionalized youth, caregiving quality impacted reward processing, executive functioning, and psychopathology during adolescence (Colich et al., 2020). Critically, the latter two effects of caregiving quality were stronger when assessed during adolescence (ages 12 and 16) than at age 8, suggesting an adolescent-specific effect. Additionally, acculturation following immigration might provide an important model for studying plasticity (Qu et al., 2021), though few studies have employed a sensitive periods framework. Among immigrants from Hong Kong to Canada, an early study suggested that childhood may be a sensitive period for acculturation by identifying faster rates of acculturation (self-reported identification with mainstream culture) among those who immigrated at younger ages (Cheung et al., 2010). However, this finding was not replicated in a more diverse sample of immigrants to the United States by the same research group (Chudek et al., 2015). Furthermore, there are few neuroscience studies of acculturation. Among those that exist, findings identify, for example, that genetic sensitivity to cultural influences may interact with time spent in the United States to influence brain regions associated with social cognitive functioning and decision-making in young adults (Kitayama et al., 2020; Yu et al., 2019). However, without greater attention to developmental timing, such gene by environment interaction studies cannot specifically contribute to the question of adolescence as a sensitive period for sociocultural processing. Overall, larger studies with more detailed information about the timing of immigration are needed to assess the developmental specificity of acculturation effects with respect to sensitive periods.

Discussion

This review provided major contemporary updates to the question of adolescence as a sensitive period for sociocultural processing originally proposed by Blakemore and Mills in 2014. In this discussion, we first summarize the major theoretical and empirical contributions of this review. Next, we synthesize evidence from the literature with respect to the criteria for specificity proposed in the introduction. Finally, we consider translational implications of sensitive periods for sociocultural processing in adolescence.

Overview of Theoretical and Empirical Updates

We began by refining the theoretical background and concepts informing the question of adolescence as a sensitive period for sociocultural processing. We distinguished between mechanistic versus functional definitions of sensitive periods and heightened sensitivity versus sensitive periods, and especially highlighted how research bridging heightened sensitivities to sensitive period outcomes might strengthen our understanding of adolescent sensitive periods. Next, sociocultural processing was defined for the first time in the context of sensitive periods theories of adolescence. We noted that attention to sociocultural variability in psychology and neuroscience is both emerging and still urgently needed.

Next, we reviewed the literature pertaining to adolescent sensitive periods for sociocultural processing, attending to both developmental specificity (prioritizing findings from longitudinal research and studies that compared large age ranges) and specificity to sociocultural processing as compared to other domains. We identified evidence from four key areas, including 1) structural brain development, 2) the social cognitive development of face processing, 3) peer processes and influences, and 4) studies identifying psychobiological calibration to environments in adolescence. (For a summary, see Table 1.) Overall, while numerous studies identified findings compatible with the theory of sensitive periods for sociocultural processing in adolescence, strong and direct empirical tests of the theory in relation to specific sociocultural processes were less common. Although there may be multiple overlapping sensitive periods for sociocultural processing in adolescence (Woodard & Pollak, 2020), well-developed accounts of sociocultural processes that met stricter criteria for sensitive periods were fairly limited.

Table 1.

Summary of evidence discussed in relation to potential sensitive periods for sociocultural development

	Summary of key research findings	Relevant to HS, SP, or Both*	Developmental specificity to adolescence	Specificity to sociocultural processing
Structural brain development	Studies replicated across samples and with large age spans identify robust and unique patterns of structural brain development in adolescence in widely used metrics of volume, thickness, surface area, etc. An emerging area identifies changes with puberty. Newer metrics indirectly index mechanisms tied to the opening and closing of SPs; these suggest that SP mechanisms operate in adolescence.	Could identify biological plausibility for either depending on the region, metric, and trajectory. Newer metrics are more closely tied to SPs, but are still indirect forms of measurement	Major changes occur in adolescence. However, especially for the association cortex, development can progress beyond adolescence, too.	No. Though substantial changes are identified in social brain regions, changes in other regions are also evident.
Social cognition	Adolescence may be a period of improved recognition of peer over adult faces. Puberty is tied to greater understanding of facial expressions conveying complex socioaffective states over basic emotional states. This may be accompanied by a loss of plasticity in earlier-developing aspects of face perception.	Recognition effects may suggest HS, while learning to understand facial expressions is suggestive of a SP.	Yes, and effects tied to pubertal stage are notable.	Unclear. For higher-level aspects of face processing that advance in adolescence, clear non-social comparison stimuli may not be possible. However, the ability to understand complexity in general may also improve in adolescence.
Peer processes	Time spent with peers varies by culture and is not good evidence of SPs. There are potential sex-specific SPs for vulnerability to social media effects on well-being. Social connectedness and friendship in adolescence predicts physical and mental health in adulthood. Peers can elicit heightened self-consciousness and can diminish cognitive control and reasoning, especially in affectively laden contexts. Peers influence adolescents’ perceptions, behaviors, and attitudes for risky, prosocial, and other types of information.	SP SP Largely HS; emerging studies notably link peer influence to social functioning outcomes which may bridge HS and SP theories	Partly; the social media study identified SP windows within adolescence but did not examine children <10 years old. Partly; specific to adolescents vs. adults, but in some studies the youngest participants were in mid-adolescence. Somewhat; some studies with larger age ranges. Many adolescent-only or adolescent-to-adult samples find stronger effects in younger participants with the lower age bound yet to be identified.	Yes, to an extent. Many studies show effects of peers, especially when compared to non-social control conditions. However, some effects are not specific to peers or status information (compared to adults/other types of social information).
Calibration to social environments	Mathematical models suggest that middle childhood and adolescent sensitive periods can be favored in evolution under certain trajectories of environmental cue reliability. Studies of post-institutional caregiving indicate the importance of caregiving environments to sculpting the stress response and cognitive/ psychosocial development. Evidence is mixed as to whether age of immigration impacts rate of acculturation.	SP SP SP	Models suggest the plausibility of SPs beyond infancy, but are species-general and do not point to specific ages. Yes, and effects tied to pubertal stage are notable. No, adolescent-specific approaches were not used; some evidence that acculturation may be fastest in children rather than adolescents	No, such models are species and domain general. Likely yes. Many critical aspects of caregiving quality are social. Likely yes. Acculturation is a sociocultural phenomenon.

Note. HS: Heightened sensitivity, SP: Sensitive period;

^*This column indicates whether the form of evidence is theoretically relevant to heightened sensitivity effects, sensitive period effects, or both, but does not indicate whether the evidence itself is strong and/or specific.

Specificity of the Evidence and Considerations for Future Research

In this section, we revisit our initial criteria for establishing sensitive periods for sociocultural processing in order 1) to synthesize findings from the literature and 2) to identify progress, gaps, and opportunities for future research. We primarily sought to identify evidence pertaining to lasting impacts of adolescent experience on sociocultural development (and underlying neurodevelopment) notable in quality and extent. Below, we discuss types of specificity relevant to this definition that were present to varying degrees across the literature. We further highlight approaches (both experimental and observational) that may provide especially strong tests to guide future research. (For a detailed discussion of study designs that may enhance our understanding of sensitive periods, see Gabard-Durnam & McLaughlin, 2020).

Specificity to long-term effects.

Few studies explicitly tested whether effects were long-lasting, i.e., affected outcomes beyond adolescence. Of those that did, studies relating feelings of social connectedness in adolescence to physical and mental health in adulthood were among the most compelling (Allen at al., 2015; Narr et al., 2019). (However, the youngest participants in these studies were adolescents, making it difficult to determine if effects were truly specific to adolescence.) More commonly, studies examined outcomes a year or two after initial assessments, when study participants were still adolescents. For some aspects of social learning and skill development, it may be reasonable to assume that effects are long-lasting without requiring extensive longitudinal follow-up. For example, one study focused on identification of complex socioemotional facial expressions (Motta-Mena & Scherf, 2017), an ability that, once acquired, does not likely diminish in young adulthood. For other types of outcomes, researchers might gather data further into the future and/or incorporate information about the stability of effects estimated in the literature when framing arguments regarding sensitive periods. When conducting longitudinal follow up, researchers might also collect information about whether participants participated in effortful and intentional learning pertaining to long-term outcomes of interest, in order to better distinguish sensitive periods from adult plasticity.

Structured life course modeling is an approach that is especially well-suited for investigating potential sensitive periods over long timeframes when longitudinal observational datasets with many repeated observations are available (Smith et al., 2022). This method can be used to evaluate which of several complex life course hypotheses (sensitive periods at different developmental phases, but also recency, early life effects, and/or the accumulation of experience) for some repeated exposure best explains variation in an outcome at some point in the future, but may not be able to address confounding between timing and duration (Gabard-Durnam & McLaughlin, 2019).

Specificity to adolescent experiences as a driver of change.

Another limitation of the literature is that few studies isolated effects of adolescent experience relative to other sources of outcome variability, such as genetics and/or earlier experience. Relationships between genes, experience, and behavior over development are complex, as sensitive periods reflect the unfolding of genetic programs for acquiring and embedding experience (e.g., Kitayama et al., 2020; Yu et al., 2019) and may further be shaped by epigenetics (Dunn et al., 2019). Research that aims to estimate the contributions of adolescent experiences in light of these other sources of variability may strengthen our understanding of sensitive periods in adolescence.

Additionally, intervention approaches may be especially well-suited for identifying the importance of experience at different developmental time periods. For example, one set of studies we reviewed examined high-quality caregiving as an intervention for institutionalized youth, identifying that caregiving during puberty was important for a range of outcomes including stress physiology (DePasquale et al., 2018; Gunnar et al., 2019) and cognitive/psychosocial development more broadly (Colich et al., 2020). More targeted, short-term interventions may also shed insights into sensitive period timing. For example, one preregistered study of intensive reading intervention for dyslexia found no differences between younger and older participants’ behavioral and neural responses to the intervention (across ages 7-13; Yeatman & Huber, preprint). Similar amenability to intervention across all ages suggested that plasticity in the targeted domain might be ongoing beyond middle school. Though this example comes from the cognitive domain, it highlights a study design and analysis strategy that might be useful across domains.

Developmental specificity.

Most studies reviewed were cross-sectional or used adolescent-only samples, and it is difficult to establish evidence for sensitive periods hypotheses without comparisons across more comprehensive age ranges (Blakemore & Mills, 2014; Fuhrmann, Knoll, Blakemore, 2015; Pfeifer & Allen, 2015). Numerous studies that we reviewed reported stronger effects in younger adolescents compared to older adolescents or adults, but could not rule out other possibilities (e.g., effects in childhood or effects of earlier life experiences). Addressing developmental specificity can be resource- and time-intensive, often requiring samples with large age ranges and/or a significant amount of time to observe development longitudinally. Because puberty is a developmentally specific process in adolescence, targeted study designs with small to medium-sized samples that vary in puberty can also be informative when puberty is the driving force behind developmental change. For example, face perception studies of adolescents aged 11-14 who varied in their pubertal timing strongly implicate puberty in perceptual specialization for facial expressions conveying complex social emotions (Picci & Scherf, 2016; Motta-Mena & Scherf, 2017). This study design can also overcome common challenges in disentangling the effects of puberty and age.

Specificity to sociocultural processing.

A substantial body of literature underscores the significance of social experiences during adolescence, including interactions with peers and caregivers (see Peers Influence Decisions, Attitudes, and Behaviors). However, patterns in structural brain development do not indicate that sensitive periods are likely exclusive to social brain regions. Instead, regions supporting a variety of complex cognitive abilities mature during human adolescence, which may facilitate the navigation of both social and nonsocial challenges. Indeed, some posit that adolescence may also be a sensitive period for cognitive development (Laube, van den Bos, & Fandakova, 2020). Furthermore, while peer and status-related information is notably important to adolescents, unique effects of peers and status are not consistently observed when tested against other robust social control conditions (e.g., interactions with adults and family members or other social information) rather than non-social conditions (e.g., Chierchia et al., 2020; Cosme et al., 2022). While multiple sensitive periods can coexist in adolescence, using strong comparison conditions in future research can enhance our understanding of the specificity and limits of sensitive period effects.

Addressing sensitive periods versus periods of heightened sensitivity.

Whereas studies of sensitive periods in sensory and language domains typically rely on concrete learning benchmarks, such benchmarks are often less sharply defined in the socio-affective domain. A notable number of studies, especially those encountered in the peer domain, especially focused on heightened sensitivity to peer faces, behaviors, attitudes, etc., but did not employ the types of study designs or analyses that might speak to sensitive periods. Overall, we find a greater focus on characterizing heightened sensitivity to the stimuli themselves, rather than what is learned from them or how individuals learn to decipher them. We highlight one article that especially linked heightened peer susceptibility to the quality of peer relationships at a later time point (Reiter et al., 2021). This work suggests that some aspects of heightened peer susceptibility support social functioning. Future work might build further bridges to empirically characterize how periods of heightened sensitivity facilitate enhanced learning.

A Translational Science Lens

A major goal of developmental science is to foster healthy youth development by creating valuable and actionable insights. Much is at stake, as experiences during adolescence build upon those in the first decade of life and present opportunities to cultivate positive development or further exacerbate global inequalities (Dahl et al., 2018; Worthman, Dockray, & Marceau, 2019). Sensitive period theories can support translational efforts of developmental science as a crucial framework for understanding how experiences become neurobiologically embedded during development.

Conceptualizing adolescence as a period when social experiences support long-term psychosocial development, even in a broad, heuristic sense, can help address legal, ethical, and policy questions. For example, in evaluating the practices of solitary confinement or social isolation used as punishment in the justice and educational systems, policymakers and educators might consider the detrimental long-term effects of significant isolation on adolescents: social isolation has been noted for inducing symptoms of psychosis as early as the 19th century (Smith, 2006), and isolation in juvenile detention is associated with higher rates of suicide (Cloud et al., 2014). Weighing developmental needs for social connection along with public safety is also relevant to policy during public health crises, such as lockdowns and school closures for covid-19.

Precise and careful empirical studies on sensitive periods can also inform translational efforts targeting specific cognitive and health outcomes. Educators, medical professionals, and mental health practitioners often receive guidelines based on age groups, but these do not account for variability in pubertal onset, which can differ by many years. Considering how adolescents’ progression through puberty relates to sensitive periods in development could inform more targeted efforts to support long-term psychosocial and cognitive outcomes. Specifically, variability in the onset and course of puberty in relation to adolescent sensitive period learning likely relates to psychosocial risk and resilience. As another example, knowledge of sensitive periods may support an understanding of how and when to harness or upregulate plasticity (via pharmacological or behavioral means) to better facilitate therapeutic or cognitive interventions. Preclinical research in animal models indicates that behavioral interventions following early life adversity can be potentiated by pharmacological methods that increase plasticity (Miskolczi, Halász, & Mikics, 2019). (For additional translational issues related to sensitive periods for sociocultural processing in adolescence, see Blakemore & Mills, 2014).

Conclusions

This review provides a contemporary update to the question as to whether adolescence is a sensitive period for sociocultural processing initially proposed by Blakemore and Mills (2014). We clarified concepts and terms surrounding this question, including distinguishing heightened periods of sensitivity from sensitive periods and calling for greater consideration of culture. Our literature review highlighted recent research in key areas of interest, including structural brain development, social cognitive development, peer influences, and psychobiological calibration during adolescence. Many studies were consistent with the notion of sensitive periods for sociocultural processing, and a few areas -- especially newly developed metrics of structural brain development, face processing, and psychobiological calibration to the caregiving environment -- stand out as providing stronger evidence of such sensitive periods. Overall, however, strong and direct tests of sensitive period effects meeting strict criteria for specificity were rare, and sensitive period accounts of how specific sociocultural processes unfold were somewhat limited. Though adolescents undergo significant sociocultural development, neurobiological evidence suggests that sensitive periods are not restricted to the sociocultural domain, and further suggest that sensitive periods may extend beyond late adolescence. Future research supporting more precise understandings of adolescent sensitive period development might leverage a mix of large-scale, longitudinal data with studies using more targeted sampling and/or intervention designs. More specific investigations may enhance the translational potential of adolescent sensitive periods to support health and well-being through intervention and policy.

Highlights.

• We evaluate adolescence as a sensitive period for sociocultural processing.

• Social and neurobiological changes in adolescence point to a sensitive period.

• Yet, few studies directly and specifically test sensitive period theories.

• More specific definitions and study designs, as well as novel methods, may drive progress.

-- Box 1 --.

Animal Models in Adolescent Sensitive Periods Research: Focus on Gonadal Hormones and Plasticity in the Frontal Cortex

Through gonadectomy and hormone treatments, animal research can experimentally isolate the impact of pubertal hormones on a wide range of behaviors and neural processes. In one study, prepubertal gonadectomy in female mice prevented the shift toward greater inhibitory neurotransmission associated with the onset of sensitive periods in the frontal cortex (see Are Changes in Structural Brain Development Tied to Developmental Plasticity?; Piekarski et al., 2017b). These findings point to a mechanistic role of hormones in shifting E-I balance and triggering relatively late-emerging sensitive periods in the frontal cortex. Fast-spiking parvalbumin positive (PV+) interneurons are a subtype of cortical inhibitory neuron implicated in shifting E-I balance toward sensitive period onset in the frontal cortex (Le Magueresse & Moyer, 2013). As PV+ interneurons express estrogen receptor beta, they have been posited as a link through which gonadal hormones regulate plasticity within the frontal cortex (Delevich et al., 2021; Piekarski et al., 2017a). Overall, strengthening our understanding of the contributions of pubertal hormones to neurodevelopment through animal models may build testable neurobiological mechanisms of sensitive periods in humans.

-- Box 2 –

Animal Models in Adolescent Sensitive Periods Research: Focus on Social Isolation

Animal research, primarily carried out in social species of rodents, has identified potent and causal effects of social isolation occurring in the juvenile and peri-pubertal periods (Burke et al., 2017). In general, isolating rodents during a post-weaning, pre-adulthood period increases aggression (Toth et al., 2011), anxiety-like (Lukkes et al., 2009a) behaviors, and preferences for drug and reward-related stimuli (Whitaker et al., 2013; Walker et al., 2020). Critically, these effects are not fully ameliorated by later resocialization (return to group housing), suggesting effects of isolation specific to a period akin to adolescence. Some isolation-related phenotypes did not develop if rodents who were otherwise housed alone received an hour of daily contact with a peer that engaged in social play (Einon et al., 1978). Furthermore, previously-isolated rodents who were resocialized with other isolates saw fewer gains in myelin and mPFC activity than those resocialized with rodents who had always been socially housed (Makinodan et al., 2017). Taken together, this work suggests that social isolation during the peri-pubertal period has lasting effects, while high-quality social experiences can both prevent and support partial recovery from the effects of social isolation in rodents.

Social isolation impacts rodents’ behavior via multiple mechanisms, including altering dopaminergic (e.g., Fabricius et al., 2010; Yorgason et al., 2016) and serotonergic (e.g., Lukkes et al., 2009b) functioning within mesolimbic regions fundamentally impacting processing reward and value (the ventral tegmental area and nucleus accumbens), as well as decreased inhibitory activity within aspects of the amygdala, a key region in emotional and fear processing (Lukkes et al., 2012; for reviews, see Burke et al., 2017; Novick et al., 2018; Orben et al., 2020). Studies have also identified altered prefrontal myelination, dendritic pruning, and oligodendrocyte functioning following isolation during juvenile and adolescent phases of rodent development (Makinodan et al., 2012), with impacts on learning and goal-directed behaviors (Hinton et al., 2019).

Mapping rodent social isolation studies to the developmental timing of human adolescence is challenging and is further complicated by the fact that a wide variety of isolation timings and durations are employed across the animal literature. Nevertheless, this research provides insight into what behaviors and circuits may be impacted by social experiences in adolescent-like phases of development across species. While lengthy social isolation mechanisms cannot be ethically implemented in humans, researchers have recently begun using acute isolation paradigms with late adolescents to understand how such experiences can alter reward processing and decision-making (Tomova et al., preprint).