Hostname: page-component-586b7cd67f-rdxmf Total loading time: 0 Render date: 2024-11-27T12:02:13.709Z Has data issue: false hasContentIssue false

Functional connectivity between the nucleus accumbens and amygdala underlies avoidance learning during adolescence: Implications for developmental psychopathology

Published online by Cambridge University Press:  26 September 2024

Benjamin M. Rosenberg*
Affiliation:
Department of Psychology, University of California, Los Angeles (UCLA), Los Angeles, CA, USA
João F. Guassi Moreira
Affiliation:
Department of Psychology, University of California, Los Angeles (UCLA), Los Angeles, CA, USA
Adriana S. Méndez Leal
Affiliation:
Department of Psychology, University of California, Los Angeles (UCLA), Los Angeles, CA, USA
Natalie M. Saragosa-Harris
Affiliation:
Department of Psychology, University of California, Los Angeles (UCLA), Los Angeles, CA, USA
Elizabeth Gaines
Affiliation:
Department of Psychology, University of California, Los Angeles (UCLA), Los Angeles, CA, USA
Wesley J. Meredith
Affiliation:
Department of Psychology, University of California, Los Angeles (UCLA), Los Angeles, CA, USA
Yael Waizman
Affiliation:
Department of Psychology, University of Southern California, Los Angeles, CA, USA
Emilia Ninova
Affiliation:
College of Social Work, Florida State University, Tallahassee, FL, USA
Jennifer A. Silvers
Affiliation:
Department of Psychology, University of California, Los Angeles (UCLA), Los Angeles, CA, USA
*
Corresponding author: Benjamin M. Rosenberg; Email: [email protected]
Rights & Permissions [Opens in a new window]

Abstract

Background:

Reward and threat processes work together to support adaptive learning during development. Adolescence is associated with increasing approach behavior (e.g., novelty-seeking, risk-taking) but often also coincides with emerging internalizing symptoms, which are characterized by heightened avoidance behavior. Peaking engagement of the nucleus accumbens (NAcc) during adolescence, often studied in reward paradigms, may also relate to threat mechanisms of adolescent psychopathology.

Methods:

47 typically developing adolescents (9.9–22.9 years) completed an aversive learning task during functional magnetic resonance imaging, wherein visual cues were paired with an aversive sound or no sound. Task blocks involved an escapable aversively reinforced stimulus (CS+r), the same stimulus without reinforcement (CS+nr), or a stimulus that was never reinforced (CS−). Parent-reported internalizing symptoms were measured using Revised Child Anxiety and Depression Scales.

Results:

Functional connectivity between the NAcc and amygdala differentiated the stimuli, such that connectivity increased for the CS+r (p = .023) but not for the CS+nr and CS−. Adolescents with greater internalizing symptoms demonstrated greater positive functional connectivity for the CS− (p = .041).

Conclusions:

Adolescents show heightened NAcc-amygdala functional connectivity during escape from threat. Higher anxiety and depression symptoms are associated with elevated NAcc-amygdala connectivity during safety, which may reflect poor safety versus threat discrimination.

Type
Regular Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2024. Published by Cambridge University Press

Introduction

Adolescence is characterized by increasing autonomy and pursuit of novel experiences, corresponding with heightened behavioral and cognitive flexibility (Crone & Dahl, Reference Crone and Dahl2012). During adolescence, changing environmental demands coincide with neurodevelopmental shifts that motivate exploration of one’s environment, while still protecting survival. Prevailing theories in developmental affective neuroscience and developmental psychopathology point to changes in both reward and threat circuitry during adolescence as being, respectively, important for approach and avoidance motivations, and more generally for healthy development and well-being (Baker & Galván, Reference Baker and Galván2020; Gee et al., Reference Gee, Bath, Johnson, Meyer, Murty, van den Bos and Hartley2018; Silk et al., Reference Silk, Davis, McMakin, Dahl and Forbes2012). However, relatively little research has looked at how these systems may work together to support adaptive associative and instrumental learning processes. The present study aims to (1) identify interactions between canonical threat and reward neurocircuitry during aversive learning and (2) explore potential links to emerging symptoms of anxiety and depression during adolescence.

Threat and safety learning

Threat and safety learning are considered central to the etiology and treatment of anxiety disorders, which are known to emerge or worsen during adolescence (Baker & Galván, Reference Baker and Galván2020; Shechner et al., Reference Shechner, Hong, Britton, Pine and Fox2014). Relatively young children can learn cues (i.e., conditional stimuli; CSs) that predict threat or safety and provide a foundation for emerging threat appraisals (Britton et al., Reference Britton, Lissek, Grillon, Norcross and Pine2011). However, growing evidence suggests that the ability to discriminate between dangerous and safe CSs emerges continues to improve across childhood (Grasser & Jovanovic, Reference Grasser and Jovanovic2021; Jovanovic et al., Reference Jovanovic, Nylocks, Gamwell, Smith, Davis, Norrholm and Bradley2014), throughout adolescence (Reinhard et al., Reference Reinhard, Slyschak, Schiele, Andreatta, Kneer, Reif, Domschke, Gamer, Pauli, Deckert and Romanos2022), and into adulthood (Lau et al., Reference Lau, Britton, Nelson, Angold, Ernst, Goldwin, Grillon, Leibenluft, Lissek, Norcross, Shiffrin and Pine2011). Furthermore, preliminary evidence suggests that adolescents tend to show less differentiation of threatening and safe stimuli than adults, corresponding with heightened avoidance of safe stimuli and overgeneralization of fear to perceptually similar stimuli (Klein et al., Reference Klein, Berger, Vervliet and Shechner2021). There is mixed evidence regarding age-related differences in fear extinction (potentially attributable to methodological differences, see Stenson et al., Reference Stenson, France and Jovanovic2023), with some studies suggesting that adolescents compared with adults tend to exhibit attenuated fear extinction (Haddad et al., Reference Haddad, Lissek, Pine and Lau2011; Linton & Levita, Reference Linton and Levita2021; Pattwell et al., Reference Pattwell, Duhoux, Hartley, Johnson, Jing, Elliott, Ruberry, Powers, Mehta, Yang, Soliman, Glatt, Casey, Ninan and Lee2012; for a review, see Baker et al., Reference Baker, Den, Graham and Richardson2014), and others suggesting no difference between adolescents and adults in fear extinction (Abend et al., Reference Abend, Gold, Britton, Michalska, Shechner, Sachs, Winkler, Leibenluft, Averbeck and Pine2020; Waters et al., Reference Waters, Theresiana, Neumann and Craske2017) or fear reinstatement (Den et al., Reference Den, Graham, Newall and Richardson2015; Waters et al., Reference Waters, Theresiana, Neumann and Craske2017). Collectively, these results suggest that the ability to discriminate threat versus safety undergoes continued refinement throughout the adolescent period.

Neuroimaging studies of threat learning in adults have highlighted roles for the amygdala (though amygdala findings are more mixed, see Fullana et al., Reference Fullana, Harrison, Soriano-Mas, Vervliet, Cardoner, Àvila-Parcet and Radua2016), dorsal anterior cingulate cortex (dACC), anterior insular cortex, and ventromedial prefrontal cortex (vmPFC) during fear acquisition (Battaglia et al., Reference Battaglia, Garofalo, di Pellegrino and Starita2020; Etkin & Wager, Reference Etkin and Wager2007; Fullana et al., Reference Fullana, Harrison, Soriano-Mas, Vervliet, Cardoner, Àvila-Parcet and Radua2016; Greco & Liberzon, Reference Greco and Liberzon2016; Milad et al., Reference Milad, Rosenbaum and Simon2014; Sehlmeyer et al., Reference Sehlmeyer, Schöning, Zwitserlood, Pfleiderer, Kircher, Arolt and Konrad2009), as well as the amygdala, hippocampus, and vmPFC in fear extinction (Greco & Liberzon, Reference Greco and Liberzon2016; Milad et al., Reference Milad, Rosenbaum and Simon2014). Relatively few neuroimaging studies of Pavlovian fear processes have been published in adolescents (Treanor et al., Reference Treanor, Rosenberg and Craske2021). However, it has been suggested that differences in fear learning observed in adolescents versus adults may be attributable to adolescents possessing relatively mature threat neurocircuitry (emphasizing the amygdala) in tandem with less mature prefrontal cortex (Grasser & Jovanovic, Reference Grasser and Jovanovic2021; Morriss et al., Reference Morriss, Christakou and Van Reekum2019). For example, adolescents compared with adults tend to show less activation of the vmPFC and dlPFC during fear extinction recall (Ganella et al., Reference Ganella, Drummond, Ganella, Whittle and Kim2018).

Adolescent exploration, reward pursuit, and risk-taking

Exploration and reward pursuit are considered hallmarks of adolescence that have the potential to promote both adaptive and maladaptive behaviors, depending on context (Ciranka & van den Bos, Reference Ciranka and van den Bos2021; Duell & Steinberg, Reference Duell and Steinberg2019; Romer et al., Reference Romer, Reyna and Satterthwaite2017). Reward-related activation of the nucleus accumbens (NAcc), a region of the ventral striatum that is responsible for dopaminergic signaling across a variety of tasks and motivated behaviors (Floresco, Reference Floresco2015; Nicola, Reference Nicola2007), has been shown to peak during adolescence (Braams et al., Reference Braams, van Duijvenvoorde, Peper and Crone2015; Cohen et al., Reference Cohen, Asarnow, Sabb, Bilder, Bookheimer, Knowlton and Poldrack2010; Schreuders et al., Reference Schreuders, Braams, Blankenstein, Peper, Güroğlu and Crone2018; Silverman et al., Reference Silverman, Jedd and Luciana2015; Urošević et al., Reference Urošević, Collins, Muetzel, Lim and Luciana2012). A wealth of prior studies has focused on how increasing striatal engagement coincides with increasing reward pursuit and sensitivity, thereby steering adolescents toward greater engagement in specific exploratory behaviors that promote new learning, including risk-taking (Barkley-Levenson & Galván, Reference Barkley-Levenson and Galván2014; Depasque & Galván, Reference DePasque and Galván2017; Galván et al., Reference Galvan, Hare, Parra, Penn, Voss, Glover and Casey2006; Galván, Reference Galván2010; Somerville et al., Reference Somerville, Jones and Casey2010; Van Duijvenvoorde et al., Reference Van Duijvenvoorde, Peters, Braams and Crone2016; Walker et al., Reference Walker, Bell, Flores, Gulley, Willing and Paul2017). Reinforcement learning rates are higher among adolescents than adults (Topel et al., Reference Topel, Ma, Sleutels, van Steenbergen, de Bruijn and van Duijvenvoorde2023) and increase throughout adolescence (Master et al., Reference Master, Eckstein, Gotlieb, Dahl, Wilbrecht and Collins2020), and this is particularly evident in paradigms involving rewards (Palminteri et al., Reference Palminteri, Kilford, Coricelli and Blakemore2016) or positive feedback (Davidow et al., Reference Davidow, Foerde, Galván and Shohamy2016). Collectively, these studies highlight adolescence as a period characterized by elevated recruitment of reward neurocircuitry and corresponding elevations in reward sensitivity, thereby establishing a basis for learning and a propensity for risk-taking.

Threat and reward: an integrated perspective

While in everyday life reward and threat neurocircuitry ostensibly work closely together to support adaptive learning and decision making (e.g., to support approach and avoidance decisions, see Peris & Galván, Reference Peris and Galván2021), they have been largely studied separately in adolescent samples (although some studies have evaluated threat and reward circuitry in relation to cognitive control of emotions, e.g., Heller et al., Reference Heller, Cohen, Dreyfuss and Casey2016). For example, the adolescent risk-taking literature has largely emphasized development of reward-related processes while ignoring threat processes (Baker & Galván, Reference Baker and Galván2020). That said, some models of adolescent motivated behavior have integrated the development of threat neurocircuitry into conceptualizations of adolescent impulsivity and risk-taking behavior, noting a tendency for peaking NAcc recruitment to “tip” adolescent behaviors away from amygdala-mediated avoidance behaviors and toward risks (Ernst et al., Reference Ernst, Pine and Hardin2006; Ernst, Reference Ernst2014; Richards et al., Reference Richards, Plate and Ernst2012).

Relatedly, it has been theorized that peaking striatal expression in adolescence may have implications for adolescent threat processing (Lago et al., Reference Lago, Davis, Grillon and Ernst2017). Indeed, there is striking evidence that striatal neurocircuitry may be crucial for supporting avoidance of or escape from threats. For example, striatal dopamine signaling, a hallmark of reward prediction error (Schultz, Reference Schultz2013), has been tied to the processing of aversive or threatening stimuli (Brooks & Berns, Reference Brooks and Berns2013; França & Pompeia, Reference França and Pompeia2023; McCutcheon et al., Reference McCutcheon, Ebner, Loriaux and Roitman2012; Verharen et al., Reference Verharen, Zhu and Lammel2020). More specifically, the NAcc is thought to coordinate with threat neurocircuitry (particularly the amygdala, hippocampus, and medial prefrontal cortex) to support fear-related avoidance behaviors in both rodents and humans (for a review, see Wong et al., Reference Wong, Wirth and Pittig2022). Similar neurocircuitry (e.g., amygdala, dACC, insula, NAcc, and vmPFC) is involved in studies of threat responding and defensive behaviors like freezing or escape (Mobbs et al., Reference Mobbs, Petrovic, Marchant, Hassabis, Weiskopf, Seymour, Dolan and Frith2007, Reference Mobbs, Marchant, Hassabis, Seymour, Tan, Gray, Petrovic, Dolan and Frith2009, Reference Mobbs, Headley, Ding and Dayan2020). In humans, the NAcc has been shown to signal imminence of both rewards and threats (Murty et al., Reference Murty, Song, Surampudi and Pessoa2023). It has likewise been theorized that the vmPFC coordinates with the ventral striatum (including the NAcc) to assess the safety or threat of a given situation, and to make subsequent decisions about how to respond (Tashjian et al., Reference Tashjian, Zbozinek and Mobbs2021). In sum, there are compelling theoretical and empirical reasons to believe that striatal reward circuitry contributes to adaptive learning and responding to threats – for example, by coordinating escape or avoidance behavior, at least in adults. However, it is relatively unknown how these brain-behavior links change during adolescent development (França & Pompeia, Reference França and Pompeia2023).

Other theoretical work suggests that positively valenced systems are inextricably linked to threat processing – for example, the removal of threat can be an inherently rewarding experience (for a review, see Rosenberg, Barnes-Horowitz, et al., Reference Rosenberg, Barnes-Horowitz, Zbozinek and Craske2024). More specifically, striatal dopamine signals the unexpected omission of aversive outcomes during Pavlovian fear extinction (Kalisch et al., Reference Kalisch, Gerlicher and Duvarci2019; Gentry et al., Reference Gentry, Schuweiler and Roesch2019; Salinas-Hernández & Duvarci, Reference Salinas-Hernández and Duvarci2021) corresponding with subjective experiences of relief (Willems et al., Reference Willems, Van Oudenhove and Vervliet2023), a positive emotion that is experienced as similarly rewarding compared with monetary gains (Leng et al., Reference Leng, Beckers and Vervliet2023) and is moderated by reward sensitivity (Leng et al., Reference Leng, Beckers and Vervliet2022, Reference Leng, Beckers and Vervliet2024). The positively valenced experience of relief is thought to (1) reinforce escape from the predicted dangerous outcomes (unconditioned stimulus; US) or instrumental avoidance of the CS (Carver, Reference Carver2009; Deutsch et al., Reference Deutsch, Smith, Kordts-Freudinger and Reichardt2015), and (2) coincide with prediction error learning when an anticipated fearful outcome surprisingly does not occur (i.e., during Pavlovian fear extinction) (Vervliet et al., Reference Vervliet, Lange and Milad2017). Positive experiences of threat omission may further relate to other positive emotions, such as “thrill,” that may be particularly desirable among adolescents and may relate to peaking NAcc expression during this period (Spielberg et al., Reference Spielberg, Olino, Forbes and Dahl2014). Additional research is needed to understand how reward and threat processes support one another, especially in adolescence.

Adolescence and the emergence of psychopathology

Adolescence coincides with peak incidence rates across a range of internalizing psychopathology, including anxiety and depression (Kessler et al., Reference Kessler, Angermeyer, Anthony, De Graaf, Demyttenaere, Gasquet and Üstün2007, 2012; Merikangas et al., Reference Merikangas, He, Burstein, Swanson, Avenevoli, Cui, Benjet, Georgiades and Swendsen2010; Paus et al., Reference Paus, Keshavan and Giedd2008; Powers & Casey, Reference Powers and Casey2015; Rapee et al., Reference Rapee, Oar, Johnco, Forbes, Fardouly, Magson and Richardson2019). Aversive learning paradigms (e.g., Pavlovian fear acquisition) may offer a window into this phenomenon, as poor discrimination of threat and safety cues during development can lead to overuse of behavioral strategies that maintain or increase fear, such as avoidance (Waters & Craske, Reference Waters and Craske2016). Likewise, individual differences in positive emotions are thought to moderate aversive learning during adolescence, suggesting that poor discrimination of threat and safety cues may also relate to emerging depression symptoms (e.g., anhedonia) (Rosenberg, Barnes-Horowitz, et al., Reference Rosenberg, Barnes-Horowitz, Zbozinek and Craske2024; Rosenberg, Young, et al., Reference Rosenberg, Barnes-Horowitz, Zbozinek and Craske2024).

Variations in the development of both threat and reward neurocircuitry may therefore be central to the onset of adolescent psychopathology (Baker & Galván, Reference Baker and Galván2020; Xie et al., Reference Xie, Zhang, Cheng and Yang2021; Zimmermann et al., Reference Zimmermann, Richardson and Baker2019). For example, during late adolescence, heightened responsivity of threat neurocircuitry (particularly the vmPFC) during Pavlovian fear conditioning is associated with increases in anxiety-specific symptoms from late adolescence into young adulthood (Peng et al., Reference Peng, Knotts, Young, Bookheimer, Nusslock, Zinbarg, Kelley, Echiverri-Cohen and Craske2023). Likewise, aberrations in reward sensitivity and associated neurocircuitry (e.g., low striatal recruitment, less striatal coordination with limbic regions) are associated with elevated risk avoidance behaviors among anxious adolescents (Baker et al., Reference Baker, Padgaonkar, Galván and Peris2024; Galván & Peris, Reference Galván and Peris2014; Peris & Galván, Reference Peris and Galván2021). Considering potential interactions between these systems, individual differences in anhedonia symptoms (i.e., low reward sensitivity or motivation) are associated with (1) atypical patterns of brain activity during Pavlovian fear extinction (Young et al., Reference Young, Bookheimer, Nusslock, Zinbarg, Damme, Chat and Craske2021; Rosenberg et al., Reference Rosenberg, Taschereau-Dumouchel, Lau, Young, Nusslock, Zinbarg and Craske2023) as well as (2) less acquisition of Pavlovian fear leading to greater generalization of fear to safe stimuli that persists into young adulthood (Rosenberg, Young, et al., Reference Rosenberg, Barnes-Horowitz, Zbozinek and Craske2024).

The present study

The present study examined interactions between threat and reward neurocircuitry among typically developing adolescents during an aversive learning task involving three types of CS: a CS reinforced with an aversive sound (CS+r), an identical CS presented without reinforcement (CS+nr), or a different CS that was never reinforced (CS−). During CS+r trials, participants were able to terminate the aversive sound (i.e., escape) by quickly pressing a button. We focused our analyses on the NAcc as well as regions that have been implicated in studies of threat learning (amygdala, vmPFC, hippocampus, insular cortex, dACC). Considering theories of how threat and reward neurocircuitry interact during the instrumental removal of threat (e.g., escape, avoidance) or unexpected omission of threat (e.g., fear extinction, threat prediction error) (Deutsch et al., Reference Deutsch, Smith, Kordts-Freudinger and Reichardt2015; Vervliet et al., Reference Vervliet, Lange and Milad2017), and considering existing literature on the coordination between the NAcc and threat-related neurocircuitry among rodents or adult humans (e.g., Wong et al., Reference Wong, Wirth and Pittig2022), we hypothesized that functional connectivity between the NAcc and threat neurocircuitry would be greater for stimuli associated with threat termination/omission (CS+r, CS+nr), and potentially greatest for the CS+r, compared with the stimulus involving no threat (CS−). Given that participants had not seen the stimuli prior to this task, and given the relatively short duration of task blocks, we hypothesized that differences among the blocks would emerge and be most notable during the final task blocks. Given prior research suggesting that anxiety and depression are associated with aberrances in safety processing and related neurocircuitry (Grasser & Jovanovic, Reference Grasser and Jovanovic2021; Odriozola & Gee, Reference Odriozola and Gee2021; Pittig et al., Reference Pittig, Treanor, LeBeau and Craske2018), we additionally explored whether symptoms of anxiety and depression were associated with functional connectivity between the NAcc and threat neurocircuitry during the task.

Materials and methods

Participants

Participants were 47 typically developing individuals ages 9.9–22.9 years (24 female, mean age = 15.23 years, SD = 3.75 years). The sample included nine individuals ages 9.9–11.9 years old, 26 individuals ages 11.9–17.9 years old, and 12 individuals older than 18 years old. Demographics are summarized in Table 1. These participants were part of a broader longitudinal study investigating the impact of early life experiences on the neural bases of socioemotional development. Participants included in the current set of analyses were those who provided usable data from an fMRI scanning session and did not have a history of early social deprivation. All research was completed at the University of California, Los Angeles (UCLA) and was approved by the UCLA Institutional Review Board. All minor participants provided informed assent, and their parents provided informed consent, and all 18+ participants provided consent.

Table 1. Participant demographics in the study

Two subjects were excluded for excess head motion (as described below, this was established as participants with > 20% of volumes having average framewise displacement exceeding 0.9 mm or global BOLD signal changes above 5 standard deviations), yielding a final sample of n = 45 subjects. Two subjects had incomplete CS+nr data and were therefore excluded from omnibus tests (n = 43). These subjects were included in multilevel models (n = 45) with incomplete CS+nr blocks included as missing data.

Aversive-learning task

Participants completed an aversive-learning task while inside of an MRI scanner. This task was identical to a version described elsewhere (see Silvers et al., Reference Silvers, Lumian, Gabard-Durnam, Gee, Goff, Fareri, Caldera, Flannery, Telzer, Humphreys and Tottenham2016). Briefly, on each trial, participants viewed one of two shapes. Trials were organized into blocks in which a CS was reinforced with an aversive sound (CS+r), the same CS was not reinforced with an aversive sound (CS+nr), or a different stimulus was not reinforced with an aversive sound (CS−) (Figure 1). On each trial, the shape presented initially had a thin border that became thick after 1000ms. Participants were instructed to make a button response as soon as they saw the border of the shape thicken. During CSr trials, at the same time that the border would begin to change, an aversive noise (US) started. Though participants were not told so, the button press terminated each trial and temporarily extinguished the US during CSr blocks.

Figure 1. Visual representation of the aversive learning paradigm including three types of trial blocks: CS+r, CS+nr, and CS−. Participants completed eight total blocks (three CS+r, three CS+nr, two CS−).

The US was a loud, metallic, high-frequency noise (Neumann et al., Reference Neumann, Waters, Westbury and Henry2008) that was titrated for each participant before the task so that it was perceived as “annoying” but not painful (maximum volume, 92 dB). This calibration has been previously used in studies of aversive learning (e.g., Silvers et al., Reference Silvers, Lumian, Gabard-Durnam, Gee, Goff, Fareri, Caldera, Flannery, Telzer, Humphreys and Tottenham2016). The US and CS+r co-terminated when participants responded and another trial immediately began.

Participants completed eight 27 s task blocks (three CS+r blocks, three CS+nr blocks, and two CS − blocks) lasting 10–15 trials each (with exact length depending on RTs during a given block). Assignment of CS + and CS − to shape was counterbalanced across participants.

Symptoms of anxiety and depression

Of the recruited sample, n = 36 parents of the included participants completed the Revised Child Anxiety and Depression Scales (RCADS-P; Chorpita et al., Reference Chorpita, Yim, Moffitt, Umemoto and Francis2000). The RCADS-P is a 47-item questionnaire that probes anxiety symptomology (e.g., “All of a sudden my child will feel really scared for no reason at all”) and depression symptomatology (e.g., “Nothing is much fun for my child anymore”) with answer options ranging from 0 to 3 (0 = Never; 1 = Sometimes; 2 = Often; 3 = Always). The range of raw RCADS-P scores is 0–141, with higher scores suggesting greater levels of anxiety and depression. RCADS-P scores can be further separated into subscales including a 37-item anxiety subscale and 10-item major depression subscale. Total scores are converted to T-scores to enhance clinical utility and interpretability.

fMRI data acquisition and analysis

Acquisition

Imaging data were acquired on a 3T Siemens Prisma scanner using a 32-channel head coil and a parallel image acquisition system (GRAPPA). A high-resolution T1-weighted, MPRAGE image was acquired for registration to functional runs (TR = 2400 ms, TE = 2.22 ms, flip angle = 8°, FOV = 256 mm2, 0.8 mm3 isotropic voxels, 208 slices). Functional images were acquired using a T2* EPI BOLD sequence. 33 axial slices were collected with a TR of 2000 ms and a 3 × 3 × 4 mm3 voxel resolution (TE = 30 ms, flip angle = 75°, FOV = 192 mm2). Participants completed the aversive-learning task by using a head-mounted mirror on the coil to view an LCD back projector screen.

Preprocessing

Before preprocessing, functional images were visually inspected for artifacts and biological abnormalities. No images contained obvious artifacts or biological abnormalities that warranted exclusion from further analysis. The scans for n = 2 subjects terminated prior to completion of the full task, yielding usable data for seven task blocks (i.e., incomplete third CS+nr block). Incomplete blocks for these subjects were modeled as missing data during analyses (see below).

We implemented the default preprocessing pipeline for volume-based analyses in the CONN functional connectivity toolbox v22a (Whitfield-Gabrieli & Nieto-Castanon, Reference Whitfield-Gabrieli and Nieto-Castanon2012)Footnote 1 . Images were realigned using the default procedure in SPM12 (Andersson et al., Reference Andersson, Hutton, Ashburner, Turner and Friston2001). We applied slice-timing correction for interleaved acquisition using the default procedure in SPM12 (Henson et al., Reference Henson, Buechel, Josephs and Friston1999). Outlier volumes were identified and censored using the Artifact Detection Tools (ART) software in CONN applying the toolbox’s default settings (acquisitions with framewise displacement above 0.9 mm or global BOLD signal changes above 5 standard deviations). Data were then spatially normalized into the standard Montreal Neurological Institute (MNI) space (Friston et al., Reference Friston, Holmes, Worsley, Poline, Frith and Frackowiak1994), resliced to 2 mm × 2 mm × 2 mm voxels, and smoothed using a Gaussian kernel with a full-width at half-maximum (FWHM) of 6 mm.

First-level fMRI analyses

Six motion regressors (x, y, and z displacement; pitch, roll, and yaw rotation) and their first- and second-level derivatives were included as first-level covariates. Physiological noise was controlled with CompCor, an algorithm in which the timeseries of activation is extracted from subject-specific tissue masks (white matter, cerebrospinal fluid), and principal components analysis is applied to estimate physiological noise reflected in these timeseries, after which the resulting components are included as covariates in a denoising regression (for additional details on this approach, see Whitfield-Gabrieli & Nieto-Castanon, Reference Whitfield-Gabrieli and Nieto-Castanon2012). Finally, we applied a band-pass filter of 0.008–0.09 Hz to further remove high-frequency activity associated with physiological functioning and low-frequency activity associated with scanner drift.

Task performance

Task-based reaction times (RTs) were evaluated for each trial throughout the task. We calculated mean RTs for each participant during each task block. Mean RTs were winsorized for each task block, such that RTs below the 5th percentile were replaced by RT values at the 5th percentile, and RTs above the 95th percentile were replaced by RT values at the 95th percentile.

We conducted multilevel modeling in Stata 18.0 to test the main effect of Block (covarying for Age, Block, and Sex), the main effect of Stimulus Type (covarying for Age, Block, and Sex), and the interaction between Stimulus Type × Block in predicting RTs (covarying for Age and Sex). Models included random effects of the intercept and slope for each subject and fixed effects for Stimulus Type and Block as a within-subject factor. Stimulus Type was modeled as categorical variables in all analyses. We did not model Block as a continuous variable, as there were an unequal number of blocks for each stimulus type (i.e., such a model would estimate non-existent RT values for CS− Block 3). Therefore, Block was modeled as a categorical variable. Significance of the Stimulus Type × Block interaction was determined by comparing models with and without the interaction term included (using the lrtest function in Stata).

Functional connectivity

Using the Harvard-Oxford atlas, we extracted one ROI for bilateral NAcc as the seed region. We additionally extracted four ROIs for threat neurocircuitry (bilateral amygdala, bilateral hippocampus, bilateral insular cortex, and dACC). As the Harvard-Oxford atlas does not include a ROI for vmPFC, we used a 5 mm sphere centered on coordinates from a prior meta-analysis on Pavlovian fear learning (Fullana et al., Reference Fullana, Harrison, Soriano-Mas, Vervliet, Cardoner, Àvila-Parcet and Radua2016). These five ROIs were binarized and combined into a single ROI capturing the Threat Network. When significant results were identified, individual ROIs were analyzed separately in follow-up analyses (see below).

We conducted generalized psychophysiological interaction (gPPI) analyses (Friston et al., Reference Friston, Buechel, Fink, Morris, Rolls and Dolan1997) to examine NAcc connectivity with the Threat Network during the task. For each of the 8 task blocks, we computed beta weights (covarying for age, sex, and mean framewise displacement) for connectivity between the NAcc and each of the five target ROIs. We used a between-conditions omnibus contrast in CONN identifying any significant pairwise effects of task block to determine if connectivity between the NAcc and Threat Network was significantly associated with the task (p-unc < .05). Follow-up analyses evaluated the same contrast for each of the five ROIs within the Threat Network (p-unc < .05).

For significant target ROIs identified in the initial gPPI analysis, beta weights for each of the eight blocks were extracted from the CONN Toolbox and imported into Stata 18.0 for multilevel modeling of connectivity throughout the task. This analysis enabled a comparison across specific stimulus/block combinations. Using the same statistical procedure outlined above in 2.5.4 Task Performance, we tested the interaction between Stimulus Type × Block in predicting connectivity betas (covarying for Age, Sex, and Mean Framewise Displacement).

Association with internalizing symptoms

For significant ROI results, follow-up analyses used multilevel modeling to evaluate the association between RCADS-P scores and gPPI beta weights for the stimuli across task blocks. Considering the reduced sample size of participants with RCADS-P data (n = 36), analyses focused principally on main effects of RCADS-P (covarying for Age, Stimulus Type, Block, and Sex) in predicting connectivity betas. We did not additionally analyze an Age × RCADS-P interaction due to power concerns and considering the exploratory nature of Age analyses (see below). Exploratory analyses also evaluated a RCADS-P × Stimulus Type interaction (covarying for Age, Block, and Sex). Similar analyses were also conducted to explore a main effect of RCADS-P or a RCADS-P × Stimulus Type interaction in predicting RTs during the task, as these behavioral results may provide added context for interpreting the neural data.

Exploratory analyses of age effects

As the ability to discriminate threat and safety tends to improve throughout adolescence, it is possible that functional connectivity between the NAcc and Threat Network becomes increasingly associated with threat discrimination as adolescents get older. Therefore, exploratory analyses tested the main effect of Age (covarying for Block, Stimulus Type, Sex, and Mean Framewise Displacement), Age × Stimulus Type interaction (covarying for Block, Sex, and Mean Framewise Displacement), and Age × Stimulus Type × Block interaction (covarying for Sex and Mean Framewise Displacement) in predicting connectivity betas for NAcc-ROI pairs significantly implicated in the main analyses. Similar analyses were also conducted to explore a main effect of Age (covarying for Stimulus Type, Block, and Sex), Age × Stimulus Type interaction (covarying for Block and Sex), and Age × Stimulus Type × Block interaction (covarying for Sex) in predicting RTs during the task, as these behavioral results may provide added context for interpreting the neural data.

Supplementary analyses

Exploratory analyses applied a seed-to-voxel approach to evaluate NAcc connectivity across the whole brain during the different task blocks (CS+r > CS−, CS+nr > CS−). These analyses are reported in the supplement. Furthermore, as the task is not optimized for a comparison of between-subjects differences in activation, we did not have a priori hypotheses about activation in this study. Univariate activation analyses are included in the supplement for ROIs implicated in significant functional connectivity results.

Results

The aversive learning task involved blocks of CS+r, CS+nr, and CS− trials. We hypothesized that functional connectivity between the NAcc and threat neurocircuitry would be greatest for the CS+r and CS+nr compared with the CS− by the end of the task. For significant target ROIs, we additionally investigated associations between task functional connectivity and symptoms of anxiety and depression. Finally, we explored age effects to determine if significant target NAcc-ROI connectivity is increasingly associated with threat discrimination as adolescents get older.

Reaction times

There was a significant main effect of Block (b_Block2 = −18.98, b_Block3 = −22.02, χ 2(2) = 11.52, p = .003), such that RTs tended to get faster across task blocks. There was a significant main effect of Stimulus Type (b_CS+nr = 19.87, b_CS− = 4.26, χ 2(2) = 9.77, p = .008), such that RTs for the CS+r tended to be faster than RTs for the CS+nr (Z = 3.02, p = .003) but not the CS− (Z = .55, p = .581). There was a significant Stimulus Type × Block interaction (b_CS+nr_Block2 = 21.29, b_CS−_Block2 = 30.66, b_CS+nr_Block3 = 57.72, χ 2(3) = 15.51, p = .001), such that RTs were faster for the CS+r Block 3 compared with the CS+nr Block 3 (Z = 4.60, p < .001) and CS− Block 2 (Z = 2.67, p = .007). Broadly, these results provide behavioral evidence that individuals discriminated between the stimuli.

Functional connectivity (gPPI) analyses

Omnibus test of task effects

There was a significant effect for the omnibus contrast (F(8,32) = 2.32, p = .044), indicating that connectivity between the NAcc and Threat Network differed significantly as a function of condition (i.e., stimulus/block combinations) during the task. Follow-up ROI-specific analyses found a significant association between the NAcc and bilateral amygdala (F(8,32) = 2.28, p = .046) and the dACC (F(8,32) = 2.51, p = .031), but not the NAcc and the bilateral hippocampus (F(8,32) = 1.45, p = .214), bilateral insula (F(8,32) = 1.09, p = .393), or vmPFC (F(8,32) = 0.74, p = .653). To better interpret these findings, multilevel model results are reported below.

Multilevel modeling

Bilateral amygdala connectivity

The full model significantly explained variance in connectivity between the NAcc and bilateral amygdala (χ2(10) = 18.49, p = .047). There was an effect nearing significance for a Stimulus Type × Block interaction (b_CS+nr_Block2 = .09, b_CS−_Block2 = −.19, b_CS+nr_Block3 = −.39, χ 2(4) = 8.69, p = .069). Follow-up analyses found a significant simple main effect of Block for the CS+r (b_Block2 = −.07, b_Block3 = .29, χ 2(2) = 7.59, p = .023; Figure 2), a marginal effect for the CS− (b_Block2 = −.25, χ 2(1) = 3.41, p = .065), and no effect for the CS+nr (b_Block2 = .02, b_Block3 = −.09, χ 2(2) = 1.07, p = .586). Pairwise comparisons similarly showed significantly greater connectivity for the CS+r by the end of the task compared with earlier CS+r blocks and compared with CS+nr or CS− blocks (Table 2).

Figure 2. NAcc-amygdala functional connectivity differentiated stimuli by the end of the task, such that connectivity tended to increase for the CS+r but decrease for the CS+nr and the CS−.

Table 2. Covariate results and significant pairwise contrasts among stimulus/block pairs in the NAcc-amygdala connectivity analyses

dACC connectivity

The full model did not significantly explain variance in NAcc-dACC connectivity (χ 2(10) = 11.37, p = .330). There was not a significant Stimulus Type × Block interaction (b_CS+nr_Block2 = .04, b_CS−_Block2 = −.13, b_CS+nr_Block3 = .36, χ 2(4) = 3.89, p = .421).

Association with trait anxiety and depression symptoms

NAcc-amygdala connectivity

There was no significant main effect of RCADS-P in predicting connectivity between the NAcc and bilateral amygdala (b = .01, Z = 1.43, p = .154). There was a significant RCADS-P × Stimulus Type interaction in predicting connectivity between the NAcc and bilateral amygdala (b_CS+nr_RCADS = .01, b_CS−_RCADS = .04, χ 2(2) = 6.37, p = .041), such that participants with greater trait anxiety and depression symptoms showed stronger positive connectivity for the CS− but not the CS+r or CS+nr (Figure 3).

Figure 3. Greater internalizing symptoms were associated with greater NAcc-amygdala functional connectivity for the CS−, but not CS+r or CS+nr, during the task.

Reaction time

There was a significant effect of RCADS-P over and above Age, Stimulus Type, Block, and Sex (b = −2.58, Z = −2.33, p = .020), such that participants with greater symptoms exhibited faster RTs. There was also a significant RCADS-P × Stimulus Type interaction (b_CS+nr_RCADS = −2.66, b_CS−_RCADS = −.58, χ 2(2) = 7.50, p = .024), such that participants with greater symptoms showed faster RTs specifically for the CS+nr, but not the CS+r or CS−.

Exploratory analyses of age effects

Bilateral amygdala connectivity

There was no significant main effect of Age (b = .02, Z = 1.44, p = .150) or Age × Stimulus Type interaction (b_CS+nr _Age = −.01, b_CS−_Age = .02, χ 2(2) = 1.87, p = .392) in predicting NAcc-amygdala connectivity during the task. There was a significant Age × Stimulus Type × Block interaction in predicting NAcc-amygdala connectivity during the task (b_CS+nr_Block2_Age = .04, b_CS−_Block2_Age = .17, b_CS+nr_Block3_Age = −.07, χ 2(3) = 16.87, p = .001), such that connectivity for the CS+r increased for older but not younger participants, and connectivity for the CS− decreased for younger but not older participants (Figure 4).

Figure 4. NAcc-amygdala functional connectivity during the task among participants younger (left) or older (right) than the median age in the sample (14.62 years). Age is shown categorically for interpretability. All analyses were conducted on continuous measures of age.

dACC connectivity

There was no significant main effect of Age (b = .00, Z = 0.30, p = .734), Age × Stimulus Type interaction (b_CS+nr _Age = −.01, b_CS−_Age = .00, χ 2(2) = .17, p = .917), or Age × Stimulus Type × Block (b_CS+nr_Block2_Age = .04, b_CS−_Block2_Age = .09, b_CS+nr_Block3_Age = −.01, χ 2(3) = 3.33, p = .344) interaction in predicting NAcc-dACC connectivity during the task.

Reaction time

There was a significant main effect for Age in predicting RT (b = −6.81, Z = −3.95, p < .001), such that older participants had faster RTs than younger participants throughout the task. There was no significant Age × Stimulus Type interaction in predicting RTs (b_CS+nr _Age = −.47, b_CS−_Age = 1.15, χ 2(2) = .66, p = .719), suggesting that older participants had faster RTs regardless of stimulus type.

Discussion

This study evaluated functional connectivity between reward and threat neurocircuitry among adolescents during an aversive learning task that included a condition with an opportunity to escape threat. Over the course of the task, we found that functional connectivity between the NAcc and bilateral amygdala became increasingly positive for an escapable aversively reinforced stimulus (CS+r), stable and nonsignificant (i.e., noncorrelated activity) for the same stimulus when it was not reinforced (CS+nr), and marginally negative for a stimulus that was never reinforced (CS−). These results suggest that, among adolescents, NAcc-amygdala connectivity may play an important role during escape from threats but not during safety. To our knowledge, the present study is the first to highlight a NAcc-amygdala circuit during adaptive threat responding in adolescent humans. Furthermore, individual differences in trait anxiety and depression symptoms were associated with more positive NAcc-amygdala connectivity for the CS−, which may reflect a link between adolescent psychopathology and aberrant safety learning. Collectively, these findings provide a novel perspective on how threat and reward neurocircuitry support adaptive behavior during adolescent development.

Reward processes in threat learning

Reward processes are considered central to survival behaviors, such as escape and avoidance of danger, and may work in part by activating reward-related neurocircuitry that supports positive emotional experiences (e.g., relief) to reinforce escape or avoidance (Rosenberg, Barnes-Horowitz, et al., Reference Rosenberg, Barnes-Horowitz, Zbozinek and Craske2024). Results from the present study support this notion, as the NAcc (frequently considered in the context of reward paradigms) and amygdala (frequently considered in the context of threat paradigms) showed greater coordination during escape from threat versus during safety.

These findings add to a growing literature showing recruitment of canonical reward neurocircuitry in the context of threat learning paradigms. For example, rodent studies have implicated the NAcc in (1) discrimination of threat and safety (Ray et al., Reference Ray, Russ, Walker and McDannald2020, Reference Ray, Moaddab and McDannald2022; Stelly et al., Reference Stelly, Haug, Fonzi, Garcia, Tritley, Magnon, Ramos and Wanat2019), (2) prediction error learning during threat learning (Delgado et al., Reference Delgado, Li, Schiller and Phelps2008; Oleson et al., Reference Oleson, Gentry, Chioma and Cheer2012), and (3) responding to the termination of pain (i.e., negative reinforcement) (Navratilova et al., Reference Navratilova, Xie, Okun, Qu, Eyde, Ci, Ossipov, King, Fields and Porreca2012). Human studies have similarly shown that the NAcc signals more strongly during instrumental avoidance learning than fear extinction (Boeke et al., Reference Boeke, Moscarello, LeDoux, Phelps and Hartley2017; Garrison et al., Reference Garrison, Erdeniz and Done2013), suggesting that the NAcc may play a role during the execution of motivated behaviors (e.g., escape or avoidance) (Delgado et al., Reference Delgado, Jou, LeDoux and Phelps2009) but not during more passive learning of threat or safety associations (e.g., acquisition or extinction). Indeed, rodent and human studies have highlighted projections from the basolateral amygdala to the NAcc that support goal-directed actions such as instrumental avoidance and escape behavior (e.g., Ramirez et al., Reference Ramirez, Moscarello, LeDoux and Sears2015; for a review, see LeDoux & Daw, Reference LeDoux and Daw2018). The present study demonstrates a similar NAcc-amygdala circuit is evident during escape from threat among a sample of typically developing adolescent humans.

Developmental significance of reward-threat interactions

While the NAcc and associated risk-taking behaviors are central to many neuroscientific accounts of adolescence (Ernst et al., Reference Ernst, Pine and Hardin2006; Ernst, Reference Ernst2014; Richards et al., Reference Richards, Plate and Ernst2012), relatively little work has considered the role that the NAcc may play in threat processes during adolescence. However, as in reward paradigms, threat-related activation of the NAcc also increases during pubertal maturation and is theorized to relate to positive emotions that are triggered by successfully escaping danger (Spielberg et al., Reference Spielberg, Olino, Forbes and Dahl2014). The exploratory analyses of age effects complement and extend this prior work, as older versus younger adolescents exhibited greater threat-related engagement of a NAcc-amygdala circuit for the CS+r. Older adolescents also exhibited faster RTs during the aversive learning task, although this effect was not specific to the CS+r. Together, these results suggest that threat-related NAcc-amygdala connectivity strengthens during adolescence in support of adaptive threat learning.

These findings may be integrated within the broader literature on NAcc engagement during adolescent development. For example, it is possible that increasing NAcc engagement during adolescence not only promotes approach behavior in potentially threatening situations but also serves a protective role by enabling the successful deployment of amygdala-mediated avoidance or escape behaviors in threatening situations. If so, this may have implications for existing conceptualizations of “positive risk-taking” in adolescence, which have traditionally considered the relative benefits and costs associated with the risky behavior (Duell & Steinberg, Reference Duell and Steinberg2019, Reference Duell and Steinberg2021). For example, positive risk-taking behaviors (e.g., initiating a friendship with a new classmate or trying out for a sport) during adolescence may additionally open up new opportunities for adolescents to learn that (1) they are capable of successfully escaping danger, or (2) a given situation is not as dangerous as was originally predicted (i.e., fear extinction), processes which are considered anxiolytic (Craske et al., Reference Craske, Treanor, Zbozinek and Vervliet2022). Rewarding positive emotions, such as relief or thrill, may draw upon similar neurocircuitry and reinforce positive risk-taking behaviors, ultimately promoting adaptive threat learning during adolescence.

Implications for developmental psychopathology

The present study highlighted associations between adolescent internalizing symptoms and elevated NAcc-amygdala connectivity during safety learning. These results are consistent with existing models of developmental psychopathology, wherein inaccurate discrimination of threat and safety cues is considered a hallmark of emerging anxiety symptoms (Britton et al., Reference Britton, Lissek, Grillon, Norcross and Pine2011) and further relates to transdiagnostic symptom dimensions, such as anhedonia, from adolescence into young adulthood (Rosenberg, Young, et al., Reference Rosenberg, Barnes-Horowitz, Zbozinek and Craske2024). If elevated NAcc-amygdala coordination occurs in safe contexts for some adolescents, those adolescents may be more likely to activate motivated behaviors designed to circumvent threats (e.g., avoidance or escape). For example, we found that greater internalizing symptoms were also associated with faster RTs for the safe CS+nr, which may reflect similar inaccuracies in threat and safety discrimination. Such a possibility deserves careful consideration given that contemporary models of clinical anxiety emphasize over-reliance on behavioral avoidance (e.g., in safe contexts), as this can limit opportunities for fear extinction (Beckers et al., Reference Beckers, Hermans, Lange, Luyten, Scheveneels and Vervliet2023; Graham & Milad, Reference Graham and Milad2011; Pittig et al., Reference Pittig, Treanor, LeBeau and Craske2018), increase fear renewal following extinction (Arnaudova et al., Reference Arnaudova, Kindt, Fanselow and Beckers2017), and preserve or even increase conditional fear responses that further reinforce instrumental avoidance behaviors (Lovibond et al., Reference Lovibond, Mitchell, Minard, Brady and Menzies2009; Pittig et al., Reference Pittig, Wong, Glück and Boschet2020). Additional research is needed to test these possibilities.

Importantly, as the peak onset for anxiety and depressive disorders occurs at the same time that NAcc responsivity tends to peak during adolescence (Kessler et al., Reference Kessler, Angermeyer, Anthony, De Graaf, Demyttenaere, Gasquet and Üstün2007, Reference Kessler, Avenevoli, Costello, Georgiades, Green, Gruber and Merikangas2012; Merikangas et al., Reference Merikangas, He, Burstein, Swanson, Avenevoli, Cui, Benjet, Georgiades and Swendsen2010; Paus et al., Reference Paus, Keshavan and Giedd2008; Powers & Casey, Reference Powers and Casey2015; Rapee et al., Reference Rapee, Oar, Johnco, Forbes, Fardouly, Magson and Richardson2019), it is possible that individual differences in safety-related NAcc expression may inform which individuals are most likely to develop symptoms. Indeed, greater NAcc activation during active avoidance has been shown to correlate with anxiety symptoms among adults (Levita et al., Reference Levita, Hoskin and Champi2012), suggesting that emerging aberrances in threat-related NAcc activity may precede anxiety symptoms long-term. Healthy brain and behavioral development during adolescence may therefore involve a tenuous balance across systems responsible for processing rewards and threats to promote adaptive responding in the face of true danger without promoting overreliance on avoidance strategies for managing fear. Future research should evaluate (1) how the NAcc-amygdala circuit relates to safety learning among clinically anxious adolescents, (2) if NAcc-amygdala connectivity during adolescence predicts anxiety symptoms into adulthood, and (3) if emerging symptoms of anxiety become increasingly associated with aberrant NAcc-amygdala connectivity as adolescents get older.

While the present study did not consider what factors drive individual differences in NAcc-amygdala threat processes, it stands to reason that environmental exposures may strongly influence how reward and threat neurocircuitry interact during adolescent development. For example, experiences of early life adversity have been shown to alter both threat and reward circuitry (McLaughlin et al., Reference McLaughlin, DeCross, Jovanovic and Tottenham2019), particularly recruitment of the amygdala and ventral striatum (Fareri & Tottenham, Reference Fareri and Tottenham2016). It is thought that these alterations confer greater risk for psychopathology, in part, due to their association with aberrances in threat and reward learning (McLaughlin et al., Reference McLaughlin, DeCross, Jovanovic and Tottenham2019). For example, escapable threats extinguish more readily than inescapable threats, suggesting that uncontrollable environmental factors during development may be especially likely to confer increased risk for psychopathology (Cohodes et al., Reference Cohodes, Odriozola, Mandell, Caballero, McCauley, Zacharek, Hodges, Haberman, Smith, Thomas, Meisner, Ellis, Hartley and Gee2023; Hartley et al., Reference Hartley, Gorun, Reddan, Ramirez and Phelps2014). Future research is needed to examine if early life adversity alters recruitment of the NAcc-amygdala circuit during aversive learning, thereby biasing behavioral responses toward maladaptive responses (e.g., escape or avoidance in safe contexts).

Limitations and future directions

The present study should be interpreted in light of several limitations. First, results from the present study should be replicated in a larger sample of adolescents across a wider variety of aversive learning paradigms – for example, given the “social reorientation” that occurs during adolescence, it would be especially informative to examine approach and avoidance behavior around valenced social stimuli. Second, while the majority of participants were within an age range typically overlapping with conventional definitions of adolescence, the age range of our sample spans early adolescence to early adulthood. Longitudinal research with an explicit focus on age (rather than exploratory cross-sectional analysis of age as a moderator) is needed to evaluate developing threat and reward neurocircuitry, as this could more definitively test if the NAcc-amygdala aversive learning circuit emerges continuously throughout the adolescent window. Third, although prior studies suggest activation of the NAcc and amygdala both increase during instrumental avoidance (LeDoux & Daw, Reference LeDoux and Daw2018), the present study was not optimized to evaluate directionality of brain connectivity. Fourth, as individual differences in fear and threat processes are most detectable in studies with ambiguous stimuli (i.e., “strong situation effect,” see Lissek et al., Reference Lissek, Pine and Grillon2006), it is possible that symptom associations would not be observed in other unambiguous task designs (e.g., inclusion of CS+r blocks without CS+nr blocks). Finally, as anxiety and depression symptoms were assessed at trait levels, the clinical applicability of these results should be interpreted with caution. Additional research is needed to replicate and extend these findings among clinically anxious or depressed youth and among individuals at risk for psychopathology (e.g., those exposed to early life adversity).

Conclusion

In conclusion, the present study evaluated functional connectivity between the NAcc and canonical threat neurocircuitry among adolescents during an aversive learning task. Results indicated that greater connectivity between the NAcc and bilateral amygdala was associated with escapable threat versus safety. Trait-level anxiety and depression symptoms were associated with greater connectivity between the NAcc and bilateral amygdala during safe blocks, suggesting persistent engagement of defensive neurocircuitry as a mechanism of emerging psychopathology among adolescents. Additional research is needed to replicate and extend these findings in other samples and with other paradigms.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/S095457942400141X.

Acknowledgements

We thank our participants for taking part in this study.

Funding statement

This research was supported by the National Science Foundation (NSF 1848004), the Hellman Fellow Award, and an APF Visionary Award (awarded to JAS). BMR was supported by the National Institute of Mental Health under award number T32MH015750. JFGM, ASML, and NMSG were supported by the National Science Foundation Graduate Research Fellowship (DGE-1650604). JFGM and NMSG were supported by a National Institute of Child Health and Human Development T32 Predoctoral Training Grant. NMSH was supported by the UCLA Edwin W. Pauley Fellowship, and the UCLA Eugene V. Cota-Robles Fellowship Grant. WJM was supported by funding from the UCLA graduate division. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Competing interests

None.

References

Abend, R., Gold, A. L., Britton, J. C., Michalska, K. J., Shechner, T., Sachs, J. F., Winkler, A. M., Leibenluft, E., Averbeck, B. B., & Pine, D. S. (2020). Anticipatory threat responding: Associations with anxiety, development, and brain structure. Biological Psychiatry, 87(10), 916925. https://doi.org/10.1016/j.biopsych.2019.11.006 CrossRefGoogle ScholarPubMed
Andersson, J. L., Hutton, C., Ashburner, J., Turner, R., & Friston, K. (2001). Modeling geometric deformations in EPI time series. Neuroimage, 13(5), 903919. https://doi.org/10.1006/nimg.2001.0746 CrossRefGoogle ScholarPubMed
Arnaudova, I., Kindt, M., Fanselow, M., & Beckers, T. (2017). Pathways towards the proliferation of avoidance in anxiety and implications for treatment. Behaviour Research and Therapy, 96, 313. https://doi.org/10.1016/j.brat.2017.04.004 CrossRefGoogle ScholarPubMed
Baker, A. E., & Galván, A. (2020). Threat or thrill? The neural mechanisms underlying the development of anxiety and risk taking in adolescence. Developmental Cognitive Neuroscience, 45, 100841. https://doi.org/10.1016/j.dcn.2020.100841 CrossRefGoogle ScholarPubMed
Baker, A. E., Padgaonkar, N. T., Galván, A., & Peris, T. S. (2024). Anxiety may alter the role of fronto-striatal circuitry in adolescent risky decision-making. Journal of Affective Disorders, 348, 238247. https://doi.org/10.1016/j.jad.2023.12.063 CrossRefGoogle ScholarPubMed
Baker, K. D., Den, M. L., Graham, B. M., & Richardson, R. (2014). A window of vulnerability: Impaired fear extinction in adolescence. Neurobiology of Learning and Memory, 113, 90100. https://doi.org/10.1016/j.nlm.2013.10.009 CrossRefGoogle ScholarPubMed
Barkley-Levenson, E., & Galván, A. (2014). Neural representation of expected value in the adolescent brain. Proceedings of the National Academy of Sciences, 111(4), 16461651. https://doi.org/10.1073/pnas.1319762111 CrossRefGoogle ScholarPubMed
Battaglia, S., Garofalo, S., di Pellegrino, G., & Starita, F. (2020). Revaluing the role of vmPFC in the acquisition of pavlovian threat conditioning in humans. Journal of Neuroscience, 40(44), 84918500. https://doi.org/10.1523/JNEUROSCI.0304-20.2020 CrossRefGoogle ScholarPubMed
Beckers, T., Hermans, D., Lange, I., Luyten, L., Scheveneels, S., & Vervliet, B. (2023). Understanding clinical fear and anxiety through the lens of human fear conditioning. Nature Reviews Psychology, 2(4), 233245. https://doi.org/10.1038/s44159-023-00156-1 CrossRefGoogle ScholarPubMed
Boeke, E. A., Moscarello, J. M., LeDoux, J. E., Phelps, E. A., & Hartley, C. A. (2017). Active avoidance: neural mechanisms and attenuation of Pavlovian conditioned responding. Journal of Neuroscience, 37(18), 48084818. https://doi.org/10.1523/JNEUROSCI.3261-16.2017 CrossRefGoogle ScholarPubMed
Braams, B. R., van Duijvenvoorde, A. C., Peper, J. S., & Crone, E. A. (2015). Longitudinal changes in adolescent risk-taking: A comprehensive study of neural responses to rewards, pubertal development, and risk-taking behavior. Journal of Neuroscience, 35(18), 72267238. https://doi.org/10.1523/JNEUROSCI.4764-14.2015 CrossRefGoogle ScholarPubMed
Britton, J. C., Lissek, S., Grillon, C., Norcross, M. A., & Pine, D. S. (2011). Development of anxiety: The role of threat appraisal and fear learning. Depression and Anxiety, 28(1), 517. https://doi.org/10.1002/da.20733 CrossRefGoogle ScholarPubMed
Brooks, A. M., & Berns, G. S. (2013). Aversive stimuli and loss in the mesocorticolimbic dopamine system. Trends in Cognitive Sciences, 17(6), 281286. https://doi.org/10.1016/j.tics.2013.04.001 CrossRefGoogle ScholarPubMed
Carver, C. S. (2009). Threat sensitivity, incentive sensitivity, and the experience of relief. Journal of Personality, 77(1), 125138. https://doi.org/10.1111/j.1467-6494.2008.00540.x CrossRefGoogle ScholarPubMed
Chorpita, B. F., Yim, L., Moffitt, C., Umemoto, L. A., & Francis, S. E. (2000). Assessment of symptoms of DSM-IV anxiety and depression in children: A revised child anxiety and depression scale. Behaviour Research and Therapy, 38(8), 835855. https://doi.org/10.1016/S0005-7967(99)00130-8 CrossRefGoogle ScholarPubMed
Ciranka, S., & van den Bos, W. (2021). Adolescent risk-taking in the context of exploration and social influence. Developmental Review, 61, 100979. https://doi.org/10.1016/j.dr.2021.100979 CrossRefGoogle Scholar
Cohen, J. R., Asarnow, R. F., Sabb, F. W., Bilder, R. M., Bookheimer, S. Y., Knowlton, B. J., & Poldrack, R. A. (2010). A unique adolescent response to reward prediction errors. Nature Neuroscience, 13(6), 669671. https://doi.org/10.1038/nn.2558 CrossRefGoogle ScholarPubMed
Cohodes, E. M., Odriozola, P., Mandell, J. D., Caballero, C., McCauley, S., Zacharek, S. J., Hodges, H. R., Haberman, J. T., Smith, M., Thomas, J., Meisner, O. C., Ellis, C. T., Hartley, C. A., & Gee, D. G. (2023). Neural effects of controllability as a key dimension of stress exposure. Development and Psychopathology, 35(1), 218227. https://doi.org/10.1017/S0954579421001498 CrossRefGoogle ScholarPubMed
Craske, M. G., Treanor, M., Zbozinek, T. D., & Vervliet, B. (2022). Optimizing exposure therapy with an inhibitory retrieval approach and the OptEx nexus. Behaviour Research and Therapy, 152, 104069. https://doi.org/10.1016/j.brat.2022.104069 CrossRefGoogle ScholarPubMed
Crone, E. A., & Dahl, R. E. (2012). Understanding adolescence as a period of social-affective engagement and goal flexibility. Nature Reviews Neuroscience, 13(9), 636650. https://doi.org/10.1038/nrn3313 CrossRefGoogle ScholarPubMed
Davidow, J. Y., Foerde, K., Galván, A., & Shohamy, D. (2016). An upside to reward sensitivity: The hippocampus supports enhanced reinforcement learning in adolescence. Neuron, 92(1), 9399. https://doi.org/10.1016/j.neuron.2016.08.031 CrossRefGoogle ScholarPubMed
Delgado, M. R., Jou, R. L., LeDoux, J., & Phelps, L. (2009). Avoiding negative outcomes: Tracking the mechanisms of avoidance learning in humans during fear conditioning. Frontiers in Behavioral Neuroscience, 3, 765. https://doi.org/10.3389/neuro.08.033.2009 CrossRefGoogle ScholarPubMed
Delgado, M. R., Li, J., Schiller, D., & Phelps, E. A. (2008). The role of the striatum in aversive learning and aversive prediction errors. Philosophical Transactions of the Royal Society B: Biological Sciences, 363(1511), 37873800. https://doi.org/10.1098/rstb.2008.0161 CrossRefGoogle ScholarPubMed
Den, M. L., Graham, B. M., Newall, C., & Richardson, R. (2015). Teens that fear screams: A comparison of fear conditioning, extinction, and reinstatement in adolescents and adults. Developmental Psychobiology, 57(7), 818832. https://doi.org/10.1002/dev.21330 CrossRefGoogle ScholarPubMed
DePasque, S., & Galván, A. (2017). Frontostriatal development and probabilistic reinforcement learning during adolescence. Neurobiology of Learning and Memory, 143, 17. https://doi.org/10.1016/j.nlm.2017.04.009 CrossRefGoogle ScholarPubMed
Deutsch, R., Smith, K. J., Kordts-Freudinger, R., & Reichardt, R. (2015). How absent negativity relates to affect and motivation: An integrative relief model. Frontiers in Psychology, 6, 113521. https://doi.org/10.3389/fpsyg.2015.00152 CrossRefGoogle ScholarPubMed
Duell, N., & Steinberg, L. (2019). Positive risk taking in adolescence. Child Development Perspectives, 13(1), 4852. https://doi.org/10.1111/cdep.12310 CrossRefGoogle ScholarPubMed
Duell, N., & Steinberg, L. (2021). Adolescents take positive risks, too. Developmental Review, 62, 100984. https://doi.org/10.1016/j.dr.2021.100984 CrossRefGoogle Scholar
Ernst, M. (2014). The triadic model perspective for the study of adolescent motivated behavior. Brain and Cognition, 89, 104111. https://doi.org/10.1016/j.bandc.2014.01.006 CrossRefGoogle Scholar
Ernst, M., Pine, D. S., & Hardin, M. (2006). Triadic model of the neurobiology of motivated behavior in adolescence. Psychological Medicine, 36(3), 299312. https://doi.org/10.1017/S0033291705005891 CrossRefGoogle ScholarPubMed
Etkin, A., & Wager, T. D. (2007). Functional neuroimaging of anxiety: a meta-analysis of emotional processing in PTSD, social anxiety disorder, and specific phobia. American Journal of Psychiatry, 164(10), 14761488. https://doi.org/10.1176/appi.ajp.2007.07030504 CrossRefGoogle ScholarPubMed
Fareri, D. S., & Tottenham, N. (2016). Effects of early life stress on amygdala and striatal development. Developmental Cognitive Neuroscience, 19, 233247. https://doi.org/10.1016/j.dcn.2016.04.005 CrossRefGoogle ScholarPubMed
Floresco, S. B. (2015). The nucleus accumbens: An interface between cognition, emotion, and action. Annual Review of Psychology, 66(1), 2552. https://doi.org/10.1146/annurev-psych-010213-115159 CrossRefGoogle ScholarPubMed
França, T. F., & Pompeia, S. (2023). Reappraising the role of dopamine in adolescent risk-taking behavior. Neuroscience & Biobehavioral Reviews, 147, 105085. https://doi.org/10.1016/j.neubiorev.2023.105085 CrossRefGoogle ScholarPubMed
Friston, K. J., Holmes, A. P., Worsley, K. J., Poline, J. P., Frith, C. D., & Frackowiak, R. S. (1994). Statistical parametric maps in functional imaging: a general linear approach. Human brain Mapping, 2(4), 189210. https://doi.org/10.1002/hbm.460020402 CrossRefGoogle Scholar
Friston, K. J., Buechel, C., Fink, G. R., Morris, J., Rolls, E., & Dolan, R. J. (1997). Psychophysiological and modulatory interactions in neuroimaging. Neuroimage, 6(3), 218229. https://doi.org/10.1006/nimg.1997.0291 CrossRefGoogle ScholarPubMed
Fullana, M. A., Harrison, B. J., Soriano-Mas, C., Vervliet, B., Cardoner, N., Àvila-Parcet, A., & Radua, J. (2016). Neural signatures of human fear conditioning: An updated and extended meta-analysis of fMRI studies. Molecular Psychiatry, 21(4), 500508. https://doi.org/10.1038/mp.2015.88 CrossRefGoogle ScholarPubMed
Galván, A. (2010). Adolescent development of the reward system. Frontiers in Human Neuroscience, 4, 1018. https://doi.org/10.3389/neuro.09.006.2010 Google ScholarPubMed
Galvan, A., Hare, T. A., Parra, C. E., Penn, J., Voss, H., Glover, G., & Casey, B. J. (2006). Earlier development of the accumbens relative to orbitofrontal cortex might underlie risk-taking behavior in adolescents. Journal of Neuroscience, 26(25), 68856892. https://doi.org/10.1523/JNEUROSCI.1062-06.2006 CrossRefGoogle ScholarPubMed
Galván, A., & Peris, T. S. (2014). Neural correlates of risky decision making in anxious youth and healthy controls. Depression and Anxiety, 31(7), 591598. https://doi.org/10.1002/da.22276 CrossRefGoogle ScholarPubMed
Ganella, D. E., Drummond, K. D., Ganella, E. P., Whittle, S., & Kim, J. H. (2018). Extinction of conditioned fear in adolescents and adults: A human fMRI study. Frontiers in Human Neuroscience, 11, 647. https://doi.org/10.3389/fnhum.2017.00647 CrossRefGoogle ScholarPubMed
Garrison, J., Erdeniz, B., & Done, J. (2013). Prediction error in reinforcement learning: a meta-analysis of neuroimaging studies. Neuroscience & Biobehavioral Reviews, 37(7), 12971310. https://doi.org/10.1016/j.neubiorev.2013.03.023 CrossRefGoogle ScholarPubMed
Gee, D. G., Bath, K. G., Johnson, C. M., Meyer, H. C., Murty, V. P., van den Bos, W., & Hartley, C. A. (2018). Neurocognitive development of motivated behavior: Dynamic changes across childhood and adolescence. Journal of Neuroscience, 38(44), 94339445. https://doi.org/10.1523/JNEUROSCI.1674-18.2018 CrossRefGoogle ScholarPubMed
Gentry, R. N., Schuweiler, D. R., & Roesch, M. R. (2019). Dopamine signals related to appetitive and aversive events in paradigms that manipulate reward and avoidability. Brain Research, 1713, 8090. https://doi.org/10.1016/j.brainres.2018.10.008 CrossRefGoogle ScholarPubMed
Graham, B. M., & Milad, M. R. (2011). The study of fear extinction: Implications for anxiety disorders. American Journal of Psychiatry, 168(12), 12551265. https://doi.org/10.1176/appi.ajp.2011.11040557 CrossRefGoogle Scholar
Grasser, L. R., & Jovanovic, T. (2021). Safety learning during development: Implications for development of psychopathology. Behavioural Brain Research, 408, 113297. https://doi.org/10.1016/j.bbr.2021.113297 CrossRefGoogle ScholarPubMed
Greco, J. A., & Liberzon, I. (2016). Neuroimaging of fear-associated learning. Neuropsychopharmacology, 41(1), 320334. https://doi.org/10.1038/npp.2015.255 CrossRefGoogle ScholarPubMed
Haddad, A. D., Lissek, S., Pine, D. S., & Lau, J. Y. (2011). How do social fears in adolescence develop? Fear conditioning shapes attention orienting to social threat cues. Cognition & Emotion, 25(6), 11391147. https://doi.org/10.1080/02699931.2010.524193 CrossRefGoogle ScholarPubMed
Hartley, C. A., Gorun, A., Reddan, M. C., Ramirez, F., & Phelps, E. A. (2014). Stressor controllability modulates fear extinction in humans. Neurobiology of Learning and Memory, 113, 149156. https://doi.org/10.1016/j.nlm.2013.12.003 CrossRefGoogle ScholarPubMed
Heller, A. S., Cohen, A. O., Dreyfuss, M. F., & Casey, B. J. (2016). Changes in cortico-subcortical and subcortico-subcortical connectivity impact cognitive control to emotional cues across development. Social Cognitive and Affective Neuroscience, 11(12), 19101918. https://doi.org/10.1093/scan/nsw097 Google ScholarPubMed
Henson, R. N. A., Buechel, C., Josephs, O., & Friston, K. J. (1999). The slice-timing problem in event-related fMRI. NeuroImage, 9, 125.Google Scholar
Jovanovic, T., Nylocks, K. M., Gamwell, K. L., Smith, A., Davis, T. A., Norrholm, S. D., & Bradley, B. (2014). Development of fear acquisition and extinction in children: Effects of age and anxiety. Neurobiology of Learning and Memory, 113, 135142. https://doi.org/10.1016/j.nlm.2013.10.016 CrossRefGoogle ScholarPubMed
Kalisch, R., Gerlicher, A. M., & Duvarci, S. (2019). A dopaminergic basis for fear extinction. Trends in Cognitive Sciences, 23(4), 274277. https://doi.org/10.1016/j.tics.2019.01.013 CrossRefGoogle ScholarPubMed
Kessler, R. C., Angermeyer, M., Anthony, J. C., De Graaf, R. O. N., Demyttenaere, K., Gasquet, I., & Üstün, T. B. (2007). Lifetime prevalence and age-of-onset distributions of mental disorders in the world health organization’s world mental health survey initiative. World Psychiatry, 6(3), 168–76.Google ScholarPubMed
Kessler, R. C., Avenevoli, S., Costello, E. J., Georgiades, K., Green, J. G., Gruber, M. J., & Merikangas, K. R. (2012). Prevalence, persistence, and sociodemographic correlates of DSM-IV disorders in the national comorbidity survey replication adolescent supplement. Archives of General Psychiatry, 69(4), 372380. https://doi.org/10.1001/archgenpsychiatry.2011.160 Google ScholarPubMed
Klein, Z., Berger, S., Vervliet, B., & Shechner, T. (2021). Fear learning, avoidance, and generalization are more context-dependent for adults than adolescents. Behaviour Research and Therapy, 147, 103993. https://doi.org/10.1016/j.brat.2021.103993 CrossRefGoogle ScholarPubMed
Lago, T., Davis, A., Grillon, C., & Ernst, M. (2017). Striatum on the anxiety map: Small detours into adolescence. Brain Research, 1654, 177184. https://doi.org/10.1016/j.brainres.2016.06.006 CrossRefGoogle ScholarPubMed
Lau, J. Y., Britton, J. C., Nelson, E. E., Angold, A., Ernst, M., Goldwin, M., Grillon, C., Leibenluft, E., Lissek, S., Norcross, M., Shiffrin, N., & Pine, D. S. (2011). Distinct neural signatures of threat learning in adolescents and adults. Proceedings of the National Academy of Sciences of the United States of America, 108(11), 45004505. https://doi.org/10.1073/pnas.1005494108 CrossRefGoogle ScholarPubMed
LeDoux, J., & Daw, N. D. (2018). Surviving threats: Neural circuit and computational implications of a new taxonomy of defensive behaviour. Nature Reviews Neuroscience, 19(5), 269282. https://doi.org/10.1038/nrn.2018.22 CrossRefGoogle ScholarPubMed
Leng, L., Beckers, T., & Vervliet, B. (2022). No joy-why bother? Higher anhedonia relates to reduced pleasure from and motivation for threat avoidance. Behaviour Research and Therapy, 159, 104227. https://doi.org/10.1016/j.brat.2022.104227 CrossRefGoogle ScholarPubMed
Leng, L., Beckers, T., & Vervliet, B. (2023). What a relief! The pleasure of threat avoidance. Emotion. https://doi.org/10.1037/emo0001312 CrossRefGoogle ScholarPubMed
Leng, L., Beckers, T., & Vervliet, B. (2024). Anhedonia influences threat avoidance and relief: A conceptual replication. Journal of Mood & Anxiety Disorders, 5, 100050. https://doi.org/10.1016/j.xjmad.2024.100050 CrossRefGoogle Scholar
Levita, L., Hoskin, R., & Champi, S. (2012). Avoidance of harm and anxiety: A role for the nucleus accumbens. Neuroimage, 62(1), 189198. https://doi.org/10.1016/j.neuroimage.2012.04.059 CrossRefGoogle ScholarPubMed
Linton, S. R., & Levita, L. (2021). Potentiated perceptual neural responses to learned threat during Pavlovian fear acquisition and extinction in adolescents. Developmental Science, 24(5), e13107. https://doi.org/10.1111/desc.13107 CrossRefGoogle ScholarPubMed
Lissek, S., Pine, D. S., & Grillon, C. (2006). The strong situation: A potential impediment to studying the psychobiology and pharmacology of anxiety disorders. Biological Psychology, 72(3), 265270. https://doi.org/10.1016/j.biopsycho.2005.11.004 CrossRefGoogle ScholarPubMed
Lovibond, P. F., Mitchell, C. J., Minard, E., Brady, A., & Menzies, R. G. (2009). Safety behaviours preserve threat beliefs: Protection from extinction of human fear conditioning by an avoidance response. Behaviour Research and Therapy, 47(8), 716720. https://doi.org/10.1016/j.brat.2009.04.013 CrossRefGoogle ScholarPubMed
Master, S. L., Eckstein, M. K., Gotlieb, N., Dahl, R., Wilbrecht, L., & Collins, A. G. (2020). Disentangling the systems contributing to changes in learning during adolescence. Developmental Cognitive Neuroscience, 41, 100732. https://doi.org/10.1016/j.dcn.2019.100732 CrossRefGoogle Scholar
McCutcheon, J. E., Ebner, S. R., Loriaux, A. L., & Roitman, M. F. (2012). Encoding of aversion by dopamine and the nucleus accumbens. Frontiers in Neuroscience, 6, 137. https://doi.org/10.3389/fnins.2012.00137 CrossRefGoogle ScholarPubMed
McLaughlin, K. A., DeCross, S. N., Jovanovic, T., & Tottenham, N. (2019). Mechanisms linking childhood adversity with psychopathology: Learning as an intervention target. Behaviour Research and Therapy, 118, 101109. https://doi.org/10.1016/j.brat.2019.04.008 CrossRefGoogle ScholarPubMed
Merikangas, K. R., He, J.-P., Burstein, M., Swanson, S. A., Avenevoli, S., Cui, L., Benjet, C., Georgiades, K., & Swendsen, J. (2010). Lifetime prevalence of mental disorders in US adolescents: Results from the national comorbidity survey replication-adolescent supplement (NCS-A). Journal of the American Academy of Child & Adolescent Psychiatry, 49(10), 980989. https://doi.org/10.1016/j.jaac.2010.05.017 CrossRefGoogle ScholarPubMed
Milad, M. R., Rosenbaum, B. L., & Simon, N. M. (2014). Neuroscience of fear extinction: implications for assessment and treatment of fear-based and anxiety related disorders. Behaviour Research and Therapy, 62, 1723. https://doi.org/10.1016/j.brat.2014.08.006 CrossRefGoogle ScholarPubMed
Mobbs, D., Headley, D. B., Ding, W., & Dayan, P. (2020). Space, time, and fear: Survival computations along defensive circuits. Trends in Cognitive Sciences, 24(3), 228241. https://doi.org/10.1016/j.tics.2019.12.016 CrossRefGoogle ScholarPubMed
Mobbs, D., Marchant, J. L., Hassabis, D., Seymour, B., Tan, G., Gray, M., Petrovic, P., Dolan, R. J., & Frith, C. D. (2009). From threat to fear: The neural organization of defensive fear systems in humans. Journal of Neuroscience, 29(39), 1223612243. https://doi.org/10.1523/JNEUROSCI.2378-09.2009 CrossRefGoogle ScholarPubMed
Mobbs, D., Petrovic, P., Marchant, J. L., Hassabis, D., Weiskopf, N., Seymour, B., Dolan, R. J., & Frith, C. D. (2007). When fear is near: Threat imminence elicits prefrontal-periaqueductal gray shifts in humans. Science, 317(5841), 10791083. https://doi.org/10.1126/science.1144298 CrossRefGoogle ScholarPubMed
Morriss, J., Christakou, A., & Van Reekum, C. M. (2019). Multimodal evidence for delayed threat extinction learning in adolescence and young adulthood. Scientific Reports, 9(1), 7748. https://doi.org/10.1038/s41598-019-44150-1 CrossRefGoogle ScholarPubMed
Murty, D. V., Song, S., Surampudi, S. G., & Pessoa, L. (2023). Threat and reward imminence processing in the human brain. Journal of Neuroscience, 43(16), 29732987. https://doi.org/10.1523/JNEUROSCI.1778-22.2023 CrossRefGoogle ScholarPubMed
Navratilova, E., Xie, J. Y., Okun, A., Qu, C., Eyde, N., Ci, S., Ossipov, M. H., King, T., Fields, H. L., & Porreca, F. (2012). Pain relief produces negative reinforcement through activation of mesolimbic reward-valuation circuitry. Proceedings of the National Academy of Sciences, 109(50), 2070920713. https://doi.org/10.1073/pnas.1214605109 CrossRefGoogle ScholarPubMed
Neumann, D. L., Waters, A. M., Westbury, H. R., & Henry, J. (2008). The use of an unpleasant sound unconditional stimulus in an aversive conditioning procedure with 8-to 11-year-old children. Biological Psychology, 79(3), 337342. https://doi.org/10.1016/j.biopsycho.2008.08.005 CrossRefGoogle Scholar
Nicola, S. M. (2007). The nucleus accumbens as part of a basal ganglia action selection circuit. Psychopharmacology, 191(3), 521550. https://doi.org/10.1007/s00213-006-0510-4 CrossRefGoogle ScholarPubMed
Odriozola, P., & Gee, D. G. (2021). Learning about safety: Conditioned inhibition as a novel approach to fear reduction targeting the developing brain. American Journal of Psychiatry, 178(2), 136155. https://doi.org/10.1176/appi.ajp.2020.20020232 CrossRefGoogle ScholarPubMed
Oleson, E. B., Gentry, R. N., Chioma, V. C., & Cheer, J. F. (2012). Subsecond dopamine release in the nucleus accumbens predicts conditioned punishment and its successful avoidance. Journal of Neuroscience, 32(42), 1480414808. https://doi.org/10.1523/JNEUROSCI.3087-12.2012 CrossRefGoogle ScholarPubMed
Palminteri, S., Kilford, E. J., Coricelli, G., & Blakemore, S. J. (2016). The computational development of reinforcement learning during adolescence. PLoS Computational Biology, 12(6), e1004953. https://doi.org/10.1371/journal.pcbi.1004953 CrossRefGoogle ScholarPubMed
Pattwell, S. S., Duhoux, Séphanie, Hartley, C. A., Johnson, D. C., Jing, D., Elliott, M. D., Ruberry, E. J., Powers, A., Mehta, N., Yang, R. R., Soliman, F., Glatt, C. E., Casey, B. J., Ninan, I., & Lee, F. S. (2012). Altered fear learning across development in both mouse and human. Proceedings of the National Academy of Sciences, 109(40), 1631816323. https://doi.org/10.1073/pnas.1206834109 CrossRefGoogle ScholarPubMed
Paus, T., Keshavan, M., & Giedd, J. N. (2008). Why do many psychiatric disorders emerge during adolescence? Nature Reviews Neuroscience, 9(12), 947957. https://doi.org/10.1038/nrn2513 CrossRefGoogle ScholarPubMed
Peng, Y., Knotts, J. D., Young, K. S., Bookheimer, S. Y., Nusslock, R., Zinbarg, R. E., Kelley, N. J., Echiverri-Cohen, A. M., & Craske, M. G. (2023). Threat neurocircuitry predicts the development of anxiety and depression symptoms in a longitudinal study. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging, 8(1), 102110. https://doi.org/10.1016/j.bpsc.2021.12.013 Google ScholarPubMed
Peris, T. S., & Galván, A. (2021). Brain and behavior correlates of risk taking in pediatric anxiety disorders. Biological Psychiatry, 89(7), 707715. https://doi.org/10.1016/j.biopsych.2020.11.003 CrossRefGoogle ScholarPubMed
Pittig, A., Treanor, M., LeBeau, R. T., & Craske, M. G. (2018). The role of associative fear and avoidance learning in anxiety disorders: Gaps and directions for future research. Neuroscience & Biobehavioral Reviews, 88, 117140. https://doi.org/10.1016/j.neubiorev.2018.03.015 CrossRefGoogle ScholarPubMed
Pittig, A., Wong, A. H., Glück, V. M., & Boschet, J. M. (2020). Avoidance and its bi-directional relationship with conditioned fear: Mechanisms, moderators, and clinical implications. Behaviour Research and Therapy, 126, 103550. https://doi.org/10.1016/j.brat.2020.103550 CrossRefGoogle ScholarPubMed
Powers, A., & Casey, B. J. (2015). The adolescent brain and the emergence and peak of psychopathology. Journal of Infant, Child, and Adolescent Psychotherapy, 14(1), 315. https://doi.org/10.1080/15289168.2015.1004889 CrossRefGoogle Scholar
Ramirez, F., Moscarello, J. M., LeDoux, J. E., & Sears, R. M. (2015). Active avoidance requires a serial basal amygdala to nucleus accumbens shell circuit. Journal of Neuroscience, 35(8), 34703477. https://doi.org/10.1523/JNEUROSCI.1331-14.2015 CrossRefGoogle ScholarPubMed
Rapee, R. M., Oar, E. L., Johnco, C. J., Forbes, M. K., Fardouly, J., Magson, N. R., & Richardson, C. E. (2019). Adolescent development and risk for the onset of social-emotional disorders: A review and conceptual model. Behaviour Research and Therapy, 123, 103501. https://doi.org/10.1016/j.brat.2019.103501 CrossRefGoogle ScholarPubMed
Ray, M. H., Moaddab, M., & McDannald, M. A. (2022). Threat and bidirectional valence signaling in the nucleus accumbens core. Journal of Neuroscience, 42(5), 817833. https://doi.org/10.1523/JNEUROSCI.1107-21.2021 CrossRefGoogle ScholarPubMed
Ray, M. H., Russ, A. N., Walker, R. A., & McDannald, M. A. (2020). The nucleus accumbens core is necessary to scale fear to degree of threat. Journal of Neuroscience, 40(24), 47504760. https://doi.org/10.1523/JNEUROSCI.0299-20.2020 CrossRefGoogle ScholarPubMed
Reinhard, J., Slyschak, A., Schiele, M. A., Andreatta, M., Kneer, K., Reif, A., Domschke, K., Gamer, M., Pauli, P., Deckert, J. C., & Romanos, M. (2022). Fear conditioning and stimulus generalization in association with age in children and adolescents. European Child & Adolescent Psychiatry, 31(10), 15811590. https://doi.org/10.1007/s00787-021-01797-4 CrossRefGoogle ScholarPubMed
Richards, J. M., Plate, R. C., & Ernst, M. (2012). Neural systems underlying motivated behavior in adolescence: Implications for preventive medicine. Preventive Medicine, 55, S7S16. https://doi.org/10.1016/j.ypmed.2011.11.016 CrossRefGoogle ScholarPubMed
Romer, D., Reyna, V. F., & Satterthwaite, T. D. (2017). Beyond stereotypes of adolescent risk taking: Placing the adolescent brain in developmental context. Developmental Cognitive Neuroscience, 27, 1934. https://doi.org/10.1016/j.dcn.2017.07.007 CrossRefGoogle ScholarPubMed
Rosenberg, B. M., Barnes-Horowitz, N. M., Zbozinek, T. D., & Craske, M. G. (2024). Reward processes in extinction learning and applications to exposure therapy. Journal of Anxiety Disorders, 106, 102911. https://doi.org/10.1016/j.janxdis.2024.102911 CrossRefGoogle Scholar
Rosenberg, B. M., Taschereau-Dumouchel, V., Lau, H., Young, K. S., Nusslock, R., Zinbarg, R. E., & Craske, M. G. (2023). A multivoxel pattern analysis of anhedonia during fear extinction: Implications for safety learning. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging, 8(4), 417425. https://doi.org/10.1016/j.bpsc.2021.12.008 Google ScholarPubMed
Rosenberg, B. M., Young, K. S., Nusslock, R., Zinbarg, R. E., & Craske, M. G. (2024). Anhedonia is associated with overgeneralization of conditioned fear during late adolescence and early adulthood. Journal of Anxiety Disorders, 10, 2880. https://doi.org/10.1016/j.janxdis.2024.102880 Google Scholar
Salinas-Hernández, X. I., & Duvarci, S. (2021). Dopamine in fear extinction. Frontiers in Synaptic Neuroscience, 13, 10. https://doi.org/10.3389/fnsyn.2021.635879 CrossRefGoogle ScholarPubMed
Schreuders, E., Braams, B. R., Blankenstein, N. E., Peper, J. S., Güroğlu, B., & Crone, E. A. (2018). Contributions of reward sensitivity to ventral striatum activity across adolescence and early adulthood. Child Development, 89(3), 797810. https://doi.org/10.1111/cdev.13056 CrossRefGoogle ScholarPubMed
Schultz, W. (2013). Updating dopamine reward signals. Current Opinion in Neurobiology, 23(2), 229238. https://doi.org/10.1016/j.conb.2012.11.012 CrossRefGoogle ScholarPubMed
Sehlmeyer, C., Schöning, S., Zwitserlood, P., Pfleiderer, B., Kircher, T., Arolt, V., & Konrad, C. (2009). Human fear conditioning and extinction in neuroimaging: a systematic review. PloS one, 4(6), e5865. https://doi.org/10.1371/journal.pone.0005865 CrossRefGoogle ScholarPubMed
Shechner, T., Hong, M., Britton, J. C., Pine, D. S., & Fox, N. A. (2014). Fear conditioning and extinction across development: Evidence from human studies and animal models. Biological Psychology, 100, 112. https://doi.org/10.1016/j.biopsycho.2014.04.001 CrossRefGoogle ScholarPubMed
Silk, J. S., Davis, S., McMakin, D. L., Dahl, R. E., & Forbes, E. E. (2012). Why do anxious children become depressed teenagers? The role of social evaluative threat and reward processing. Psychological Medicine, 42(10), 20952107. https://doi.org/10.1017/S0033291712000207 CrossRefGoogle ScholarPubMed
Silverman, M. H., Jedd, K., & Luciana, M. (2015). Neural networks involved in adolescent reward processing: An activation likelihood estimation meta-analysis of functional neuroimaging studies. NeuroImage, 122, 427439. https://doi.org/10.1016/j.neuroimage.2015.07.083 CrossRefGoogle ScholarPubMed
Silvers, J. A., Lumian, D. S., Gabard-Durnam, L., Gee, D. G., Goff, B., Fareri, D. S., Caldera, C., Flannery, J., Telzer, E. H., Humphreys, K. L., & Tottenham, N. (2016). Previous institutionalization is followed by broader amygdala-hippocampal–PFC network connectivity during aversive learning in human development. Journal of Neuroscience, 36(24), 64206430. https://doi.org/10.1523/JNEUROSCI.0038-16.2016 CrossRefGoogle ScholarPubMed
Somerville, L. H., Jones, R. M., & Casey, B. J. (2010). A time of change: Behavioral and neural correlates of adolescent sensitivity to appetitive and aversive environmental cues. Brain and Cognition, 72(1), 124133. https://doi.org/10.1016/j.bandc.2009.07.003 CrossRefGoogle ScholarPubMed
Spielberg, J. M., Olino, T. M., Forbes, E. E., & Dahl, R. E. (2014). Exciting fear in adolescence: Does pubertal development alter threat processing? Developmental Cognitive Neuroscience, 8, 8695. https://doi.org/10.1016/j.dcn.2014.01.004 CrossRefGoogle ScholarPubMed
Stelly, C. E., Haug, G. C., Fonzi, K. M., Garcia, M. A., Tritley, S. C., Magnon, A. P., Ramos, M. A. P., & Wanat, M. J. (2019). Pattern of dopamine signaling during aversive events predicts active avoidance learning. Proceedings of the National Academy of Sciences, 116(27), 1364113650. https://doi.org/10.1073/pnas.1904249116 CrossRefGoogle ScholarPubMed
Stenson, A. F., France, J. M., & Jovanovic, T. (2023). Getting better with age? A review of psychophysiological studies of fear extinction learning across development. Current Topics in Behavioral Neurosciences, 64, 213236. https://doi.org/10.1007/7854_2023_441 CrossRefGoogle Scholar
Tashjian, S. M., Zbozinek, T. D., & Mobbs, D. (2021). A decision architecture for safety computations. Trends in Cognitive Sciences, 25(5), 342354. https://doi.org/10.1016/j.tics.2021.01.013 CrossRefGoogle ScholarPubMed
Topel, S., Ma, I., Sleutels, J., van Steenbergen, H., de Bruijn, E. R., & van Duijvenvoorde, A. C. (2023). Expecting the unexpected: A review of learning under uncertainty across development. Cognitive, Affective, & Behavioral Neuroscience, 1-21(3), 718738. https://doi.org/10.3758/s13415-023-01098-0 CrossRefGoogle Scholar
Treanor, M., Rosenberg, B. M., & Craske, M. G. (2021). Pavlovian learning processes in pediatric anxiety disorders: A critical review. Biological Psychiatry, 89(7), 690696. https://doi.org/10.1016/j.biopsych.2020.09.008 CrossRefGoogle ScholarPubMed
Urošević, S., Collins, P., Muetzel, R., Lim, K., & Luciana, M. (2012). Longitudinal changes in behavioral approach system sensitivity and brain structures involved in reward processing during adolescence. Developmental Psychology, 48(5), 14881500. https://psycnet.apa.org/doi/10.1037/a0027502 CrossRefGoogle ScholarPubMed
Van Duijvenvoorde, A. C., Peters, S., Braams, B. R., & Crone, E. A. (2016). What motivates adolescents? Neural responses to rewards and their influence on adolescents’ risk taking, learning, and cognitive control. Neuroscience & Biobehavioral Reviews, 70, 135147. https://doi.org/10.1016/j.neubiorev.2016.06.037 CrossRefGoogle ScholarPubMed
Verharen, J. P., Zhu, Y., & Lammel, S. (2020). Aversion hot spots in the dopamine system. Current Opinion in Neurobiology, 64, 4652. https://doi.org/10.1016/j.conb.2020.02.002 CrossRefGoogle ScholarPubMed
Vervliet, B., Lange, I., & Milad, M. R. (2017). Temporal dynamics of relief in avoidance conditioning and fear extinction: Experimental validation and clinical relevance. Behaviour Research and Therapy, 96, 6678. https://doi.org/10.1016/j.brat.2017.04.011 CrossRefGoogle ScholarPubMed
Walker, D. M., Bell, M. R., Flores, C., Gulley, J. M., Willing, J., & Paul, M. J. (2017). Adolescence and reward: Making sense of neural and behavioral changes amid the chaos. Journal of Neuroscience, 37(45), 1085510866. https://doi.org/10.1523/JNEUROSCI.1834-17.2017 CrossRefGoogle ScholarPubMed
Waters, A. M., & Craske, M. G. (2016). Towards a cognitive-learning formulation of youth anxiety: A narrative review of theory and evidence and implications for treatment. Clinical Psychology Review, 50, 5066. https://doi.org/10.1016/j.cpr.2016.09.008 CrossRefGoogle ScholarPubMed
Waters, A. M., Theresiana, C., Neumann, D. L., & Craske, M. G. (2017). Developmental differences in aversive conditioning, extinction, and reinstatement: A study with children, adolescents, and adults. Journal of Experimental Child Psychology, 159, 263278. https://doi.org/10.1016/j.jecp.2017.02.012 CrossRefGoogle ScholarPubMed
Whitfield-Gabrieli, S., & Nieto-Castanon, A. (2012). Conn: a functional connectivity toolbox for correlated and anticorrelated brain networks. Brain Connectivity, 2(3), 125141. https://doi.org/10.1089/brain.2012.0073 CrossRefGoogle Scholar
Willems, A. L., Van Oudenhove, L., & Vervliet, B. (2023). Omissions of threat trigger subjective relief and reward prediction error-like signaling in the human reward system. eLife, 12, RP91400. https://doi.org/10.7554/eLife.91400.1 Google Scholar
Wong, A. H., Wirth, F. M., & Pittig, A. (2022). Avoidance of learnt fear: Models, potential mechanisms, and future directions. Behaviour Research and Therapy, 151, 104056. https://doi.org/10.1016/j.brat.2022.104056 CrossRefGoogle ScholarPubMed
Xie, S., Zhang, X., Cheng, W., & Yang, Z. (2021). Adolescent anxiety disorders and the developing brain: Comparing neuroimaging findings in adolescents and adults. General Psychiatry, 34(4), e100411. https://doi.org/10.1136gpsych-2020-100411 CrossRefGoogle ScholarPubMed
Young, K. S., Bookheimer, S. Y., Nusslock, R., Zinbarg, R. E., Damme, K. S., Chat, I. K. Y., ... & Craske, M. G. (2021). Dysregulation of threat neurocircuitry during fear extinction: the role of anhedonia. Neuropsychopharmacology, 46(9), 16501657. https://doi.org/10.1038/s41386-021-01003-8 CrossRefGoogle ScholarPubMed
Zimmermann, K. S., Richardson, R., & Baker, K. D. (2019). Maturational changes in prefrontal and amygdala circuits in adolescence: Implications for understanding fear inhibition during a vulnerable period of development. Brain Sciences, 9(3), 65. https://doi.org/10.3390/brainsci9030065 CrossRefGoogle ScholarPubMed
Figure 0

Table 1. Participant demographics in the study

Figure 1

Figure 1. Visual representation of the aversive learning paradigm including three types of trial blocks: CS+r, CS+nr, and CS−. Participants completed eight total blocks (three CS+r, three CS+nr, two CS−).

Figure 2

Figure 2. NAcc-amygdala functional connectivity differentiated stimuli by the end of the task, such that connectivity tended to increase for the CS+r but decrease for the CS+nr and the CS−.

Figure 3

Table 2. Covariate results and significant pairwise contrasts among stimulus/block pairs in the NAcc-amygdala connectivity analyses

Figure 4

Figure 3. Greater internalizing symptoms were associated with greater NAcc-amygdala functional connectivity for the CS−, but not CS+r or CS+nr, during the task.

Figure 5

Figure 4. NAcc-amygdala functional connectivity during the task among participants younger (left) or older (right) than the median age in the sample (14.62 years). Age is shown categorically for interpretability. All analyses were conducted on continuous measures of age.

Supplementary material: File

Rosenberg et al. supplementary material

Rosenberg et al. supplementary material
Download Rosenberg et al. supplementary material(File)
File 628.5 KB