Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Mar 18.
Published in final edited form as: Dev Psychobiol. 2023 May;65(4):e22386. doi: 10.1002/dev.22386

Neurophysiology of predictable unpleasant event processing in pre-adolescents and early adolescents, part II: Reflex and event-related potential markers of defensive reactivity and peripheral attention modulation

Christopher T Sege 1, Lisa M McTeague 1, Molly Kegley 1,2, Curtisha Shacklewood 1, Colleen A Halliday 1, Casey D Calhoun 3, Jane E Joseph 1, Zachary W Adams 4, Greg Hajcak 5, Carla Kmett Danielson 1
PMCID: PMC10948024  NIHMSID: NIHMS1894388  PMID: 37073586

Abstract

The ability to anticipate and process predictable unpleasant events, while also regulating emotional reactivity, is an adaptive skill. The current article and a companion in this issue test for potential changes in predictable event processing across the childhood-to-adolescence transition, a key developmental period for biological systems that support cognitive/ emotional abilities. While the companion article focuses on neurophysiology of predictable event processing itself, the present article examines peripheral emotional response regulation and attention modulation that coincides with event processing. A total of 315 third-, sixth-, or ninth-grade individuals saw 5-s cues predicting “scary,” “every day,” or uncertain pictures, and here, blink reflexes and brain event-related potentials (ERPs) elicited by peripheral noise probes are analyzed. During the cue, blink reflexes and probe ERP (P200) amplitudes were increased when the cue predicted scary, compared to everyday, content. After picture onset, reflex enhancement by scary content then disappeared for predictable images, whereas ERP modulation was similar regardless of predictability. Patterns are similar to those in adults and suggest (1) sustained defensive response priming and enhancement of peripheral attention during aversive anticipation, and (2) an ability, even in pre-adolescents, to downregulate defensive priming while maintaining attentional modulation once an awaited predictable aversive event occurs.

Keywords: adolescent, attention, electrophysiology, emotion, stress

1 |. INTRODUCTION

The transition from childhood to adolescence is a critical period in which to study cognitive/emotional development for several reasons—including that marked shifts in psychological functioning in this period (see Costello et al., 2003) coincide with commensurate changes in the biological systems critical to that functioning (e.g., prefrontal cortex systems; Caballero et al., 2016). To contribute to the literature on cognitive/emotional development in the adolescent transition, this article and a companion in this issue test for potential change in one adaptive ability—the ability to process upcoming unpleasant events—from ages 8–15 (a range that, for most, spans a pubertal process that critically contributes to prefrontal cortex development; see Howard, 2019; Stephens & Marcin, 2022). While a companion article examines neural concomitants of predictable event processing itself, the present article tests for potential age-related change in how emotional reactivity and peripheral attention are modulated in concert with that processing—and in so doing, this paper tests if the ways in which automatic emotion regulation systems operate in predictably unpleasant contexts matures across the adolescent transition.

An ability to modulate/regulate emotional reactivity in contexts of upcoming unpleasant events is adaptive inasmuch as it supports optimal engagement with those events by preventing interference from potentially disruptive emotional reactions (see Gross & Thompson, 2007). Speaking to relevance of studying regulatory processes in 8to 15-year-olds, a number of studies find evidence for key changes in higher order emotion regulation during this age range—for example such that adolescents, compared to pre-adolescents, show generally increased reliance on emotion regulation in naturalistic settings (Silvers, 2022), marked changes in the types of regulation strategies that are selected and brain activity during execution of those strategies (see Helion et al., 2019), and evidence of greater stress response regulation in dynamic challenge tasks like the Trier Stress Test (Seddon et al., 2020). Building from this work, the present paper examines possible age-related change in automatic/implicit emotional modulation specifically in the context of predictable aversive events—using a startle elicitation method that can probe modulation of emotional reactivity as it co-occurs with predictable event processing itself.

1.1 |. Probing affective reactivity in emotional contexts: the blink startle reflex

A startle elicitation method can be used to study how emotional responding is automatically modulated/regulated across different contexts by probing implicit changes in motive reflex reactivity while a primary (i.e., foreground) task is being performed. In the startle method, brief probes (often acoustic bursts) are presented as to-be-ignored stimuli during a task, and facial electromyography (EMG) is used to measure the blink component of an elicited startle reflex that is inherently protective (i.e., defensive) and thus primed (potentiated) under aversive conditions (sub-cortical limbic structures, especially amygdala, mediate this priming; see Lang & Davis, 2006). Though work is needed on how youth modulate startle in predictable unpleasant exposure contexts, such work can then be informed by research with adults that has probed reflex reactivity in predictable event tasks wherein initial cues reliably predict emotional content of subsequent images. In these studies, reflexes elicited during the cuing phase show sustained reflex priming in unpleasant conditions—such that blinks elicited through-out unpleasant-predicting cues are potentiated compared to blinks elicited during neutral-predicting cues (Dichter et al., 2002; Nitschke et al., 2002; Sabatinelli et al., 2001; Sege et al., 2014). In addition, studies that continue probing reflex reactivity after picture onset show that priming during unpleasant events once they occur is then reduced by reliable cuing—that is, potentiation typically seen during un-cued unpleasant, relative to neutral, pictures (see Bradley et al., 2001) is reduced (Sege et al., 2015) or eliminated (Sege et al., 2014) when pictures have been reliably cued. Together, findings for adults thus indicate sustained activation of a (subcortical/limbic) defensive system throughout anticipation of predictable aversive events, and then downregulation of such activation after an event occurs when content of that event is predictable.

In addition to studies of startle reflex modulation during cued unpleasant event processing in adults, tests of related conditioning/generalization and general predictable/unpredictable threat contexts with younger individuals could also inform the study of emotional response modulation during predictable event processing in this age range. In conditioning/generalization work in a similar age range as in the current study (Glenn et al., 2012), some evidence for age-related changes arises such that pre-adolescent children show less learning of blink potentiation during conditioned cues and also atypical generalization to similar cues as compare to older youth—overall suggesting changes across the transition toward adolescence in ability to learn conditioned associations. Conversely, in an oft-used no threat/predictable threat/unpredictable threat (NPU) task—wherein startle is probed throughout time blocks wherein unpleasant stimuli will either certainly not occur (no threat), occur sporadically but be preceded by a countdown (predictable threat), or occur unpredictably with no countdown (unpredictable threat)—pre-adolescents and early adolescents each show, similarly to adults, reflex potentiation in contexts where exposure is predictable and further potentiation when exposure is unpredictable (Nelson & Hajcak, 2017). Taken together, conditioning and NPU results then suggest that an ability to learn predictive associations may develop during the adolescent transition, whereas ability to modulate defensive reactivity in unconditioned threat contexts does not—such that the latter is present even prior to adolescence.

While it could be that an ability to modulate defensive reactivity in unconditioned threat contexts matures prior to adolescence, existing data also cannot rule out an alternative that developmental findings in a conditioning task do not reflect changes in learning ability, but rather changes in the ability to modulate startle during cued anticipation periods (as opposed to in more diffuse periods of threat as modeled in NPU). Moreover, existing child and adolescent research also does not test for potential changes in how startle is modulated after onset of a predicted unpleasant event across the adolescent transition. Related to this, research does show that startle priming during unpredictable unpleasant (compared to emotionally neutral) picture viewing is similar in youth as in adults (Dhamija et al., 2017), but this work has not tested the impact of content predictability on these modulation patterns. Given this gap, current work also probes startle reactivity after onset of cued pictures to test if the transition to adolescence impacts downregulation of defensive activation once predictable unpleasant events occur.

1.2 |. Peripheral attention in emotional contexts: the probe-locked event-related potential

In addition to examining defensive reflex modulation, this research also studies modulation of peripheral attention during predictable event processing by measuring electroencephalography and deriving an event-related brain potential (ERP) index of cortical processing of the probe (see Schupp et al., 1997). Startle probes evoke an ERP that is very typical for punctate sensory stimuli, and which is characterized by N100 (reflecting early stimulus detection; Näätänen & Picton, 1986), P200 (reflecting secondary perceptual processing; Tremblay et al., 2001), and P300 (thought to reflect preferential detection of salient stimuli; see Polich, 2007) peaks. When studying emotional modulation of probe processing, effects are then often most striking for P300—in particular such that (in adults) P300 amplitudes are reduced for probes presented in unpleasant or pleasant, compared to neutral, contexts to indicate withdrawal of selective processing resources from the probe in order to allocate them toward more task-central emotional stimuli (Schupp et al., 1997).

Just as interestingly regarding probe P300, reduced amplitudes also coincide with different blink reflex modulation patterns across emotional contexts (e.g., inhibited blink reactivity in pleasant contexts but enhanced blinks in unpleasant contexts; Cuthbert et al., 1998; Schupp et al., 1997). Together, probe P300 and blink findings then suggest a dissociation of probe-related cortical processing from probe-related defensive reactivity—such that, in unpleasant contexts, attention can be directed away from the task-irrelevant probe while defensive reactivity to that same probe is also enhanced (Cuthbert et al., 1998). By studying probe ERPs in younger individuals, current work can then test if developmental stage impacts attentional monitoring of peripheral stimuli in addition to, or instead of, defensive reactivity; a potentially critical aim inasmuch as developments in selective attention during the adolescent transition have also been found to predict adaptive emotion regulation development (Schweizer et al., 2020). And, while work with NPU (Nelson & Hajcak, 2017) and unpredictable picture viewing (Latham, Cook, Simmons, Byrne et al., 2017) tasks shows similar probe ERP modulation (reduced P300 amplitudes) as in adults, probe ERP has not been studied during discrete content-predictive cues—and so it is not known if peripheral attention modulation during aversive event anticipation and processing changes across the adolescent transition.

1.3 |. The present study

To summarize, this article and a companion in the current issue examine potential changes in predictable unpleasant processing from pre-adolescence to early adolescence—and this article focuses on how defense reactivity and peripheral attention are modulated concurrently with processing of the predictable events themselves. To model predictable event contexts, a task was deployed in which each trial shows an initial cue whose color indicates if a subsequent picture will depict an unpleasant scene, emotionally neutral scene, or unpredictable content (i.e., cue followed equally often by an unpleasant or neutral scene). In addition to task stimuli, task-irrelevant startling noises are also presented during each cue and picture to probe modulation of defensive reactivity and peripheral attention as a function of the type of stimulus being processed for the foreground task. To test for potential changes in modulation at different stages of the transition from childhood to early adolescence, third-grade (i.e., largely 8- to 9-year-old, pre-adolescent), sixth-grade, and ninth-grade (largely 14- to 15-year-old, early adolescent) samples were recruited to do the cued image task (data in both papers from the same third-, sixth-, and ninth-grade samples).

Across grade groups in this study, possible change in emotional modulation was captured by measuring blink reflex reactivity to the peripheral startling probes. Regarding predictions, while no research has examined reflex reactivity during cue-elicited anticipation of unpleasant images in youth, lack of age effects on reflex priming in a different general threat (NPU) context (Nelson & Hajcak, 2017) suggests that reflex priming during cue-elicited discrete anticipation of upcoming aversive events might also be apparent even in pre-adolescent individuals. For the perhaps more interesting picture interval, meanwhile, findings of emotion regulation development across the adolescent transition (see Silvers, 2022) suggest a hypothesis that less developed regulation capacities might also be reflected in a weaker effect of content predictability on aversive reflex priming—that is, reflexes are still enhanced during predictable unpleasant, relative to neutral, pictures—in younger individuals. Moreover, given the timing of biological changes in the adolescent transition, it might also be that an impact of predictability on reflex priming is particularly attenuated in third-grade (i.e., largely pre-adolescent) individuals, whereas sixth graders (a transitional group with respect to developmental neurobiological changes; Marceau et al., 2011; Peper et al., 2009) show a marginal predictability effect and ninth graders (i.e., largely early adolescent) show a reliable effect.

Next, to test effects on how peripheral attention is modulated during predictable event processing, ERPs elicited by the startling probes were also measured. Given that these ERPs have not been studied in discrete cued anticipation contexts, predictions were based on findings from the NPU task (Nelson & Hajcak, 2017) and were that, in a cuing context as in the general threat context modeled by NPU, peripheral attention would be attenuated while processing unpleasant, as compared to emotionally neutral, stimuli even for pre-adolescent (third grade) individuals. Finally, because current data were collected in a large, demographically diverse sample, analyses also include preliminary tests of potential differences as a function of key biological variables of sex and race/ethnicity. Importantly, it should be emphasized that these analyses are a highly preliminary exploration that might give a first indication of potential differences that might warrant further study. At the same time, it must also be acknowledged that much more sophisticated analyses with proper modeling of sociocultural constructs are needed to fully understand any differences that do arise.

2 |. MATERIALS AND METHODS

2.1 |. Participants

Children and adolescents were recruited from the community as part of a large NIMH-funded longitudinal project (the “CHARM” study) that examines how stressful life events impact development of threat-related negative valence systems during a childhood-to-adolescence transition. The CHARM project was designed as an accelerated longitudinal study such that third-, sixth-, and ninth-grade cohorts were recruited and each followed for 2 years. This paper reports on time 1 data for an anticipation task administered as part of the larger task battery. All procedures were IRB approved and written assent/consent was obtained from participants and their legal guardians.

The study sample includes 100 third-grade, 111 sixth-grade, and 104 ninth-grade individuals who completed study procedures. An additional 14 third graders, 15 sixth graders, and 11 ninth graders chose not to complete the study task. Of the 315 participants who did complete this study, some data were not available for 17 third graders, 18 sixth graders, and 19 ninth graders due to recording problems or excessive movement. Specifically, 12 third-grade, 14 sixth-grade, and 12 ninth-grade participants had usable ERP but not blink startle data, while 2 third-grade, 3 sixth-grade, and 4 ninth-grade participants had usable blink startle but not EEG data. Main analyses report findings for the 83 third graders, 93 sixth graders, and 84 ninth graders with complete data for both measures.1

Examination of demographic characteristics related to data missingness also indicated that data for one or both measures were more likely to be missing for participants who did not identify as White/European American (22.4% with missing data; 33/147 participants) than for participants who did identify as White/European American (11.3% with missing data; 19/168 participants). Data missingness did not differ as a function of other demographic variables (biological sex, grade cohort). Potential implications of differences in data missingness as a function of race identification are considered in the discussion.

2.2 |. Design and materials

Participants completed an anticipation task in which one cue reliably predicted “something scary,” another predicted “something you might see every day,” and a third was followed equally often by scary or everyday pictures (Figure 1). All task stimuli were shown centrally (576×768 pixels) on a computer 90 cm away, such that they subtended ~5° visual angle from fixation. Every trial began with 5 s presentation of a red, green, or yellow rectangle that indicated if a subsequent image would be scary, quotidian, or not predictable (color meaning counterbalanced across participants). After the cue, a picture appeared immediately and was shown for 3 s. Finally, an inter-trial interval then showed a fixation cross for 4–6 s.

FIGURE 1.

FIGURE 1

Open in a new tab

Study task design.

To serve as anticipated content in the task, 36 color pictures from the International Affective Picture System (Lang et al., 2008) were selected such that scary and everyday categories each contained 18 images. Scary images depicted snakes, spiders, or threat (e.g., pointed guns), while everyday images depicted mundane objects (e.g., a rolling pin).2 Based on normative ratings (Lang et al., 2008), sets were constructed such that “scary” images were significantly more unpleasant and intense/ arousing than “everyday” images.3 Images were divided so that 12 scary and 12 everyday images followed a predictive cue, while six scary and six everyday images followed the nonpredictive cue. Though the nature of the task unavoidably resulted in presentation of a greater number of content nonpredictive cues or of content-predictable pictures, for statistical reasons (e.g., avoiding unequal regression to the mean across conditions) we also viewed it as important to equalize trial numbers in analyses. Toward this end, task orders were constructed a priori such that images with predictable content were divisible into subsets of 12 images (six scary, six everyday) to be used in analyses and 12 images (six scary, six everyday) that were excluded from analyses. Images to be used in analyses were presented at similar points in the experiment as images whose content was unpredictable, and unpredictable and analyzed predictable image subsets were also matched for content (e.g., each subset contained a scary snake picture, a common household everyday object, etc.), visual complexity, and aversiveness/arousal (the unanalyzed predictable image subset was also matched on these factors). Every image was presented once for a total of 36 trials.

To probe modulation of startle reactivity, 95-decibel white noise bursts (50 ms, instantaneous rise time) were also presented, binaurally over headphones, throughout the task. Two probes occurred on each trial, with (a) one probe during the cuing interval, either 2.5 or 1 s prior to picture onset, and (b) one probe during picture perception, 2 s after picture onset. Probes were presented at different points across the cue interval to keep the task consistent with its instantiation in prior work that examined how startle modulation might change with the approach of an anticipated stimulus (Sege et al., 2014). An equal number of probes (6) occurred at each time point for scary and everyday content. In addition, four probes were presented before the task to habituate startle responding to a stable level (Blumenthal et al., 2005).

2.3 |. Data collection

Measures of primary interest in this study were reflex and neural physiology indices collected during task completion. As a reflex measure of interest, blink reactivity to startle-eliciting probes was indexed by measuring orbicularis oculi electromyography (EMG) using two Ag/AgCl sensors placed under the left eye. As a neural measure of interest, continuous electroencephalography (EEG) was collected using a Brain Products actiCHamp amplifier and 32 active sensors arrayed in the International 10−20 system and held on the head using EasyCap cloth caps. EEG data were online referenced to an electrical signal within the amplifier. Finally, in addition to reflex EMG and EEG, vertical and horizontal eye movements were indexed by measuring electrooculogram (EOG) signal with Ag/AgCl electrodes placed above and below the right eye and also on the outer canthi of each eye. All signals (EEG, EMG, and EOG) were sampled at 500 Hz.

2.4 |. Procedure

Prior to completion of the study task, demographics information including age and race were collected from parents/legal guardians via structured self-report questionnaire. Because the task reported here was administered as part of a larger project, consent/sensor attachment occurred ~30 min–1 h before task administration. Just prior to the anticipation task, participants were told which color would signal “something scary,” which would signal “something you might see every day,” and which meant “you won’t know if the next picture will be scary or something you see might every day.” Participants were also told that they would be asked after the task to report how many predictable things they saw (to encourage engagement). Finally, participants were told to ignore noises heard over headphones. Following instructions, participants completed the task with no interruptions.

2.5 |. Data processing

After data collection, processing of startle blink reflexes began with filtering (90–250 Hz), rectification, and integration (20 ms time constant) of orbicularis oculi EMG. Magnitudes of blink reflexes were then scored by using Brain Vision software to deviate the EMG value at response peak from the value at the start of the response. For analysis, startles were standardized for each participant, relative to that participant’s mean and standard deviation, and expressed as T scores.

To process EEG data offline, signal was re-referenced to linked mastoids and filtered using Butterworth low-pass and high-pass filters with one-third amplitude cutoffs of 30 Hz and 0.01 Hz (8 dB/ octave roll-off). Next, data were segmented from 200 ms pre-probe onset to 1000 ms post-probe onset. After segmentation, eye movement artifact was removed by inputting EOG data collected from around the eyes into a blink detection algorithm for correction (Gratton et al., 1983); remaining artifact was then removed based on criteria of overly large (≥200 μV) or small (≤0.50 μV) voltage shifts in rolling 200 ms windows moving through each segment in 50 ms increments (Keil et al., 2014). Individual channels with artifact were removed on each trial, and trials with more than six artifactual channels were removed prior to averaging (on average, <1 trial removed per condition). Finally, data were baseline-corrected for each segment and trial-averaged within condition (with weighted averaging at each sensor to correct for unequal trial numbers).

Following pre-processing, probe-related ERP components were quantified for analysis by (1) visually inspecting ERP waveforms and scalp topographies to derive a time window and sensor cluster of maximal voltage, and (2) averaging signal within the selected window and cluster to derive a single numerical index of component amplitude. For each component, identified time windows and sensor clusters were compared against prior literature to determine their validity. Though significant deviations from prior literature would have been noted, observed timings and distributions were consistent with what has been reported for each component. Since determination of time and sensor windows was data driven, actual windows are presented in Section 4.

2.6 |. Data analysis

In analyses of task data using JASP version 0.16.3, experimental factors (cue condition, cue-interval probe time, picture content, and picture content predictability) were treated as repeated measures (Jennings, 1987). For the cue interval, startle and probe ERPs were analyzed in 3 (Condition: scary-predicting cue, everyday-predicting cue, content non-predictive cue) × 2 (probe time: 2.5 s or 1 s pre-picture onset) × 3 (grade: third, sixth, ninth) mixed-effects analyses of variance (ANOVAs). For the picture interval, startle and probe ERPs were analyzed in separate 2 (picture content: scary, everyday) × 2 (predictability: predictable content, unpredictable content) × 3 (Grade: third, sixth, ninth) mixed-effects ANOVAs.

Along with primary aim analyses, preliminary exploration of effects of sex or race was also done (recruitment for the larger project emphasized representativeness, relative to recruitment area, for these factors). Biological sex and race were explored in separate analyses by adding each between-subjects factor along with grade to analyses described above (full factorial ANOVA tested, including all higher level interactions). For statistical reasons (i.e., to ensure roughly equivalent group sizes), race was explored using a two-level factor (White/European American vs. minority) in this preliminary analysis.

For all omnibus tests, Greenhouse–Geisser corrected F-values are reported in accordance with recommendations for psychophysiological research (Keselman, 1998). Effect size estimates are also presented for each statistic (partial eta-squared for F-values, Cohen’s d for t statistics).

3 |. RESULTS

3.1 |. Sample characteristics

Table 1 provides age, biological sex, and race/ethnicity information for each grade (data are the same as in Study 1 but are recapitulated here for thoroughness). The entire sample was slightly more male (51.5%) than female. Slightly over half (54.6%) of the sample identified as White/European American, while 30.8% identified as Black/African American and 14.6% as multiracial or another designation. Note that 10.4% of the sample identified as Hispanic. Pearson χ2 tests indicated that ninth graders had a higher percentage of females than did sixth graders (Χ2(1,176) = 5.4, p = .02), but they did not differ from the third graders (Χ2(1,165) = 2.4, p = .12), and the third- and sixth-grade cohorts did not differ (Χ2(1,174) = 0.5, p = .47). Grades did not differ in percentage of females identified as White/European American (Χ2 (2,247) = 0.97, p = .62), Black/African American, (Χ2(2,247) = 2.6, p = .28), or another designation (Χ2(2,247) = 2.0, p = .38). They also did not differ in percentage of females who are identified as Hispanic (Χ2(2,247) = 0.3, p = .85).

TABLE 1.

Sample characteristics.

Third grade (n = 83) Sixth grade (n = 93) Ninth grade (n = 84)
M (SD), age 8.5 (0.5) 11.5 (0.8) 14.4 (0.6)
N (%), female 39 (47.0) 38 (40.9) 49 (58.3)
Race
N (%), White/European American 48 (57.8) 51 (54.8) 43 (51.2)
N (%), Black/African American 20 (24.1) 32 (34.4) 28 (33.3)
N (%), Other race 15 (18.1) 10 (10.8) 13 (15.5)
Ethnicity
N (%), Hispanic 9 (10.8) 8 (8.6) 9 (10.7)
Open in a new tab

3.2 |. Defensive reactivity to peripheral probes: Startle blink reflex

Anticipation:

Figure 2A depicts how blink startle reflex reactivity was modulated throughout anticipation. During anticipation, startle reactivity was modulated by condition (F(2,257)= 10.3, p<.001, ηp2 = 0.04) but not probe time (F(1,258) = 0.1, p = .73, ηp2 = 0.00) or the condition × probe time interaction (F(2,257) = 1.7, p = .19, ηp2 = 0.01). Anticipatory startle was also not modulated by grade (F(2,257) = 1.2, p = .31, ηp2 = 0.01) or an interaction of grade with condition (F(4,514) = 0.2, p = .92, ηp2 = 0.00), probe time (F(2,257) = 0.5, p = .60, ηp2 = 0.00), or condition × probe time (F(4,514) = 0.6, p = .69, ηp2 = 0.00) effects. Decomposition of the condition main effect revealed that startle magnitudes were increased during anticipation of scary, as compared to everyday (t(259) = 4.3, p < .001, d = 0.27) or unpredictable (t(259) = 3.1, p = .003, d = 0.19) contents. Blinks did not differ across everyday-predicting and nonpredictive cues (t(259) = 1.4, p = .17, d = 0.09).

FIGURE 2.

FIGURE 2

Open in a new tab

Blink reflex modulation during anticipation and perception. Panels (a) and (b) depict data for the entire sample; panel (a) depicts change in reactivity across the cue, and panel (b) depicts reactivity in the viewing interval. Panel (c) depicts reactivity in late anticipation and picture viewing within each grade group (scale identical across all charts, and identical to scales in panels (a) and (b). Dashed line in all charts indicates mean startle across all trials (T-score = 50). All error bars depict standard error of the mean for each data point.

The addition of biological sex to omnibus analyses did not reveal a moderating impact on any condition, grade, or experimental condition × grade effects. For exploratory race analyses, an interaction of race, grade, condition, and probe time did arise (F(2,251) = 3.1, p = 0.01) (no other main effects or interactions significant). Follow-up of the four-level interaction indicated that reflex magnitudes were greater for (1) early everyday cue probes in the ninth-grade White compared to the ninth-grade minority participants t(254) = 2.1, p = .04; (2) early content nonpredictive cue probes in the third-grade White compared to the third-grade minority participants (t(254) = 3.9, p < .001), and the ninth-grade White compared to the ninth-grade minority participants (t(254) = 3.9, p < .001); (3) late everyday cue probes in the third-grade White compared to the third-grade minority participants (t(254) = 3.0, p = .003), and; (4) late scary cue probes in the third-grade White compared to the third-grade minority participants (t(254) = 3.2, p = .002), and the ninth-grade White compared to the ninth-grade minority participants (t(254) = 1.9, p = .05). Examination of condition effects within each race group also indicated that White participants evinced increased startle reactivity late in scary-predicting compared to everyday-predicting (t(138) = 3.9, p < .001) and content non-predictive (t(138) = 3.2, p = .002) cues, whereas minority participants evinced nonsignificant differences between scary-predicting compared to everyday-predicting (t(117) = 1.6, p = .12), or content non-predictive (t(117) = 1.7, p = .09) cues.

Picture viewing:

After picture onset, effects of picture content (F(1,256) = 7.9, p = .006, ηp2 = 0.04) and content predictability (F(1,256) = 0.1, p = .74, ηp2 < 0.01) were modified by a content × predictability interaction (F(1,256) = 9.2, p = .003, ηp2 = 0.04) (Figure 2B). As with anticipatory startle, a grade effect was not significant (F(2,257) = 0.9, p = .40, ηp2 = 0.01) and grade did not moderate effects of content (F(2,257) = 0.03, p = .97, ηp2 = 0.00), content predictability (F(2,257) = 0.6, p = .53, ηp2 = 0.01), or the content × predictability interaction (F(2,257) = 1.5, p = .23, ηp2 = 0.01). Decomposition of the content × predictability main effect revealed that across participants (1) reflexes elicited during scary pictures whose content could not be predicted were larger than reflexes elicited during everyday pictures whose content could not be predicted (t(259) = 4.3, p < .001, d = 0.27), and (2) reflexes elicited during scary pictures whose content could be predicted did not differ from blinks elicited during everyday images whose content could be predicted (t(259) = 0.08, p = .94, d = 0.01). Comparing content-unpredictable to content-predictable pictures also showed that startle was larger during unpredictable scary images than during predictable scary images (t(259) = 2.5, p = .01, d = 0.16) and startle was smaller during unpredictable everyday images than during predictable everyday images (t(259) = 2.1, p = .04, d = 0.15).

The addition of biological sex or race/ethnicity to omnibus analyses did not reveal moderating effects for picture-interval startle. Thus, for both biological sex and race, interactions with valence, predictability, grade, or any interaction of these effects were nonsignificant for the startle reflex.

3.3 |. Cortical processing of peripheral probes: Event-related potentials

Figures 3 and 4 show time series and scalp distributions for probe ERPs in the anticipatory and picture viewing intervals, respectively. Probe ERPs were characterized by prototypical N100, P200, and P300 components that were averaged in the following windows of maximal voltage: (1) For the N100, maximal negativity arose in a window from 84 to 124 ms post-probe and in a frontocentral sensor cluster (FC1, FC2, Cz); (2) for the P200, positivity was maximal in a window from 156 to 196 ms post-probe and in a frontocentral sensor cluster (Fz, FC1, FC2, Cz); and (3) for P300, positivity was maximal in a window from 266 to 316 ms post-probe and in a centroparietal sensor cluster (Cz, CP1, CP2, Pz).

FIGURE 3.

FIGURE 3

Open in a new tab

Startle probe event-related potential across scary-predicting, everyday-predicting, and content nonpredictive cues. Waveforms depict signal averaged across sensors in the cluster used for P300 analysis (Cz, CP1, CP2, Pz). In each waveform, boxes indicate analysis time windows for N100 (light gray), P200 (dark gray), and P300 (black) components. Scalp topographies depict amplitude of the P300.

FIGURE 4.

FIGURE 4

Open in a new tab

Probe event-related potential for probes presented during picture viewing. Waveforms depict signal averaged across sensors in the cluster used for P300 analysis (Cz, CP1, CP2, Pz). In each waveform, boxes indicate analysis time windows for N100 (light gray), P200 (dark gray), and P300 (black) components. Scalp topographies depict amplitude of the P300.

Anticipation:

In the anticipatory interval, N100 amplitude was not modulated by condition (F(2, 256) = 0.5, p = .59, ηp2 < 0.01), probe presentation time (F(2, 256) = 0.1, p = .81, ηp2 < 0.01), or their interaction (F(2,256) = 0.4, p = .67, ηp2 < 0.01). For P200, conversely, an interaction of condition and probe presentation time (F(2,256) = 4.2, p = .02, ηp2 = 0.02), qualified condition (F(2,256) = 0.4, p = .64, ηp2 < 0.01), and probe time (F(2, 256) = 0.7, p = .41, ηp2 < 0.01) effects. In decomposing the condition × probe time interaction, follow-up paired-samples tests indicated no condition differences early in the cue but significant differences later in the interval—such that P200 amplitude was increased late in scary-predicting cues as compared to late in everyday-predicting cues (t(258) = 2.1, p = .04, d = 0.14). Comparison across probe times also indicated that P200 amplitudes increased from early to late probes during the scary-predicting cue (t(258) = 2.2, p = .03, d = 0.15), whereas they did not change from early to late probes during cues predicting every day or unknown content. Grade did not moderate any N100 or P200 effects.

In the P300 window, amplitude was modulated by probe presentation time (F(1,257) = 9.4, p = .002, ηp2 = 0.04), but not condition (F(2,256) = 0.6, p = .53, ηp2 < 0.01), or the condition × probe time interaction (F(2,257) = 0.6, p = .34, ηp2 < 0.01). Decomposition of the probe time main effect revealed that, across conditions, P300 amplitude increased from early to later probe presentation times (t(259) = 3.1, p = .002, d = 0.19). In addition to experimental effects, grade also modulated overall probe P300 amplitude (F(2,257) = 6.7, p = .002, ηp2 = 0.05), such that third graders had increased P300s as compared to sixth graders (t(257) = 2.5, p = .01, d = 0.35), and ninth graders (t(257) = 3.6, p < .001, d = 0.54). Despite the overall grade effect, grade did not moderate condition (F(4,514) = 1.1, p = .34, ηp2 = 0.01), probe time (F(2, 257) = 0.4, p = .66, ηp2 < 0.01), or interaction (F(4, 514) = 0.8, p = .54, ηp2 = 0.01) effects.

The addition of biological sex to omnibus analyses did not reveal significant interactions with experimental condition, grade, or experimental condition × grade effects for the N100, P200, and P300. For the preliminary analysis of race, an interaction of race, grade, condition, and probe time arose for the P300 component (F(2, 253) = 2.6, p = .04; no other main effects or interactions significant). Follow-up of the four-level interaction indicated that P300 amplitude was increased for (1) early scary cue probes in the sixth-grade White compared to the sixth-grade minority participants (t(254) = 2.1, p = .04), and the ninth-grade White compared to the ninth-grade minority participants (t(254) = 2.6, p = .001); (2) early everyday cue probes in the third-grade White compared to the third-grade minority participants (t(254) = 2.6, p = .01), the sixth-grade White compared to the sixth-grade minority participants (t(254) = 3.0, p = .003), and the ninth-grade White compared to the ninth-grade minority participants (t(254) = 2.5, p = .01); (3) early content nonpredictive cue probes in the third-grade White compared to the third-grade minority participants (t(254) = 3.1, p = .003), and ninth-grade White compared to the ninth-grade minority participants (t(254) = 2.6, p = .009); (4) late scary cue probes in the third-grade White compared to the third-grade minority participants (t(254) = 3.5, p < .001), and the sixth-grade White compared to the sixth-grade minority participants (t(254) = 1.9, p = .06; (5) late everyday cue probes in the third-grade White compared to the third-grade minority participants (t(254) = 3.5, p < .001), and the sixth-grade White compared to the sixth-grade minority participants (t(254) = 2.2, p = .03); and (6) late content nonpredictive cue probes in the sixth-grade White compared to the sixth-grade minority participants (t(254) = 2.1, p = .04), and the ninth-grade White compared to the ninth-grade minority participants (t(254) = 2.7, p = .007). Next, comparisons across conditions within each grade/race subgroup indicated only one significant difference in any group—an increased P300 amplitude for early everyday compared to early content nonpredictive cue probes in the sixth-grade White participants (t(50) = 2.7, p = .01).

Picture viewing:

Figure 4 depicts the ERP for startle probes presented during picture viewing. Prominent N100, P200, and P300 components were again apparent, and again, N100 amplitude was not modulated by picture content (F(1,257) = 0.5, p = .48, ηp2 < 0.01), content predictability (F(1,257) = 0.1, p = .82, ηp2 < 0.01), or their interaction (F(1,257) = 0.2, p = .68, ηp2 < 0.01). For P200, amplitude was not modulated by predictability (F(1,257) = 0.16, p = .69, ηp2 < 0.01) or the content × predictability interaction (F(1,257) = 0.00, p = .95, ηp2 < 0.01), but it was modulated by content (F(1,257) = 7.2, p = .008, ηp2 = 0.02). Follow-up decomposition of the content main effect indicated that, regardless of content predictability, P200 amplitude increased during scary, relative to everyday, pictures (t(259) = 2.0, p = .05, d = 14). Grade did not affect N100 or P200.

In the P300 window, amplitude was not modulated by content predictability (F(1,257) = 0.56, p = .46, ηp2 < 0.01) or the content × predictability interaction (F(1,257) = 0.00, p = .97, ηp2 < 0.01), but it was modulated by picture content (F(1,257) = 4.9, p = .03, ηp2 = 0.02). Decomposition of the content main effect revealed that, across predictability conditions, P300 amplitude was reduced during scary, relative to everyday, pictures (t(259) = 2.3, p = .02, d = 0.14). In addition to this experimental effect, grade also modulated P300 amplitudes (F(1,257) = 3.8, p = .02, ηp2 = 0.03), such that overall P300 amplitude was more positive for third graders than for ninth graders (t(257) = 2.7, p = .007). At the same time, grade did not affect P300 modulation by content, content predictability, or their interaction.

The addition of biological sex to omnibus analyses did not reveal any significant interactions with experimental conditions, grade, and experimental condition × grade effects for the N100, P200, and P300. For the preliminary analysis of race, race × grade × content predictability interactions arose for N100 (F(2, 253) = 6.1, p = .003) and P200 (F(2, 253) = 8.9, p < .001) components (no other main effects or interactions significant). For the N100, the sixth-grade White participants showed greater N100 negativity during unpredictable-content compared to predictable-content picture viewing (t(50) = 2.2, p = .03), while all other groups showed no difference. N100 amplitudes did not differ when compared across grade/race subgroups within predictable-content or unpredictable-content conditions. For the P200, positivity was increased for probes during predictable-content compared to unpredictable-content conditions for the third-grade minority (t(34) = 2.6, p = .01) and the sixth-grade White (t(50) = 2.0, p = .05) participants. At the same time, the third-grade White participants showed greater P200 amplitude in unpredictable-content compared to predictable-content conditions (t(34) = 2.1, p = .04) and all other groups showed no difference. As with N100, P200 amplitudes did not differ across grade/race sub-groups within predictable-content or unpredictable-content conditions.

4 |. DISCUSSION

In companionship with another article in this issue that describes electrocortical markers of anticipation and perceptual processing of task events themselves, current data characterize how peri-adolescent individuals regulate defensive reactivity and modulate peripheral attention concurrently with that processing. Results from the cuing interval revealed that pre-adolescent and early-adolescent individuals, like adults (see Sege et al., 2014), potentiate blink reflexes throughout anticipation of scary, relative to everyday, images—suggesting sustained activation of a defensive response system. In addition, an increase of probe P200 during scary- compared to everyday-predicting cues suggests enhanced attentional registration of peripheral stimuli during aversive anticipation to go along with the enhanced defensive responding. Moving into the exposure window, downregulation of defensive response (startle) priming was then apparent during predictable aversive events for all grades—that is, the priming that arose when viewing unpredictable scary, compared to everyday, pictures disappeared when image content was made predictable (as also seen in adults; Sege et al., 2015). At the same time, probe ERPs were still modulated during predictable scary (relative to everyday) images, and this time reduced positivity in the P300 window was apparent (regardless of grade) as, again, has been observed in adults (see Schupp et al., 1997). Together, picture interval data suggest that defensive priming is downregulated for predictable aversive events even as attentional modulation is not altered by predictability—and this is true even for youth who have not yet gone through the adolescent transition.

Potentiated blink reflex reactivity during anticipation of upcoming aversive events is in line with work that also finds adult-like blink potentiation in diffuse contexts of threatened aversive exposure (i.e., the NPU task; Nelson & Hajcak, 2017). Taken with the NPU data, current results provide further evidence that a basic ability to prime defensive system activation in unconditioned threat anticipation contexts (be they diffuse or discrete cue contexts) is mature even prior to the adolescent transition—whereas fear conditioning data suggest that an ability to learn predictive cue-stimulus associations does go through development during the transition to adolescence (Glenn et al., 2012). Regarding increased probe P200 positivity in the present task context, meanwhile, this is in contrast with an NPU task wherein (for children and adolescents as for adults) reduced rather than increased postivities are observed under threat and this difference arises in the P300, not the P200, window (Nelson & Hajcak, 2017). Together, current and NPU data further support some independence of probe-related defensive reactivity and attention—such that reflex priming can occur regardless of whether attention is allocated toward or away from the probe (see Cuthbert et al., 1998 for a similar interpretation). As for central developmental aims of this study, meanwhile, current findings suggest that neither defensive system activation nor modulation of peripheral attention during predictable event anticipation change across the transition to early adolescence.

Moving from anticipation to the subsequent exposure period, blink reflex and ERP modulation patterns—that is, reduced reflex priming for predictable unpleasant pictures that coincided with no impact of predictability on probe P300 modulation—are consistent with an emotion regulation capacity that supports attenuating limbic/sub-cortical defensive reactivity even as perceptual/ attentional responses remain unchanged. Regarding development effects, current results suggest that the capacity to automatically downregulate emotional reactivity during predictable unpleasant events is in place even prior to adolescence, unlike volitional/ controlled regulation strategies that do appear to develop across the adolescent transition (see Silvers, 2022). Put in neurobiological terms, current findings could also be analogous to findings of a downregulation of unconditioned emotional responses in conditioning contexts that are conceptualized to be driven by bottom-up formation of cue—stimulus associations (Ison et al., 1990; Lykken & Tellegen, 1974)—and appear to be especially associated with amygdala activity (Wood et al., 2012). To the degree current cued picture data and prior conditioning data reflect a similar underlying process, current findings then suggest that this basic capacity to implicitly regulate emotional reactivity as a function of cued predictability is early-maturing and, thus, in place prior to the adolescent transition. While more complex regulatory functions and the frontal circuits that support them might still be developing (e.g., Crone, 2014), then current data suggest that some capacities to adaptively regulate emotional reactivity are present even prior to the adolescent transition.

In addition to main-aim analyses that compared peripheral defensive reactivity and attention across pre-adolescent and early adolescent cohorts, potential differences as a function of biological sex and race were explored in a very preliminary manner. While no differences arose due to biological sex, one critical finding with regard to race was that data missingness differed such that those who were identified as White/European American were missing more data than those who did identify as White/European American. Data were excluded because of poor recording quality and/or excessive movement (i.e., disengagement from task instructions). In light of this, differences in data missingness could be due to recording differences related to an inability of the equipment used here to collect high-quality data across all hair/scalp conditions (Louis et al., 2022), or to differences in general task engagement that perhaps stem from different historical interactions with research in the United States (Fryer et al., 2016). Independently of the cause of missingness, however, one key implication is that current findings could be more representative of White/European American individuals than of non-White individuals. As a guide for future work, current findings then further indicate the need to replicate neuroscience methods with more equitable data collection equipment and methods due to the potential for systematic bias inherent in current widely used systems (e.g., Etienne et al., 2020).

Keeping this in mind, when data from the still substantial number of Minority participants who did have data (n = 108) were compared to data from White/European American youth, it was found that White/European-American individuals generally showed greater aversive reflex potentiation during anticipation than did others, and probe ERP (specifically P300) positivities were also generally increased in White/European-American individuals. In the subsequent picture-viewing interval, meanwhile, blink reactivity differences did not arise and race also did not alter the way probe ERPs were modulated by picture content or predictability—suggesting that race perhaps did not impact defensive or attentional engagement during picture viewing itself. Together, results are inconsistent with an interpretation that aversive stimuli were differentially salient to each group (given that picture-interval data did not change) and they instead could suggest differences in how defensive reactivity is modulated in anticipation. Again, however, more work is needed to rule out other factors that could contribute to race/ethnicity effects including data collection factors, experimental factors (e.g., under-representation of minorities in task stimuli), or differential relationships with research writ large across race/ethnicity groups (e.g., Fryer et al., 2016). Though current interpretations are highly speculative, then continued research that addresses systematic measurement biases and includes more sophisticated modeling with other important sociocultural correlates is highly needed to determine the generalizability of neuroscience findings across individuals of different sociocultural backgrounds—research that is still sorely lacking in neuroscience (e.g., Rapp et al., 2021).

As discussed in the companion article, in addition to further examining important sociodemographic variables, it will also be critical for future research to address limitations of current work with respect to accounting for the marked inter-individual variability that exists in pubertal timing (by carefully measuring pubertal status; Marceau et al., 2011; Stroud & Davila, 2014) and examining development throughout the child-to-adult transition (by expanding the recruitment window to include younger children, older adolescents, and adults). In addition, future work should address limitations related to the highly routinized nature of this task context and the lack of a pleasant condition to determine if effects are specific to unpleasant stimuli. Finally, future work should test neural mechanisms of predictable unpleasant processing in the functional magnetic resonance imaging scanner—and as informed also by companion work to this article, special attention should be paid to whether different areas of a multi-regional predictable event processing network show different patterns of change in activation across the adolescent transition (e.g., such that executive control regions show changing activity patterns across the adolescent transition, whereas emotional response regions do not).

5 |. CONCLUSIONS

Even prior to future work, current results and those of the companion article begin to describe how neurophysiological correlates of predictable aversive event processing might, for many youth, be consistent with adults in many ways, and thus reflect basic processing capacities established prior to adolescence (while also pointing to a need to further explore potential sociodemographic moderators). In concert with the companion article, it should be pointed out that one potential age-related difference that did arise was in how aversive anticipation processing itself was carried out—such that younger individuals appeared to show different anticipatory ERP modulation than adults (see companion article) even as current data also revealed adult-like defensive reflex priming. As a whole, these findings not only further support at least partial independence of defensive regulation and attentional/executive processing (such that a sub-cortical/limbic defensive response system can be activated regardless of how anticipatory attention is deployed)—they also raise an intriguing possibility that operation of these systems in predictable event contexts might mature at different rates (a possibility which should now be further examined in continued research).

Echoing broader interpretations in the companion article, this paper then also again exemplifies the fact that real-world emotional processing is not a monolithic phenomenon that develops as a whole, but instead involves complex interactions of higher order top-down (e.g., attentional deployment) and basic bottom-up (e.g., defensive priming) systems that could show varying developmental effects (Lang & Bradley, 2013; LeDoux, 2013; Mobbs et al., 2015). While the companion article emphasizes variability of processing across contexts (in that case, anticipation vs. perception), data presented here emphasize that developmental effects might also vary across dimensions of processing (e.g., attentional processing vs. defensive modulation) that co-occur in any given context (see Lang, Bradley, & Cuthbert, 1997 for review). As stated in the companion article, the variegated nature of emotion processing raises a possibility that understanding emotional experience changes in the adolescent transition may depend just as much on studying aspects of processing that do not change as it does on studying aspects of processing that do change. Related to this, one intriguing possible interpretation of current data is that a co-occurrence of underdeveloped representational/attentional processing (as suggested by cue-related ERP data in the companion) with mature activation of defensive systems during anticipation of unpleasant events could create emotional ambiguity in these contexts that, in turn, makes them more difficult to manage for younger individuals. While this interpretation is highly speculative at this point, it nevertheless again emphasizes the importance of further studying how aspects of emotion processing that do not change in the adolescent transition might interact with processes that do change to contribute to emotional tumult that arises for many at this age (e.g., Costello et al., 2003).

To summarize, current data characterize modulations of peripheral defensive responding and attentional monitoring during predictable aversive event processing that appear similar to patterns also seen in adults. Current findings suggest that pre-adolescent and early-adolescent individuals show defensive activation priming and enhanced peripheral stimulus processing during aversive anticipation, and then they downregulate defensive reactivity while not changing peripheral stimulus processing once the predictable aversive event occurs. Together with companion article results, current data indicate many aspects of predictable aversive event processing that are stable even prior to the adolescent transition, while also indicating a need to further explore potential sociocultural moderators of these effects along with boundary conditions. Importantly, current findings also emphasize a general need to account for multi-dimensionality and context sensitivity in emotion processing when studying how it changes across adolescence.

ACKNOWLEDGMENTS

This study was supported in part by the National Institute of Mental Health (NIMH) under Grant # R01 MH112209 and the National Center for Advancing Translational Sciences (NCATS) under Grant # UL1 TR001450. The first author’s contributions were supported in part by NIMH Grant # K23 MH123931–01A1. NIMH and NCATS are part of the National Institutes of Health (NIH). The content of this study is sole responsibility of the authors and does not necessarily represent official views of NIH.

Funding information

National Institute of Mental Health, Grant/Award Number: R01 MH112209; National Center for Advancing Translational Sciences, Grant/Award Number: UL1 TR001450

Footnotes

1

Analyses were replicated for each individual measure using complete data for that measure. Patterns of effects did not change in any instance.

2

IAPS Numbers for analyzed images: Scary: 1050, 1052, 1201, 1205, 2691, 3530, 6230, 6260, 6300, 6510, 6571, 9427; Everyday: 5390, 5500, 5534, 5750, 5800, 7002, 7009, 7039, 7061, 7090, 7175, 7190. IAPS Numbers for unanalyzed images: Scary: 1200, 1202, 1300, 1525, 2811, 6231; Everyday: 5731, 7004, 7035, 7041, 7100, 7165.

3

Normative ratings of hedonic valence (1 = very unpleasant, 9 = very pleasant, 5 = neutral) and arousal (1 = not arousing, 9 = very arousing) are available. Scary images were more unpleasant (M= 3.0; t(34) = 13; p < .001) and arousing (M = 6.5; t(34) = 19.8, p < . 001) than everyday images (Munpleasantness = 5.4;Marousal = 2.9).

CONFLICTS OF INTEREST STATEMENT

The authors declare no conflicts of interest.

DATA AVAILABILITY STATEMENT

Data are available from the corresponding author upon request.

REFERENCES

  1. Blumenthal TD, Cuthbert BN, Filion DL, Hackley S, Lipp OV, & van Boxtel A. (2005). Committee report: Guidelines for human startle eyeblink electromyographic studies. Psychophysiology, 42(1), 1–15. 10.1111/j.1469-8986.2005.00271.x [DOI] [PubMed] [Google Scholar]
  2. Bradley MM, Codispoti M, Cuthbert BN, & Lang PJ (2001). Emotion and motivation I: Defensive and appetitive reactions in picture processing. Emotion, 1(3), 276–298. 10.1037/1528-3542.1.3.276 [DOI] [PubMed] [Google Scholar]
  3. Caballero A, Granberg R, & Tseng KY (2016). Mechanisms contributing to prefrontal cortex maturation during adolescence. Neuroscience and Biobehavioral Reviews, 70, 4–12. 10.1016/j.neubiorev.2016.05.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Costello EJ, Mustillo S, Erkanli A, Keeler G, & Angold A. (2003). Prevalence and development of psychiatric disorders in childhood and adolescence. Archives of General Psychiatry, 60(8), 837–844. 10.1001/archpsyc.60.8.837 [DOI] [PubMed] [Google Scholar]
  5. Crone EA (2014). The role of the medial frontal cortex in the development of cognitive and social-affective performance monitoring. Psychophysiology, 51(10), 943–950. 10.1111/psyp.12252 [DOI] [PubMed] [Google Scholar]
  6. Cuthbert BN, Schupp HT, Bradley MM, McManis M, & Lang PJ (1998). Probing affective pictures: Attended startle and tone probes. Psychophysiology, 35(3), 344–347. 10.1017/S0048577298970536 [DOI] [PubMed] [Google Scholar]
  7. Dhamija D, Tuvblad C, Dawson ME, Raine A, & Baker LA (2017). Heritability of startle reactivity and affect modified startle. International Journal of 1Psychophysiology, 115, 57–64. 10.1016/j.ijpsycho.2016.09.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dichter GS, Tomarken AJ, & Baucom BR (2002). Startle modulation before, during and after exposure to emotional stimuli. International Journal of Psychophysiology, 43(2), 191–196. 10.1016/S01678760(01)00170-2 [DOI] [PubMed] [Google Scholar]
  9. Etienne A, Laroia T, Wiegle H, Afelin A, Kelly SK, Krishnan A, & Grover P. (2020). Novel electrodes for reliable EEG recordings on coarse and curly hair. Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society Annual International Conference, 2020. 6151–6154. https://doi.org/10/1109/EMBC44109.2020.9176067 [DOI] [PubMed] [Google Scholar]
  10. Fryer CS, Passmore SR, Maietta RC, Petruzzelli J, Casper ET, Brown NA, Butler J, Garza MA, Thomas SB, & Quinn SC (2016). The symbolic value and limitations of racial concordance in minority engagement research. Qualitative Health Research, 26(6), 1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Glenn CR, Klein DN, Lissek S, Britton JC, Pine DS, & Hajcak G. (2012). The development of fear learning and generalization in 8–13 year-olds. Developmental Psychobiology, 54(7), 675–684. 10.1002/dev.20616 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Gratton G, Coles MG, & Donchin E. (1983). A new method for off-line removal of ocular artifact. Electroencephalography and Clinical Neurophysiology, 55(4), 468–484. 10.1016/0013-4694(83)90135-9 [DOI] [PubMed] [Google Scholar]
  13. Gross JJ, & Thompson RA (2007). Emotion regulation: Conceptual foundations. In Gross JJ (Ed.), Handbook of emotion regulation (pp. 3–24) The Guilford Press. [Google Scholar]
  14. Helion C, Krueger SM, & Ochsner KN (2019). Chapter 1–Emotion regulation across the life span. Handbook of Clinical Neurology, 163, 257–280. 10.1016/B978-0-12-804281-6.00014-8 [DOI] [PubMed] [Google Scholar]
  15. Howard S. (2019). Timing of puberty: Why is it changing and why does it matter? The Endocrinologist, 134. [Google Scholar]
  16. Ison JR, Sanes JN, Foss JA, & Pinckney LA (1990). Facilitation and inhibition of the human startle blink reflexes by stimulus anticipation. Behavioral Neuroscience, 104(3), 418–429. 10.1037/0735-7044.104.3.418 [DOI] [PubMed] [Google Scholar]
  17. Jennings JR (1987). Editorial policy on analyses of variance with repeated measures. Psychophysiology, 24(4), 474–475. 10.1111/j.1469-8986.1987.tb00320.x [DOI] [Google Scholar]
  18. Keil A, Debener S, Gratton G, Junghöfer M, Kappenman ES, Luck SJ, Luu P, Miller GA, & Yee CM (2014). Committee report: Publication guidelines and recommendations for studies using electroencephalography and magnetoencephalography. Psychophysiology, 51(1), 1–21. [DOI] [PubMed] [Google Scholar]
  19. Keselman HJ (1998). Testing treatment effects in repeated measures designs: An update for psychophysiological researchers. Psychophysiology, 35(4), 470–478. 10.1111/1469-8986.3540470 [DOI] [PubMed] [Google Scholar]
  20. Lang PJ, Bradley MM, & Cuthbert BN (1997). Motivated attention: Affect, activation, and action. In Lang PJ, Simons RF, & Balaban MT (Eds.), Attention and orienting: Sensory and motivational processes (pp. 97–135). Lawrence Erlbaum Associates Publishers. [Google Scholar]
  21. Lang PJ, & Bradley MM (2013). Appetitive and defensive motivation: Goal-directed or goal-determined? Emotion Review, 5(3), 230–234. 10.1177/1754073913477511 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Lang PJ, Bradley MM, & Cuthbert BN (2008). International affective picture system (IAPS): Affective ratings of pictures and instruction manual. Technical Report A-8. University of Florida, Gainesville, FL. [Google Scholar]
  23. Lang PJ, & Davis M. (2006). Emotion, motivation, and the brain: Reflex foundations in animal and human research. Progress in Brain Research, 156, 3–29. 10.1016/S0079-6123(06)56001-7 [DOI] [PubMed] [Google Scholar]
  24. Latham MD, Cook N, Simmons JG, Byrne ML, Kettle JWL, Schwartz O, Vijayakumar N, Whittle S, & Allen NB (2017). Physiological correlates of emotional reactivity and regulation in early adolescents. Biological Psychology, 127, 229–238. 10.1016/j.biopsycho.2017.07.018 [DOI] [PubMed] [Google Scholar]
  25. LeDoux JE (2013). Rethinking the emotional brain. Neuron, 73(4), 653–676. 10.1016/j.neuron.2012.02.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Louis CC, Webster CT, Gloe LM, & Moser JS (2022). Hair me out: Highlighting systematic exlusion in psychophysiological methods and recommendations to increase inclusion. Frontiers in Human Neuroscience, 16, 1058953. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lykken DT, & Tellegen A. (1974). On the validity of the preception hypothesis. Psychophysiology, 11(2), 125–132. 10.1111/j.1469-8986.1974.tb00833.x [DOI] [PubMed] [Google Scholar]
  28. Marceau K, Ram N, Houts RM, Grimm KJ, & Susman EJ (2011). Individual differences in boys’ and girls’ timing and temp of puberty: Modeling development with nonlinear growth models. Developmental Psychology, 47(5), 1389–1409. 10.1037/a0023838 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Mobbs D, Hagan CC, Dalgleish T, Silston B, & Prévost C. (2015). The ecology of human fear: Survival optimization and the nervous system. Frontiers in Neuroscience, 9, 55. 10.3389/fnins.2015.00055 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Näätänen R, & Picton TW (1986). N2 and automatic versus controlled processes. Electroencephalography and Clinical Neurophysiology, 38, 169–186. [PubMed] [Google Scholar]
  31. Nelson BD, & Hajcak G. (2017). Anxiety and depression symptom dimensions demonstrate unique relationships with the startle reflex in anticipation of unpredictable threat in 8 to 14 year-old girls. Journal of Abnormal Child Psychology, 45(2), 397–410. 10.1007/s10802-016-0169-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Nitschke JB, Larson CL, Smoller MJ, Navin SD, Pederson AJC, Ruffalo D, Mackiewicz KL, Gray SM, Victor E, & Davidson RJ (2002). Startle potentiation in aversive anticipation: Evidence for state but not trait effects. Psychophysiology, 39(2), 254–258. 10.1111/1469-8986.3920254 [DOI] [PubMed] [Google Scholar]
  33. Peper JS, Brouwer RM, Schnack HG, van Baal GC, van Leeuwen M, van den Berg SM, Delemarre-Van de Waal HA, Boomsma DI, Kahn RS, & Pol HEH (2009). Sex steroids and brain structure in pubertal boys and girls. Psychoneuroendocrinology, 34(3), 332–342. 10.1016/j.psyneuen.2008.09.012 [DOI] [PubMed] [Google Scholar]
  34. Polich J. (2007). Updating P300: An integrative theory of P3a and P3b. Clinical Neurophysiology, 118(10), 2128–2148. 10.1016/j.clinph.2007.04.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Rapp AM, Tan PZ, Grammer JK, Gehring WJ, Miller GA, & Chavira DA (2021). Cultural group differences in the association of neural sensitivity to social feedback and social anxiety among diverse adolescents. Journal of Psychiatric Research, 143, 400–408. 10.1016/j.jpsychires.2021.09.036 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Sabatinelli D, Bradley MM, & Lang PJ (2001). Affective startle modulation in anticipation and perception. Psychophysiology, 38(4), 719–722. 10.1111/1469-8986.3840719 [DOI] [PubMed] [Google Scholar]
  37. Schupp HT, Cuthbert BN, Bradley MM, Birbaumer N, & Lang PJ (1997). Probe P3 and blinks: Two measures of affective startle modulation. Psychophysiology, 34(1), 1–6. 10.1111/j.1469-8986.1997.tb02409.x [DOI] [PubMed] [Google Scholar]
  38. Schweizer S, Gotlib IA, & Blakemore SJ (2020). The role of affective control in emotion regulation during adolescence. Emotion, 20(1), 80–86. 10.1037/emo0000695 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Seddon JA, Rodriguez VJ, Provencher Y, Raftery-Helmer J, Hersh J, Labelle PR, & Thomassin K. (2020). Meta-analysis of the effectiveness of the Trier Social Stress Test in eliciting physiological stress responses in children and adolescents. Psychoneuroendocrinology, 116, 104582. 10.1016/j.psyneuen.2020.104582 [DOI] [PubMed] [Google Scholar]
  40. Sege CT, Bradley MM, & Lang PJ (2014). Startle modulation during emotional anticipation and perception. Psychophysiology, 51(10), 977–981. 10.1111/psyp.12244 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Sege CT, Bradley MM, & Lang PJ (2015). Prediction and perception: Defensive startle modulation. Psychophysiology, 52(12), 1664–1668. 10.1111/psyp.12546 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Silvers JA (2022). Adolescence as a pivotal period for emotion regulation development. Current Opinion in Psychology, 44, 258–263. 10.1016/j.copsyc.2021.09.023 [DOI] [PubMed] [Google Scholar]
  43. Stroud C, & Davila J. (2014). Pubertal timing. In Levesque RJR (Ed.) Encyclopedia of Adolescence (pp. 2252–2258). New York, NY: Springer. [Google Scholar]
  44. Stephens C, & Marcin A. (2022). Navigating puberty: The Tanner Stages. Healthline. Retrieved from. https://www.healthline.com/health/parenting/stages-of-puberty [Google Scholar]
  45. Tremblay K, Kraus N, McGee T, Ponton C, & Otis B. (2001). Central auditory plasticity: Changes in the N1-P2 complex after speech-sound training. Ear and Hearing, 22(2), 79–90. 10.1097/00003446-200104000-00001 [DOI] [PubMed] [Google Scholar]
  46. Wood KH, ver Hoef LW, & Knight DC (2012). Neural mechanisms underlying the conditioned diminution of the unconditioned fear response. Neuroimage, 60, 787–799. 10.1016/j.neuroimage.2011.12.048 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Data are available from the corresponding author upon request.

RESOURCES