NeuroImage 92 (2014) 267–273

Contents lists available at ScienceDirect

NeuroImage journal homepage: www.elsevier.com/locate/ynimg

The cognitive and neurobiological effects of daily stress in adolescents Ahrareh Rahdar, Adriana Galván ⁎ Department of Psychology, UCLA, USA

a r t i c l e

i n f o

Article history: Accepted 3 February 2014 Available online 16 February 2014 Keywords: Functional MRI Daily stress Adolescents Dorsolateral prefrontal cortex Cognition Response inhibition

a b s t r a c t Increased stress reactivity during adolescence coincides with maturation of cognitive abilities and development of the prefrontal cortex. Although the effects of early-life, chronic, and pervasive stress on cognition have been extensively explored across development, very little is known about the effects of naturalistic, daily stress on adolescent cognition. In this study, our goal was to use a naturalistic approach to determine whether participants' own stressful experiences from daily life impacted cognitive performance and associated neural correlates. Adolescent and adult participants provided daily ratings of stress and underwent functional magnetic resonance imaging (fMRI) twice: once under a self-reported “high-stress” state and once under a self-reported “low-stress” state. While in the scanner, participants performed a response inhibition task. Behaviorally, all participants exhibited worse response inhibition under high, versus low, stress states, an effect that was significantly stronger in adolescents. At the neural level, there was a significant age by stress interaction, such that adolescents exhibited less recruitment of the dorsolateral prefrontal cortex (DLPFC) during inhibition under high-stress versus lowstress; adults evinced the opposite activation pattern in DLPFC. These data suggest that the developing brain may be a more vulnerable target to the cognitive and neurobiological effects of daily stress. © 2014 Elsevier Inc. All rights reserved.

Introduction Relative to other developmental periods, adolescence is marked by stressful experiences and increased stress reactivity (Stroud et al., 2009), as well as psychosocial, physical and neurobiological changes (Persike and Seiffge-Krenke, 2012; Schlegel, 2001). Across different cultures, adolescents report having increased daily stress in the form of pressures from family (Bynner, 2000), school (Arnett, 2002), peers (Eccles et al., 1993; Hand and Furman, 2009) and romantic relationships (Kuttler and LaGreca, 2004). Daily stress arises from pressures of the recent past or pressures of the near future, and is the most common form of stress (Miller et al., 1994). The literature is rich with developmental studies examining the effects of chronic stress (a prolonged stressful period, often leading to serious physical or psychiatric illness, Baum and Polsusnzy, 1999) on cognition (e.g., Duckworth et al., 2012; Lupien et al., 2009; Pollak, 2005). Little is known, however, about the effects of daily stress on adolescent cognition. In contrast to chronic stress, daily stress in this study was defined as discrete (not prolonged) emotional strain resulting from demanding daily circumstances. This lack of empirical consideration is surprising, taking into account that both cognition and stress reactivity are in significant flux during this developmental time period. Our goal was to fill this gap in knowledge by examining the effects of naturally occurring (as opposed to ⁎ Corresponding author at: 1285 Franz Hall, Box 951563, Department of Psychology, UCLA, Los Angeles, CA 90095, USA. E-mail address: [email protected] (A. Galván).

http://dx.doi.org/10.1016/j.neuroimage.2014.02.007 1053-8119/© 2014 Elsevier Inc. All rights reserved.

laboratory-induced) daily stress on cognitive control and associated neural correlates in adolescents. Understanding the specific effects of daily stress is important as stress exacerbates arousal-based decisions and behavior in adolescents (Figner et al., 2009). During this developmental window, there are marked changes in cognitive abilities (increased reasoning and impulse control) and the neurobiological systems that support them (Casey et al., 2005). The brain undergoes remarkable development across adolescence (Eiland and Romeo, 2013; Galván et al., 2006; Gogtay et al., 2004; Shaw et al., 2008; Sowell et al., 1999), and the dorsolateral prefrontal cortex (PFC) is the last brain region to mature structurally and functionally (Casey et al., 2008; Chiron et al., 1992; Chugani et al., 1987; Fuster, 2001; Lewis, 1970). This region is critically involved in cognitive control, the regulation of emotional behaviors, and decision-making (Miller and Cohen, 2001). Furthermore, it is vulnerable to the effects of stress. In rats (Radley et al., 2006) and monkeys (Spinelli et al., 2009), stress reduces dendritic branching in the medial prefrontal cortex and neuronal reorganization in frontostriatal circuitry (Dias-Ferreira et al., 2009). In humans, laboratory-induced stressors alter prefrontal activation in response to performance anxiety (Dedovic et al., 2009), physiological (Porcelli and Delgado, 2009; Porcelli et al., 2012) and social stress (Eisenberger et al., 2007). Using a more naturalistic approach, Liston et al. (2009) found that a real-life, acute and discrete stressor (individuals preparing for a major academic exam vs. individuals undergoing no major psychosocial stress) selectively impaired attentional control and disrupted functional connectivity within the frontoparietal network (Liston et al., 2009).

268

A. Rahdar, A. Galván / NeuroImage 92 (2014) 267–273

A long line of adult research shows that acute stress negatively affects cognition (Galván and McGlennen, 2012; Janis, 1993; Keinan, 1987; Mather and Lighthall, 2012; Porcelli and Delgado, 2009; Preston et al., 2007), in learning, memory, decision and inhibition domains (Mather and Lighthall, 2012; Roozendaal, 2002; Sandi, 2013; Wolf, 2006). For instance, response inhibition performance in adult males is significantly impaired following acute stress (Scholz et al., 2009). Animal research has also found that rodents (Bondi et al., 2008; Hennessy et al., 1973; Lapiz-Bluhm et al., 2009; Micco et al., 1979) and monkeys exposed to stress-level cortisol treatments have impaired response inhibition (Lyons et al., 2000), which is mediated via stressinduced atrophy of prefrontal neurons (Liston et al., 2006; Radley et al., 2004). These findings have been instrumental in establishing the mechanism by which acute stress can dysregulate cognition. However, this work has been limited to adult and animal populations so the ontogenetic effects remain unknown. Current study The current study examines how daily stress impacts cognitive control and its neural correlates in adolescents. We employed an Ecological Momentary Assessment (EMA) approach to monitor participants' naturalistic stress for two weeks through text messaging. This approach is a key divergence from previous work as participants were not subjected to the most commonly used tools to induce stress, such as a laboratory manipulation of stress (e.g., public speaking, cold pressor task) or recollection of previous stressors, which both incur potential limitations; the former can suffer from a lack of ecological validity, while the latter may be limited by recall bias and/or higher-order regulation of the stressful experience. EMA is an optimal way to assess daily stress, as it minimizes recall bias and maximizes ecological validity (Bolger et al., 2003). This method has been successful in the stress literature in determining how individual differences in various domains predict daily stress reactivity (e.g., Almeida and Kessler, 1998). Further, the within-subject design allowed us to use subjects as their own controls, thus allowing us to ask novel questions. For example: How does an individual's engagement of frontal circuitry change based on daily stress? Each participant visited the laboratory twice, once on a day when they endorsed a high level of daily stress and once on a day when they endorsed a low level of daily stress. This novel approach allowed us to examine within-person, as well as developmental effects, thereby precluding potential confounds related to individual differences in laboratory-stress reactivity. At the lab, participants performed a Go/ No-go task while undergoing fMRI to assess cognitive control and as a probe for prefrontal function. Behaviorally, we predicted that daily stress would negatively affect cognitive control performance in both adolescents and adults, albeit with a stronger effect in adolescents. Furthermore, based on previous work in human adults (Ossewaarde et al., 2011; Porcelli and Delgado, 2009; Treadway et al., 2013) and in rodents (for review, see McEwen and Morrison, 2013) we predicted that compromised cognitive performance would be paralleled by reduced cortical engagement. Specifically, we hypothesized that the largest neural effect would be observed in the DLPFC, which undergoes significant development during adolescence.

consent was obtained from all adult participants, and assent was obtained from all participants under the age of 18 in accordance with procedures approved by the UCLA Institutional Review Board. The Wechsler Abbreviated Scale of Intelligence (WASI) was administered to estimate IQ; adolescents (M = 114.78) and adults (M = 117.52) did not significantly differ in IQ. Ethnic composition did not differ between age groups [adolescents: 48% Caucasian, 17% African-America, 26% Hispanic/Latino, 4% Asian-America, and 4% other; adults: 48% Caucasian, 13% AfricanAmerica, 8% Hispanic/Latino, 26% Asian-America, and 4% other] nor did socioeconomic status [χ2(1, 43) = .69, p = .24], which was categorized based on maternal education. Procedure Participants were asked to come to the lab for an initial intake, during which a short battery of questionnaires was completed, and study procedures were explained. They then completed a baseline period to assess their normative stress assessment, followed by the experimental phase when they completed two scans (see Fig. 1 for an overview of the study design). Ecological Momentary Assessment (EMA) Daily stress was assessed using the EMA method, a procedure in which participants are contacted daily via smart phone in order to capture naturally-occurring stressors as they unfold (Bolger et al., 2003). EMA has previously been found to be a successful method of capturing naturally occurring stress (e.g., Almeida et al., 2009; Galván and McGlennen, 2012). In the current study, participants were contacted three times per day for two weeks and asked to indicate overall stress level using a Likert scale (1 = no stress; 7 = very stressed). Participants were not required to have a phone with a texting plan to participate, though all participants in the current study used their own personal cellular phone. The first three days of texting were used to establish a “baseline” composite stress rating for each participant, by averaging the three overall stress ratings obtained throughout the day. We then used these baseline ratings to categorize the two laboratory visits into high-stress state and low-stress state. In order to qualify as a highstress state, participants had to endorse at least one point higher than baseline. In order to qualify as a low-stress state, participants had to endorse at least one point lower than baseline. Each participant was asked to visit UCLA when experiencing a high-stress state and when experiencing a low-stress state (stress state visit order was counterbalanced across participants). During these visits participants completed an fMRI scan in which they performed a cognitive control task (i.e., Go/No-go). The average duration between stress reporting (stressor) and brain scan was 2 h and 5 min (SD = 67 min). There were no significant differences in duration based on stress state (t(44) = 1.38, p = .17) and did not differ by age group [(adolescents: t(21) = 1.40, p = .18; high-stress scan: 2 h 18 min; low-stress scan: 1 h 51 min; Mdifference = 27 min); (adults: t(22) = .42, p = .69; high-stress scan: 2 h 12 min; low-stress scan: 2 h 7 min; Mdifference = 5 min)]. To determine whether the stress rating following this duration (i.e. the stress level at the scan) was most likely

Materials and methods Participants Participants included 45 right-handed English-speakers (n = 22 adolescents, ages 15–17, M = 16.5, SD = .76, 13 males; and n = 23 adults ages 25–30, M = 27.5, SD = 1.67, 10 males). Participants were recruited via advertisements on the UCLA campus, surrounding neighborhoods, and through Craigslist. Exclusion criteria included metal in the body (e.g., braces, permanent retainers), a diagnosis or a psychiatric or developmental disorder, claustrophobia, or pregnancy. Informed

Fig. 1. Study design.

A. Rahdar, A. Galván / NeuroImage 92 (2014) 267–273

reflective of acute stress, stress recovery or prolonged stress, we calculated the difference between the last daily stress level reported and the stress level reported at the scan, which yielded no significant differences [t(44) = .41, p = .69; (Mdifference = −0.2) and did not differ by age group [(adolescents: t(21) = 1.40, p = .18; high-stress scan: 2 h 18 min; low-stress scan: 1 h 51 min; Mdifference = 27 min); (adults: t(22) = .42, p = .69; high-stress scan: 2 h 12 min; low-stress scan: 2 h 7 min; Mdifference = 5 min)]. Sixteen percent of participants (2 adults; 5 adolescents) were unable to receive a scan due to scheduling conflicts the first time they were asked to visit UCLA when experiencing a highstress state; in these cases, we scanned them on the next occurrence (day) they reported a high-stress state. This procedural approach allowed us to obtain high-stress and low-stress state scans from all participants. Saliva collection Salivary cortisol is commonly used as an indicator of an organism's response to stress (Hellhammer et al., 2009). Cortisol concentrations in saliva have been highly correlated with those in plasma, urine, and cerebrospinal fluid (Stetler and Miller, 2008). For the current study participants were asked to collect saliva (by passive drool) at home on 2 consecutive days during the baseline phase using Salivette® test tubes (Sarstedt, Germany). Participants were instructed to collect 1 ml (1000 μl) of saliva at awakening, 30 min after awakening, at 4:30 pm, and 8:30 pm (participants noted exact time of collection). This method and timing of collection have been found to be the most common and meaningful measures of cortisol (Clements, 2013). Additionally, saliva was collected from each participant prior to and after each of their high-stress and low-stress state scans. All samples were stored in a −20 ºC in a laboratory freezer at UCLA until time of shipment. Salivary cortisol analyses were conducted as described previously by Strahler et al. (2010). fMRI Go/No-go paradigm Participants completed a standard Go/No-go (GNG) task to examine neural correlates of cognitive control (Fig. 2). Participants were presented with a series of rapid trials (1 s each), each displaying a single letter, and were instructed to respond with a button press as quickly as possible to all letters except for X. The X occurred on 25% of trials. Thus,

269

participants developed a pre-potent response to press (go) upon stimulus onset, and inhibit the go response on X trials (no-go). Response inhibition was operationalized as successful no-go trials (overriding the prepotent “go” response). Participants completed 5 blocks during one functional run. Each block contained 10 no-go trials and 30 go trials. The intertrial interval (ITI) was jittered according to a random gamma distribution (M = 0.75 s). Each block (40 trials and ITIs) lasted 70 s, and each block was separated by a twelve-second rest period. Statistical analysis of Go/No-go task A 2 (stress state) × 2 (age group) repeated measures analysis of variance (ANOVA) was conducted on the behavioral task data using SPSS. For the sake of clarity, in the current paper successful go trials will be identified as “hits”, successful no-go trials will be identified as “inhibition”, unsuccessful go trials will be identified as “misses” and unsuccessful no-go trials will be identified as “false alarms”. MRI data acquisition Functional images were acquired on a 3 Tesla Siemens Trio MRI scanner. Using a gradient-echo, echo-planar image (EPI) sequence (TR = 2 s, TE = 30 ms, flip angle: 90°, 271 volumes, 34 slices, slice thickness 4 mm). A T2*weighted, matched bandwidth (MBW), highresolution, anatomical scan and magnetization-prepared rapidacquisition gradient echo (MPRAGE) scan were acquired for registration purposes (TR: 2.3; TE: 2.1; FOV: 256; matrix: 192 × 192; sagittal plane; slice thickness: 1 mm; 160 slices). The orientation for the MBW and EPI scans was oblique axial to maximize brain coverage. MRI data preprocessing and analysis Preprocessing and statistical analyses were carried out using FSL 4.1.6 (www.fmrib.ox.ac.uk/fsl). Preprocessing included motion correction, non-brain matter removal using FSL BET, spatial smoothing (5 mm FWHM Gaussian kernel) to increase the signal-to-noise ratio, and filtered in the temporal domain using a nonlinear high-pass filter (100-s cutoff). EPI images were registered to the MBW, then to the MPRAGE, and finally into standard MNI space (MNI152, T1 2 mm) using linear registration with FSL FLIRT.

Fig. 2. Schematic of Go/No-go fMRI task.

270

A. Rahdar, A. Galván / NeuroImage 92 (2014) 267–273

The FSL FEAT package was used for statistical analysis. One general linear model (GLM) was defined for the GNG task, which included multiple regressors for each event type: successful go trials, unsuccessful go trials, successful no-go trials, and unsuccessful no-go trials (false alarms). Events were modeled with the onset time of each trial and duration of 1 s, and with a canonical (double-gamma) HRF. The rest periods and jittered inter-trial intervals were not explicitly modeled and therefore served as an implicit baseline. The following contrasts were generated: go N baseline and go N no-go to examine “hits”, nogo N baseline and no-go N go to examine “inhibition” and false alarms N correct go to examine “false alarms”. Temporal derivatives and motion parameters were included as covariates of no interest. A 2 (stress state) × 2 (age group) repeated measures ANOVA was conducted at the group level using the FMRIB Local Analysis of Mixed Effects (FLAME1) module in FSL (Beckmann et al., 2003). Z (Gaussianised T) statistic images were thresholded using clusters determined by Z N 2.3 and a (whole-brain corrected) cluster significance threshold of p b 0.05 using Gaussian random field theory (Poline et al., 1997). Tests were corrected for family-wise errors (FWE). Anatomical localization within each cluster was obtained by searching within maximum likelihood regions from the FSL Harvard–Oxford probabilistic atlas. All fMRI data shown were cluster-corrected for multiple comparison at z = 2.3, p b 0.05. Results Stress data To examine ratings of stress at the scans, a 2 (stress state) × 2 (age group) repeated measures ANOVA was conducted, controlling for gender. There was a main effect of stress state on self-reported stress ratings [F(1,43) = 7.5, p = 0.009] such that participants had higher stress ratings on scans conducted on their self-reported high-stress (M = 4.2, SD = 1.53) versus low-stress (M = 2.5, SD = 1.42) state. There was no significant main effect of age [F(1,43) = .91, p = .35] or age × stress state interaction [F(1,43) = 2.3, p = .13]. There were no significant associations between stress ratings and behavioral performance or stress ratings and brain activation. There was also a main effect of stress state on cortisol levels collected at the scanner [F(1,43) = 4.34, p = 0.04], such that there were higher cortisol levels on high-stress state (M = 11.3, SD = 8.11) versus lowstress state (M = 9.05, SD = 6.22) scans but no significant main effect of age [F(1,43) = 2.1, p = .15] or age × stress level interaction. There were no significant associations between cortisol and behavioral performance or cortisol and brain activation. Behavioral results 2 (stress state) × 2 (age group) repeated measures ANOVAs, with gender entered as a covariate, on behavioral performance revealed significant main effects of stress state on successful go trials [F(1,43) = 6.82, p b 0.02], inhibition [F(1,43) = 6.24, p = .01], misses [F(1,43) = 11.23, p = .002], and false alarm trials [F(1,43) = 6.24, p = .01]. There was a significant age × stress interaction on successful inhibition

[F(1,43) = 10.74, p = 0.002] and false alarms [F(1,43) = 10.74, p = 0.002] (Fig. 3). Post hoc tests revealed worse performance in adolescents under the high-stress state (Table 1 for behavioral means). 2 (stress state) × 2 (age group) repeated measures ANOVAs on reaction time revealed a significant main effect of age [F(1,43) = 7.17, p = .01] on reaction time during correct go trials such that adults were faster (M =454.72, SD = 7.07) than adolescents (M = 471.6, SD = 25.93). There were no significant main effects of stress or age × stress interaction on reaction time. D' prime analyses revealed a main effect of stress state on response bias [F(1,43) = 4.82, p = 0.03], such that there was worse performance under high stress [M = 2.07, SD = 0.76)] versus low stress [M = 2.34, SD = 1.04)]. There were no age effects or interactions with age. Analyses with age as a continuous variable revealed a significant negative association between age and reaction time on correct go trials (r = −.38, p = 0.01), a trend towards a positive association between age and response inhibition under high stress state (r = .3, p = 0.06) and a trend towards a negative association between age and false alarms under high stress state (r = −.28, p = 0.06). fMRI results First, the main effects of hits and inhibition (no-go) trials on neural activation were examined, controlling for gender. The omnibus GLM analysis of the imaging data identified activation in the medial frontal cortex during go responses compared to baseline and in the cingulate, bilateral caudate, and left putamen during inhibition (no-go) trials compared to baseline. The omnibus GLM analysis for the no-go N go contrast revealed a main effect of activation in the ventromedial PFC, precuneus, cingulate gyrus, bilateral accumbens, precentral gyrus and occipital cortex (Table 2). The omnibus GLM analysis for the go N no-go contrast identified activation in the ventromedial PFC, cingulate, precentral gyrus, right hippocampus, and precuneus. An analysis of the false alarms N correct go trials revealed activation in the cerebellum and angular gyrus. All coordinates are listed in Table 2. There was a main effect of stress on activation during the go N baseline contrast, in posterior cingulate (PCC) (−6, −50, 8, 3.39, 521) and medial PFC (−8, 38, 14, 4.04, 447). Post hoc tests revealed greater activation to high stress versus low stress in PCC (Mhigh stress = −0.03, SD = .16; Mlow stress = − 0.16, SD = .23) and in mPFC (Mhigh stress = .06, SD = .11; Mlow stress = −0.04, SD = .15). There was also a main effect of stress on activation during the no-go N baseline contrast in medial PFC (−8, 48, 18); post hoc tests revealed greater activation during high stress (Mhigh stress = .16, SD = .18) versus low stress (Mlow stress = −0.01, SD = .17). There was a main effect of age on insula activation (38, −12, 16) in the false alarm N go contrast, with adults (M = .09, SD = .05) showing greater activation than adolescents (M = .01, SD = .16) (Table 3). A 2 (stress state) × 2 (age group) ANOVA, controlling for gender, revealed a significant interaction in the right DLPFC (22, 40, 40; z = 4.34), extending into middle frontal gyrus (26, 18, 44, z = 3.36) and superior frontal gyrus (26, 20, 58, z = 3.81) during inhibition (no-go N go contrast) [F(1,43) = 4.93, p = 0.03] (Fig. 4A). To decompose the interaction, we extracted the beta values, which revealed that adolescents showed a trend [t(22) = 2.04, p = 0.055] towards significantly less DLPFC activation on high-stress state (M = − .10, SD = .13) versus

Fig. 3. Behavioral performance. An age × stress interaction during inhibition (A) and false alarms (B).

A. Rahdar, A. Galván / NeuroImage 92 (2014) 267–273 Table 3 Main effects of stress and age group on neural activation.

Table 1 Behavioral means.

Trial type a

Hits Inhibitiona Missesa False alarma a b c

271

High stress (% M, SD)

Low stress (%, M, SD)

Contrast

Anatomical region

R/L

x

Adults

Adolescents

Adults

Adolescents

76.88 (6.23) 90.12b (5.93) 23.1 (6.23) 9.87c (5.93)

75.47 (8.67) 83.35b (14.22) 24.52 (8.67) 15.05c (14.02)

78.5 (7.39) 90.52 (6.64) 19.7 (9.36) 9.47 (6.64)

78.5 (9.8) 88.43 (11.12) 20.36 (11.38) 11.56 (11.12)

Stress Go N baseline Posterior cingulate No-go N baseline

Medial PFC L Medial PFC

L −6 L

−8 −50 −8

38 8 48

Insula

R

38

−12

Significant main effects of stress. Significant interaction of stress × age on correct inhibition. Significant interaction of stress × age on false alarms.

low-stress state (M = −.03, SD = .19) scans; t-tests revealed a significant difference in DLPFC activation between high (M = .005, SD = .14) and low (M = − .09, SD = .12) stress scans in adults [t(22) = 4.32, p b 0.01]. We conducted an additional post-hoc test to ensure statistical independence using a DLPFC 10-mm sphere ROI (x = 30, y = 34, z = 38) that was identified in a previous study from our laboratory in which the Go/No-go task was used to test response inhibition (Telzer et al., 2013). Results from this analysis revealed similar findings for adults but not for adolescents: adults showed a significant difference in DLPFC activation between high (M = .17, SD = .08) and low (M = .11, SD = .03) stress scans [t(22) = 5.3, p b 0.01] (Fig. 4B). There was a marginal difference in DLPFC activation in adolescents for the high-stress state (M = .07, SD = .19) and low-stress state (M = .10, SD = .27) scans [t(22) = 2.14, p = .057]. There were no other significant interactions with age in any other contrasts. Correlation analyses aimed to determine whether neural activation on either stress state was associated with task performance did not yield any significant findings. Discussion The goal of this study was to examine whether adolescents and adults are differentially susceptible to the cognitive and neurobiological effects of daily stress. Understanding the effects of daily stress on the developing human brain is critical because the neural regions that subserve cognitive control are not only particularly vulnerable targets of stress, but because they undergo significant developmental maturation during adolescence (Eiland and Romeo, 2013; Mika et al., 2012; Romeo Table 2 Main effects of task on neural activation. Contrast

Anatomical region

R/L

x

y

z

z-stat

Go N baseline

Medial PFC Putamen Hippocampus Cerebellum Occipital cortex Cingulate Caudate

L L L

−8 −30 −26 0 −54 0 12 −8 −30 −8 4 −4 28 −4 4 0 6 −6 8 −26 −36 40 −24 56

62 0 −16 −44 −66 30 12 6 0 48 −12 −20 −12 −64 34 −62 −12 12 12 −26 −68 36 −68 −48

12 −2 −18 −28 34 2 8 2 −4 −12 36 64 −20 22 0 20 36 −8 −8 48 18 18 −36 42

4.01 3.28 3.77 3.46 3.39 3.19 3.9 3.32 3.17 3.55 3.11 3.41 3.99 3.44 3.12 3.64 4.4 3.54 3.12 4.43 4.44 3.9 4.93 3.93

No-go N baseline

Go N No-go

No-go N go

False alarms N correct Go

L R/L

Putamen VMPFC Cingulate Precentral gyrus Hippocampus Precuneus VMPFC Precuneus Cingulate Accumbens

L L R L R L R

Precentral gyrus Occipital cortex DLPFC Cerebellum Angular gyrus

L L R L R

R R/L

Age group False alarmsN Correct Go

y

z

z-stat

14 3.39 18

4.04

16

3.09

3.74

and McEwen, 2006), a time of considerable stress reactivity (Gunnar et al., 2009b; Stroud et al., 2009). Using a novel combination of EMA via text messaging and neuroimaging techniques, we observed that a greater perceived daily stress impaired response inhibition in both adolescents and adults, but that the effect was stronger in adolescents. Furthermore, these behavioral results were paralleled by significantly decreased engagement of the DLPFC under a high-stress state in adolescents versus adults, the latter of whom showed significantly greater activation under high relative to low stress states. Previous functional imaging studies in adult (e.g., Blasi et al., 2006; MacDonald et al., 2000; Miller and Cohen, 2001 for review) and developmental (Bunge et al., 2002; Casey et al., 1997; Rubia et al., 2006) populations have implicated the DLPFC in a broad cognitive control context, and more specifically during response inhibition in Go/No-go tasks (Bunge et al., 2002; Mostofsky et al., 2003; Steel et al., 2013). Research suggests that the DLPFC supports inhibition by inhibiting motor responses from motor and parietal cortices (Fassbender et al., 2006), by maintaining task-relevant attention (Desimone and Duncan, 1995) and by representing task set and instructions, all of which are critical to Go/No-go tasks (Courtney, 2004). It was thus unsurprising that the response inhibition task in this study elicited robust DLPFC activation in both age groups. However, the differential engagement of this region based on stress arousal revealed developmentally meaningful findings that help explain the observed behavior. The motivation for this study was based on three previously observed phenomena: stress impairs cognition in adults, stress sensitivity increases during adolescence, and neural systems that support cognition are in flux during adolescence. Interestingly, our data suggest that whereas adults exhibit greater neural activation during subjectively high stress, adolescents show no difference in neural sensitivity during low and high stress periods. The robust differences in adults but not adolescents may reflect greater experience in adults regulating cognitive control by recruiting additional neural resources during times of stress. Clearly this speculation is beyond the scope of this paper but a longitudinal study examining these effects would help address this possibility. The divergent patterns of activation had consequential effects on behavior, such that response inhibition deteriorated in adolescents relative to adults in a high-stress state. Previous work has shown that adolescents exhibit greater stress reactivity than adults to the same stressor (Dahl and Gunnar, 2009) so perhaps the decreased PFC engagement in teens versus adults in this study was influenced by overwhelming sensitivity to stress. However, we did not find greater cortisol levels or self-reported stress ratings in adolescents versus adults, which may be because all participants reported on their own naturalistic stress states, whereas previous studies have used the same laboratory stressor to measure stress reactivity across groups (Gunnar et al., 2009a). Furthermore, we did not observe significant associations between brain activation and task performance; this surprising finding may be due to insufficient statistical power and/or the rather limited variability in task performance. One further observation worth noting is that the main effects of stress were found primarily within the default mode network (DMN). Although little is known about the effect of stress on this network, one recent study found that participants who had just completed a period of prolonged stress (participants who had

272

A. Rahdar, A. Galván / NeuroImage 92 (2014) 267–273

Fig. 4. Neural effects of stress. A) An age × stress interaction yielded activation in the right dorsolateral prefrontal cortex. B) The interaction described in panel A is plotted to illustrate the significantly decreased DLPFC activation in adolescents relative to adults under high stress.

recently finished a long preparation period for a medical residence selection exam) displayed greater activation of the DMN than controls (Soares et al., 2013). Clearly, much more work that specifically targets the effects of stress on the DMN are warranted, particularly in developing population. Our study has several noteworthy strengths. First, we scanned participants under two stress states: once on a day when they reported high levels of stress and once on a day when they reported a low level of stress. This approach is a novel methodological advance in the developmental cognitive neuroscience literature as the assessments were based on naturalistic, daily experiences of adolescents, were not based on labinduced manipulations, and took advantage of teenagers' increasing use of communication technology to engage them in a study. Second, these data are particularly noteworthy because they were collected in the relatively “cold” and non-arousing nature of the laboratory. We speculate that these effects are even more pronounced under the arousing conditions of real-life in which adolescents experience stress. However, a few factors should be taken under consideration. First, we did not measure chronic stress in this sample but it is possible that the observed effects may be confounded with individual participant's stress history. Second, the relative difference in DLPFC activation between high- and low-stress states in adolescents was rather marginal, suggesting that the sample size may have been too small to detect more robust differences and that this finding warrants replication. In conclusion, we took a comprehensive and naturalistic approach to understand how daily stress impacts adolescent cognition and brain function. By using adolescents' own stress as the “manipulation” we were able to carefully determine how perceived stressors specific to the individual impact basic cognition. Importantly, we filled a gap in the literature by finding that the adolescent brain is particularly vulnerable to the effects of daily stress relative to an adult comparison group, an effect that has significant consequences for behavior. Acknowledgments We thank participants, anonymous reviewers, and the Galván Lab. This work was supported by NSF BCS0963750 to AG. Conflict of interest The authors have no conflicts of interest to report. References Almeida, D., Kessler, R., 1998. Everyday stressors and gender differences in daily distress. J. Pers. Soc. Psychol. 75, 670–680. Almeida, D., McGonagle, K., King, H., 2009. Assessing daily stress processes in social surveys by combining stressor exposure and salivary cortisol. Biodemography Soc. Biol. 55, 219–237. Arnett, J., 2002. The psychology of globalization. Am. Psychol. 57, 774–783.

Baum, A., Polsusnzy, D.M., 1999. Health psychology: mapping biobehavioral contributions to health and illness. Annual Review of Psychology 50, 137–163. Beckmann, C.F., Jenkinson, M., Smith, S.M., 2003. General multi-level linear modelling for group analysis in FMRI. Neuroimage 20, 1052–1063. Blasi, G., Goldberg, T.E., Weickert, T., Das, S., Kohn, P., Zoltick, B., Bertolino, A., Callicott, J.H., Weinberger, D.R., Mattay, V.S., 2006. Brain regions underlying response inhibition and interference monitoring and suppression. European Journal of Neuroscience 23 (6), 1658–1664. Bolger, N., Davis, A., Rafaeli, E., 2003. Diary methods: capturing life as it is lived. Annu. Rev. Psychol. 54, 579–616. Bondi, C.O., Rodriguez, G., Gould, G.G., Frazer, A., Morilak, D.A., 2008. Chronic unpredictable stress induces a cognitive deficit and anxiety-like behavior in rats that is prevented by chronic antidepressant drug treatment. Neuropsychopharmacology 33, 320–331. Bunge, S.A., Dudukovic, N.M., Thomason, M.E., Vaidya, C.J., Gabrieli, J.D., 2002. Immature frontal lobe contributions to cognitive control in children: evidence from fMRI. Neuron 33, 301–311. Bynner, J., 2000. Social change and the sequencing of developmental transitions. In: Crockett, L., Silbereisen, R. (Eds.), Negotiating Adolescence in Times of Social Change. Cambridge University Press, Cambridge, MA, pp. 89–103. Casey, B.J., Trainor, R.J., Orendi, J.L., Shuber, A.B., Nystrom, L.E., Giedd, J.N., Castellanos, F.X., Haxby, J.V., Noll, D.C., Cohen, J.D., Forman, S.D., Dahl, R.E., Rapoport, J.L., 1997. A developmental functional MRI study of prefrontal activation during performance of a Go–No-Go Task. J. Cogn. Neurosci. 9, 835–847. Casey, B., Galván, A., Hare, T., 2005. Changes in cerebral functional organization during cognitive development. Curr. Opin. Neurobiol. 15, 239–244. Casey, B., Getz, S., Galván, A., 2008. The adolescent brain. Dev. Rev. 28, 62–77. Chiron, C., Raynaud, C., Maziere, B., Zilbovicus, M., Laflamme, L., Masure, M.-C., Dulac, O., Bourguignon, M., Syrota, A., 1992. Changes in regional cerebral blood flow during brain maturation in children and adolescents. Int. J. Nucl. Med. 336, 696–703. Chugani, H., Phelps, M., Mazziotta, J., 1987. Positron emission tomography study of human brain functional development. Ann. Neurol. 22, 487–497. Courtney, S.M., 2004. Attention and cognitive control as emergent properties of information representation in working memory. Cogn. Affect. Behav. Neurosci. 4, 501–516. Dahl, R.E., Gunnar, M., 2009. Heightened stress responsiveness and emotional reactivity during pubertal maturation: implications for psychopathology. Dev. Psychopathol. 21, 1–6. Dedovic, K., Rexroth, M., Wolff, E., Duchesne, A., Scherling, C., Beaudry, T., Lue, S., Lord, C., Engert, V., Pruessner, J., 2009. Neural correlates of processing stressful information: an event-related fMRI study. Brain Res. 1293, 49–60. Desimone, R., Duncan, J., 1995. Neural mechanisms of selective visual attention. Annu Rev Neurosci 18, 193–222. Dias-Ferreira, E., Sousa, J., Melo, I., Morgado, P., Mesquita, A., Cerqueira, J., Costa, R., Sousa, N., 2009. Chronic stress causes frotostriatal reorganization and affects decisionmaking. Science 325, 621–625. Duckworth, A., Kim, B., Tsukayama, E., 2012. Life stress impairs self-control in early adolescence. Front. Psychol. 3, 608–634. Eccles, J.S., Midgley, C., Wigfield, C., Buchanan, C., Reuman, D., Flanagan, C., Iver, D.M., 1993. Development during adolescence: the impact of stage-environment fit on young adolescents' experiences in schools and families. Am. Psychol. 48, 90–101. Eiland, L., Romeo, R., 2013. Stress and the developing adolescent brain. Neuroscience 249, 162–171. Eisenberger, N.I., Taylor, S.E., Gable, S.L., Hilmert, C., Lieberman, M., 2007. Neural pathways link social support to attenuated neuroendocrine stress responses. NeuroImage 35, 1601–1612. Fassbender, C., Simoes-Franklin, C., Murphy, K., Hester, R., Meaney, J., Robertson, I.H., et al., 2006. The role of a right fronto-parietal network in cognitive control. J. Psychophysiol. 20, 286–296. Figner, B., Mackinlay, R., Wilkening, F., Weber, E., 2009. Affective and deliberative processes in risky choice: age differences in risk taking in the Columbia Card Task. J. Exp. Psychol. Learn. Mem. Cogn. 35, 709–730. Fuster, J., 2001. The prefrontal cortex—an update: time is of the essence. Neuron 30, 319–333.

A. Rahdar, A. Galván / NeuroImage 92 (2014) 267–273 Galván, A., McGlennen, K., 2012. Daily stress increases risky decision-making in adolescents: a preliminary study. Dev. Psychobiol. 54, 433–440. Galván, A., Hare, T., Parra, C., Penn, J., Voss, H., Glover, G., Casey, B., 2006. Earlier development of the accumbens relative to orbitofrontal cortex might underlie risk-taking behavior in adolescents. J. Neurosci. 26, 6885–6892. Gogtay, N., Giedd, J., Lusk, L., Hayashi, K., Greenstein, D., Vaituzis, A., Herman, D., Clasen, L., Toga, A.W., Rapoport, J., Thompson, P., 2004. Dynamic mapping of human cortical development during childhood through early adulthood. Proc. Natl. Acad. Sci. 101, 8174–8179. Gunnar, M., Talge, N., Herrera, A., 2009a. Stressor paradigms in developmental studies: what does and does not work to produce mean increases in salivary cortisol. Psychoneuroendocrinology 34, 953–967. Gunnar, M., Wewerka, S., Frenn, K., Long, J., Griggs, C., 2009b. Developmental changes in HPA activity over the transition to adolescence: normative changes and associations with puberty. Dev. Psychopathol. 21, 69–85. Hand, L., Furman, W., 2009. Rewards and costs in adolescent other-sex friendships: comparison to same-sex friendships and romantic relationships. Soc. Dev. 18, 270–287. Hellhammer, D.H., Wust, S., Kudielka, B.M., 2009. Salivary cortisol as a biomarker in stress research. Psychoneuroendocrinology 34 (2), 163–171. Hennessy, J.W., Cohen, M.E., Rosen, A.J., 1973. Adrenocortical influences upon the extinction of an appetitive runway response. Physiol Behav 11, 767–770. Janis, I., 1993. Decision-making under stress, In: Goldberger, L., Breznitz, S. (Eds.), Handbook of Stress: Theoretical and Clinical Aspects, 2nd ed. , pp. 56–74. Keinan, G., 1987. Decision making under stress: scanning of alternative under controllable and uncontrollable threats. J. Pers. Soc. Psychol. 52, 639–644. Kuttler, A., LaGreca, A., 2004. Linkages among adolescent girls' romantic relationships, best friendships, and peer networks. J. Adolesc. 27, 395–414. Lapiz-Bluhm, M.D., Soto-Piña, A.E., Hensler, J.G., Morilak, D.A., 2009. Chronic intermittent cold stress and serotonin depletion induce deficits of reversal learning in an attentional set-shifting test in rats. Psychopharmacology 202, 329–341. Lewis, D., 1970. Development of the prefrontal cortex during adolescence: insights into vulnerable neural circuits in schizophrenia. Neuropsychopharmacology 16, 386–398. Liston, C., Watts, R., Tottenham, N., Davidson, M., Niogi, S., Ulug, A., Casey, B., 2006. Frontostriatal microstructure modulates efficient recruitment of cognitive control. Cereb. Cortex 16, 553–560. Liston, C., McEwen, B.S., Casey, B., 2009. Psychosocial stress reversibly disrupts prefrontal processing and attentional control. Proc. Natl. Acad. Sci. U. S. A. 106, 912–917. Lupien, S., McEwen, B.S., Gunnar, M., Heim, C., 2009. Effects of stress throughout the lifespan on the brain, behavior and cognition. Nat. Rev. Neurosci. 10, 434–445. Lyons, D., Lopez, J., Yang, C., Schatzberg, A., 2000. Stress-level cortisol treatment impairs inhibitory control of behavior in monkeys. J. Neurosci. 20, 7816–7821. MacDonald 3rd, A.W., Cohen, J.D., Stenger, V.A., Carter, C.S., 2000. Dissociating the role of the dorsolateral prefrontal and anterior cingulate cortex in cognitive control. Science 288 (5472), 1835–1838. Mather, M., Lighthall, N., 2012. Both risk and reward are processed differently in decisions made under stress. Curr. Dir. Psychol. Sci. 21, 36–41. McEwen, B.S., Morrison, J.H., 2013. The brain on stress: vulnerability and plasticity of the prefrontal cortex over the life course. Neuron 79, 16–29. Micco, D.J., McEwen, B.S., Shein, W., 1979. Modulation of behavioral inhibition in appetitive extinction following manipulation of adrenal steroids in rats: implications for the involvement of the hippocampus. J Comp Physiol Psychol 93, 323–329. Mika, A., Mazur, G., Hoffman, A., Talboom, J., Nelson, B., Sanabria, F., Conrad, C.D., 2012. Chronic stress impairs prefrontal cortex-dependent response inhibition and spatial working memory. Behav. Neurosci. 126, 605–619. Miller, E.K., Cohen, J.D., 2001. An integrative theory of prefrontal cortex function. Annu. Rev. Neurosci. 24, 167–202. Miller, L.H., Smith, A.D., Rothstein, L., 1994. The Stress Solution: An Action Plan to Manage the Stress in your Life. Pocket Books, New York. Mostofsky, S.H., Schafer, J.G., Abrams, M.T., Goldberg, M.C., Flower, A.A., Boyce, A., et al., 2003. fMRI evidence that the neural basis of response inhibition is task-dependent. Brain Research. Cogn. Brain Res. 17, 419–430. Ossewaarde, L., Qin, S., Van Marle, H.J.F., van Wingen, G.A., Fernandez, G., Hermans, E.J., 2011. Stress-induced reduction in reward-related prefrontal cortex function. NeuroImage 55, 345–352.

273

Persike, M., Seiffge-Krenke, I., 2012. Competence in coping with stress in adolescents from three regions of the world. J. Youth Adolesc. 41, 863–879. Poline, J.-B., Worsley, K.J., Evans, A., Friston, K., 1997. Combining spatial extent and peak intensity to test for activations in functional imaging. NeuroImage 5, 83–96. Pollak, S.D., 2005. Early adversity and mechanisms of plasticity: integrating affective neuroscience with developmental approaches to psychopathology. Dev. Psychopathol. 17, 735–752. Porcelli, A., Delgado, M.R., 2009. Acute stress modulates risk taking in financial decisionmaking. Psychol. Sci. 20, 278–283. Porcelli, A., Lewis, A.H., Delgado, M.R., 2012. Acute stress influences neural circuits of reward processing. Front. Neurosci. 6, 2–9. Preston, S., Buchanan, T., Stansfield, R., Bechara, A., 2007. Effects of anticipatory stress on decision making in a gambling task. Behav. Neurosci. 121, 257–262. Radley, J., Sisti, H., Hao, J., Rocher, A., McCall, T., Hof, P., McEwen, B.S., 2004. Chronic behavioral stress induces apical dendritic reorganization. Neuroscience 125, 1–6. Radley, J., Rocher, A., Miller, M., Janssen, W., Liston, C., Hof, P., McEwen, B.S., Morrison, J., 2006. Repeated stress induces dendritic spine loss in the rat medial prefrontal cortex. Cereb. Cortex 16, 313–320. Romeo, R.D., McEwen, B.S., 2006. Stress and the adolescent brain. Ann. N. Y. Acad. Sci. 1094, 202–214. Roozendaal, B., 2002. Stress and memory: opposing effects of glucocorticoids on memory consolidation and memory retrieval. Neurobiol. Learn. Mem. 78, 578–595. Rubia, K., Smith, A.B., Woolley, J., Nosarti, C., Heyman, I., Taylor, E., Brammer, M., 2006. Progressive increase of frontostriatal brain activation from childhood to adulthood during event-related tasks of cognitive control. Hum. Brain Mapp. 27, 973–993. Sandi, C., 2013. Stress and cognition. Cogn. Sci. 4, 245–261. Schlegel, A., 2001. The global spread of adolescent culture. In: Crocket, L., Silbereisen, R. (Eds.), Negotiating Adolescence in Times of Social Change. Cambridge University Press, New York, NY, pp. 63–86. Scholz, U., LaMarca, R., UM, N., Aberle, I., Ehlert, U., Hornung, R., Martin, M., Kliegel, M., 2009. Go no-go performance under psychological stress: beneficial effects of implementation intention. Neurobiol. Learn. Mem. 91, 89–92. Shaw, P., Kabani, N., Lerch, J., Eckstrand, K., Lenroot, R., Gotgtay, N., Greenstein, D., Clasen, L., Evans, A., Rapoport, J., Giedd, J.N., Wise, S., 2008. Neurodevelopmental trajectories of the human cerebral cortex. J. Neurosci. 28, 3586–3594. Soares, J.M., Sampaio, A., Ferreira, L.M., Santos, N.C., Marques, P., Marques, F., Palha, J.A., Cerqueira, J.J., Sousa, N., 2013. Stress impact on resting state brain networks. PLoS One 8, e66500. Sowell, E., Thompson, P., Holmes, C., Jernigan, T., Toga, A., 1999. In vivo evidence for postadolescent brain maturation in frontal and striatal regions. Nat. Neurosci. 2, 859–861. Spinelli, S., Chefer, S., Suomi, S., Higley, J., Barr, C., Stein, E., 2009. Early-life stress induces long-term morphologic changes in primate brain. Arch. Gen. Psychiatry 66, 658–665. Steel, V.R., Aharoni, E., Munro, G.E., Calhoun, V.D., Nyalakanti, P., Stevens, M.C., Pearlson, G., Kiehl, K.A., 2013. A large scale (N = 102) functional neuroimaging study of response inhibition in a Go/NoGo task. Behav. Brain Res. 256, 529–536. Stetler, C.A., Miller, G.E., 2008. Social integration of daily activities and cortisol secretion: a laboratory based manipulation. Journal of Behavioral Medicine 31 (3), 249–257. Strahler, J., Mueller, A., Rosenloecher, F., Kirschbaum, C., Rohleder, N., 2010. Salivary alpha-amylase stress reactivity across different age groups. Psychophysiology 47 (3), 587–595. Stroud, L.R., Foster, E., Papandonatos, G.D., Handwerger, K., Granger, D.A., Kivlighan, K.T., Niaura, R., 2009. Stress response and the adolescent transition: performance versus peer rejection stressors. Dev. Psychopathol. 21, 47–68. Telzer, E.H., Fuligni, A.J., Lieberman, M.D., Galván, A., 2013. The effects of poor quality sleep on brain function and risk taking in adolescence. NeuroImage 71, 275–283. Treadway, M.T., Buckholtz, J.W., Zald, D.H., 2013. Perceived stress predicts altered reward and loss feedback processing in medial prefrontal cortex. Front. Hum. Neurosci. 7, 2–9. Wolf, O., 2006. Effects of stress hormones on the structure and function of the human brain. Expert. Rev. Endocrinol. Metab. 1.