Preschool- and School-Age Irritability Predict Reward-Related Brain Function

Accepted Manuscript Preschool and School-Age Irritability Predict Reward-Related Brain Function Lea Dougherty, PhD, Karen T.G. Schwartz, MS, Maria Kryza-Lacombe, MA, Jill Weisberg, PhD, Philip A. Spechler, MA, Jillian Lee Wiggins, PhD PII:

S0890-8567(18)30162-X

DOI:

10.1016/j.jaac.2018.03.012

Reference:

JAAC 2089

To appear in:

Journal of the American Academy of Child & Adolescent Psychiatry

Received Date: 12 October 2017 Revised Date:

25 February 2018

Accepted Date: 1 March 2018

Please cite this article as: Dougherty L, Schwartz KTG, Kryza-Lacombe M, Weisberg J, Spechler PA, Lee Wiggins J, Preschool and School-Age Irritability Predict Reward-Related Brain Function, Journal of the American Academy of Child & Adolescent Psychiatry (2018), doi: 10.1016/j.jaac.2018.03.012. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Preschool and School-Age Irritability Predict Reward-Related Brain Function RH = Irritability and Reward Lea Dougherty, PhD, Karen T. G. Schwartz, MS, Maria Kryza-Lacombe, MA, Jill Weisberg,

RI PT

PhD, Philip A. Spechler, MA, Jillian Lee Wiggins, PhD Clinical Guidance Supplemental Material

SC

Editorial

M AN U

Accepted April 4, 2018

Dr. Dougherty is with the Program in Neuroscience and Cognitive Science, University of Maryland, College Park. Mr. Spechler is with the University of Maryland, College Park. Drs. Wiggins and Weisberg, Mss. Schwartz and Kryzo-Lacome are with San Diego State University

TE D

and University of California, San Diego Joint Doctoral Program in Clinical Psychology.

This research was supported by the Maryland Neuroimaging Center Seed Grant Program (LRD),

EP

National Science Foundation in partnership with the University of Maryland Type: ADVANCE Program for Inclusive Excellence (LRD & Tracy Riggins), University of Maryland College of

AC C

Behavioral and Social Sciences Dean’s MRI Research Initiative RFP Program (LRD & Tracy Riggins), Behavioral and Social Sciences Dean’s Research Initiative (LRD), and the Research and Scholarship Award (LRD).

Dr. Wiggins served as the statistical expert for this research.

ACCEPTED MANUSCRIPT

We are grateful to all the participating families. We would like to recognize Sarah Blankenship, PhD, Victoria Smith, PhD, Marissa Kushner, PhD, Stephanie Merwin, PhD, and Katherine Leppert, MS, of the University of Maryland, College Park for assistance with data collection. We

Maryland, College Park, for assistance in data processing.

RI PT

also wish to thank Srikanth Padmala, PhD, and Mihai Sirbu, BS, with the University of

SC

Disclosure: The authors report no biomedical financial interests or potential conflicts of interest.

M AN U

Correspondence to Lea Dougherty, PhD, Department of Psychology, University of Maryland,

AC C

EP

TE D

College Park, MD 20742; email: [email protected].

ACCEPTED MANUSCRIPT

ABSTRACT Objective: Although chronic irritability in childhood is prevalent, impairing, and predictive of later maladjustment, its pathophysiology is largely unknown. Deficits in reward processing are

RI PT

hypothesized to play a role in irritability. The current study aimed to identify how the

developmental timing of irritability during preschool and school-age relates to reward-related brain function during school-age.

SC

Method: Children’s irritability was assessed during the preschool period (Wave 1; ages 3.0-5.9 years) and three years later (Wave 2; ages 5.9-9.6 years) using a clinical interview. At Wave 2,

M AN U

children (N=46; 28 females) performed a monetary incentive delay task in which they received rewards, if they successfully hit a target, or no reward regardless of performance, during fMRI acquisition.

Results: Children with more vs. less severe preschool irritability, controlling for concurrent

TE D

irritability, exhibited altered reward-related connectivity: right amygdala with insula and inferior parietal lobe as well as left ventral striatum with lingual gyrus, post-central gyrus, superior parietal lobe and culmen. Children with more vs. less severe concurrent irritability, controlling

EP

for preschool irritability, exhibited a similar pattern of altered connectivity between left and right amygdalae and superior frontal gyrus and between left ventral striatum and precuneus and

AC C

culmen. Neural differences associated with irritability were most evident between reward and no reward conditions when participants missed the target. Conclusion: Preschool irritability and concurrent irritability were uniquely associated with aberrant patterns of reward-related connectivity, highlighting the importance of developmental timing of irritability for brain function. Key words: irritability; reward; fMRI; connectivity; children.

ACCEPTED MANUSCRIPT

INTRODUCTION Irritability, defined as low frustration tolerance characterized by anger and temper outbursts, is a common and impairing symptom in youth. Irritability is one of the most frequent

RI PT

reasons for treatment referral and is present across multiple emotional and behavioral

disorders,1,2 including as the cardinal feature of DSM-5 disruptive mood dysregulation disorder. Chronic irritability in school-age children and adolescents predicts depressive and anxiety

SC

disorders, suicidality, and functional impairment in adulthood.1,3,4 Emerging findings also

support the significance of irritability in early childhood, demonstrating that preschool-age

M AN U

irritability predicts psychiatric disorders, functional impairment, and treatment use later in childhood, even after accounting for baseline psychopathology.5,6 Despite its prevalence and central role in developmental psychopathology, the pathophysiology of irritability is largely unknown.

TE D

Aberrant reward processing may be a pathway to chronic irritability in youth.7-10 Reward processing constitutes an interconnected neural network including the striatum, amygdala, anterior cingulate cortex, prefrontal cortex, and midbrain regions.11-13Although research is

EP

limited, some studies14-16 have demonstrated that youth with chronic irritability evidence a lower threshold for experiencing frustration when they fail to receive a reward (i.e., frustrative non-

AC C

reward7,8) compared to non-irritable youth. This observation may be due to difficulties in instrumental learning and cognitive flexibility (i.e., learning when to expect rewards and how to adjust to different reward contingencies).17,18 Prior studies have reported neural abnormalities, including decreased amygdala activation and aberrations in prefrontal recruitment, when irritable youths failed to receive a reward,14,15,19,20 and some findings indicate differences in striatal activation.14,19 However, no research has examined associations between irritability and reward-

ACCEPTED MANUSCRIPT

related network connectivity, despite that brain regions do not act in isolation. Connectivity studies addressing this important gap would provide insight into the neural circuitry underlying irritability, which could inform the development of mechanistic treatment targets.

RI PT

With the exception of Perlman et al.,19 all prior studies have been in adolescents, even though irritability in early childhood is common and impairing.5,6,21,22 Although trajectories of irritability across childhood demonstrate considerable variability,23 normatively, irritability

SC

decreases from preschool- to school-age,22,23 as children’s regulatory capacities to inhibit temper outbursts increase; thus, children who demonstrate steady or increasing irritability across

M AN U

childhood become more atypical compared to peers.23 Little work has been done, however, to examine neural correlates of irritability symptoms during preschool-age and early school-age. Identifying neural mechanisms is important because irritability during childhood is a potent predictor of later maladjustment8 and reward-related neural circuits undergo significant

Current study

TE D

development from early childhood into adulthood.24-26

This study sought to identify the unique effects of preschool irritability and concurrent

EP

irritability on reward-related neural processes during early school-age. Examining irritability across childhood may provide insight into how the developmental timing of irritability influences

AC C

the nature of brain function. Etiological pathways involving reward processing may be similar or unique depending on the developmental timing of the youth’s irritability. Moreover, previous research has demonstrated that preschool irritability predicts later psychological outcomes and functional impairment,5,6 supporting the importance of identifying irritability early and intervening; however, no prior study has tested whether preschool irritability predicts later childhood neural outcomes, which will be an important step toward identifying neural

ACCEPTED MANUSCRIPT

mechanisms of irritability. Investigating how brain function relates to preschool and school-age irritability will increase our understanding of the etiology of irritability and potentially lead to innovative treatments targeting brain mechanisms.

RI PT

The study’s focus is network connectivity, i.e., correlation of different brain regions’ activation as a function of reward/no-reward contexts, as brain regions form highly connected neural circuits. By adopting a network perspective, this study increases our understanding of how

SC

irritability relates to the functional organization of the brain during reward processing. Analyses focused on amygdala connectivity, as there have been findings in older irritable youth and adults

M AN U

pointing towards the involvement of amygdala recruitment in aberrant reward processing.14,20 Moreover, abnormalities in amygdala activation across a variety of tasks have been the most consistent finding in the irritability literature.7,8,27 At preschool and early school-age, it is expected that amygdala-prefrontal connectivity during reward processing will relate to irritability

TE D

symptom severity. The current study also examined striatal connectivity; the striatum plays a key role in reward processing,28 and abnormalities in striatal activation during reward processing have also been linked to youth irritability.14,19 Finally, consistent with previous studies,14,15,19,20

EP

we conducted traditional activation analyses, both whole-brain and regions-of-interest (ROI), to examine associations between irritability and reward-related activation in the amygdala and

AC C

ventral striatum. METHOD

Participants

Participants were recruited from a larger study investigating biological risk factors for

depression.29,30 Families were recruited via flyers and a commercial mailing list and oversampled for offspring with maternal depression; thus, the sample is enriched to achieve a wider range of

ACCEPTED MANUSCRIPT

childhood irritability symptoms.5,23 Study exclusion criteria consisted of child developmental/physical disability, non-fluent in English, and a lifetime history of psychotic or bipolar disorder in either biological parent. Data on child psychological symptoms were

RI PT

collected at two time points (Wave 1: 3-5 years; Wave 2: ~3 years post-Wave 1), and child

neuroimaging data were collected from a subset participating at Wave 2. Parents gave written

Review Board approved study procedures.

SC

informed consent, and minor participants (>7 years) gave assent. The University’s Institutional

At Wave 2, 64 children (ages 5.9-9.6 years), free of MRI contraindications, volunteered

M AN U

to participate in the neuroimaging assessment. Of the 64 volunteers, one child was not scanned due to claustrophobia, 17 were excluded due to issues with data collection (n=10 had incomplete scan data; n=3 had excessive head motion [see fMRI Data Preprocessing]; n=2 completed a different scan protocol; n=1 had inadequate neuroanatomical scan coverage; n=1 was missing

TE D

behavioral data), leaving a final sample of 46 usable datasets. No children were taking psychotropic medications. The 46 children whose neuroimaging datasets were included versus those without usable data did not differ on any demographic or clinical variable included in the

EP

study. There were no differences between the scanned subsample and those who were not scanned at Wave 2 on any demographic or clinical variable with one exception: the MRI

AC C

subsample had significantly higher irritability scores at Wave 2 (M=1.61, SD=1.61) than those who did not complete the MRI assessment at Wave 2 (M=0.94, SD=1.16, p=.023); there were no differences between the MRI subsample and children who only completed the Wave 1 assessment. See Table 1 for sample characteristics. Clinical assessments

ACCEPTED MANUSCRIPT

At Waves 1 and 2, primary caregivers were interviewed with the Preschool Age Psychiatric Assessment (PAPA31). A 6-item measure of chronic irritability symptoms was derived from the PAPA (Wave 1: ICC=.97; α=.71; Wave 2: ICC=.96; α=.64). See5,6 for a

RI PT

complete description of the scale. Items included: irritable mood, feelings of anger under minor provocation, displays of anger under minor provocation, feelings of frustration under minor provocation, episodes of temper without violence, and episodes of excessive temper, manifested

SC

by shouting, crying, or stomping, and/or involving violence/damage.

PAPA items were rated for intensity, frequency, and duration. The intensity rating

M AN U

indicates whether a symptom was absent/present and the extent to which it was intrusive, interfering, and generalizable across activities. A rating of two or higher on a two or three-point scale indicates that the symptom was present at a threshold level of intensity. Frequency items reflect the number of occurrences during the last three months. Following guidelines for chronic

TE D

irritability,1,32 items were coded as present if a child engaged in the behavior at least 45 times in the past three months. The duration criterion was coded as present if the child was rated as having at least a 30-minute duration on irritable mood, prone to frustration, annoyance or anger,

EP

or difficulty recovering from temper tantrums. The total irritability scale consisted of the sum of symptoms coded as present according to the intensity, frequency, and duration criteria. The

AC C

PAPA has been shown to have acceptable psychometric properties when used in children through age 8.33

Neuroimaging acquisition

Participating children underwent an fMRI scan using a 3T Siemens scanner equipped

with a 12-channel head coil. During task acquisition, blood oxygen level dependent (BOLD) images were acquired as 36 contiguous axial slices parallel to the AC-PC line, with whole brain

ACCEPTED MANUSCRIPT

coverage, using an echo planar single-shot gradient echo pulse sequence (matrix size=64x64, TR=2000ms, TE=25ms, flip angle=70°, FOV=192mm, voxel size=3x3x3mm, 438 images across all runs). Anatomical images (T1-weighted MPRAGE) were acquired at high resolution for

RI PT

anatomical localization and spatial normalization (176 1.0mm sagittal slices, flip angle=9°,

matrix size=256x256, FOV=250mm, voxel size=1x1x1mm). See30 for additional information. Reward processing task

SC

Neurological processes related to reward processing were assessed using a child-friendly monetary incentive delay task.13,30 We previously reported that this task is effective at eliciting

M AN U

neural responses of reward processing in this sample.30 During fMRI acquisition, children were given the opportunity to win stars by pressing a button to hit a target; stars translated into money earned (<$15). The task consisted of cue (2000ms) period, during which the child was informed of whether or not they would earn a reward (50% reward condition, 50% no-reward condition).

TE D

Across both conditions, there was a variable anticipation period in which they waited to hit the target (2500-5500ms) and feedback (1500ms) on whether they hit or missed the target. During reward conditions only, they were given feedback that hits resulted in monetary reward and

EP

misses resulted in no monetary reward. The task was projected onto a screen that participants viewed via a mirror attached to the head coil. Time was automatically adjusted (+/-50ms) based

AC C

on the participant’s performance during the task. Task runs spanned 4 minutes 52 seconds, and each child completed 3 runs with a total 60 trials across all runs (30 trials reward, 30 no-reward conditions). One child completed two of the three runs; all available data were included in the analyses. Because past findings in this sample30 were in the feedback period, the connectivity analyses reported below focused on the feedback period. Data analysis

ACCEPTED MANUSCRIPT

fMRI data preprocessing. Standard fMRI data preprocessing protocols were implemented, using Analysis of Functional NeuroImages (AFNI; https://afni.nimh.nih.gov/afni/) to perform slice-time correction, realignment of functional images, spatial smoothing at 4mm,

RI PT

and non-linear registration for spatial standardization to the Talairach template. TR pairs with frame-wise displacement exceeding 1mm were censored from participant-level analysis, and participants with mean framewise head displacement ≥.30mm or censoring of ≥35% of TRs were

SC

excluded from all analyses.

Generalized PPI. The current study used generalized psychophysiological interaction

M AN U

analysis (gPPI34) to calculate connectivity between brain areas during the feedback period of the reward task. gPPI calculates change in correlations between a seed region of interest and all other regions in each condition compared to implicit baseline. Generalized methods34 are advantageous as they allow for the evaluation of more than two task conditions in a single model. Given past

TE D

work on this task13,30 and prior fMRI work strongly implicating amygdala dysfunction in irritability,14,27 the current study focused on left and right amygdala as seeds for gPPI analyses. We also examined ventral striatum (i.e., nucleus accumbens) as the seed. Masks for seeds and

EP

other masks (see secondary analyses below) were created using the Talairach daemon atlas in AFNI35 (left amygdala=756mm3; right amygdala=972mm3; ventral striatum [nucleus

AC C

accumbens]=108mm3). The end product is a set of voxel-wise images that represent connectivity between the seed region and the rest of the brain in each condition (reward/hit, reward/miss, noreward/hit, no-reward/miss). Estimated head motion in x, y, z, roll, pitch, yaw directions and third-degree polynomials to model low-frequency drift were included in the model. Second level analyses. To identify the role that preschool and school-age irritability symptoms played in reward-related brain connectivity, second-level analyses were conducted

ACCEPTED MANUSCRIPT

with the gPPI images (reward/hit, reward/miss, no-reward/hit, no-reward/miss for each individual) using a whole-brain, group-level ANCOVA via AFNI’s 3dMVM program, including Wave 1 and 2 Irritability as quantitative, between-subjects variables, and Performance (hit, miss)

RI PT

and Condition (no reward, reward) as within-subject variables. Significant clusters in contrasts with Wave 1 Irritability thus represent brain function at Wave 2 predicted by preschool

irritability, above and beyond concurrent irritability. Contrasts with Wave 2 Irritability represent

SC

contributions of concurrent irritability to brain function at Wave 2 above and beyond preschool irritability. We report the Wave 1 Irritability (controlling for Wave 2) x Performance x Condition

M AN U

and Wave 2 Irritability (controlling for Wave 1) x Performance x Condition interactions, which examined whether irritability severity relates to brain connectivity, depending on whether participants successfully hit the target, and whether a reward was received. Where 3-way interactions were not significant, we report lower-order interactions (Irritability x Performance,

to irritability.

TE D

Irritability x Condition, Irritability main effect) to identify other connectivity disparities related

Results were corrected for multiple comparisons, employing a whole-brain corrected

EP

threshold (p<.05). The cluster threshold was calculated by 3dClustSim using the updated mixedmodel spatial autocorrelation function (-acf) and the NN1 bisided option; bisided allows both

AC C

positive and negative voxels above threshold to be clustered separately and both sets of clusters to be retained. The group mask used in the 3dClustSim calculation represented brain regions where 90% of participants had valid data. Model parameters were calculated by 3dFWHMx for each run separately, were averaged over runs for each participant, and then averaged across participants (left amygdala: .66, 3.02, 11.15; right amygdala: .66, 3.02, 11.19; left ventral striatum: .65, 3.03, 11.15; right ventral striatum: .65, 3.03, 11.16). The cluster extent threshold

ACCEPTED MANUSCRIPT

across all models was k≥28 voxels (756 mm3) with a conservative height threshold of p<.005, which is appropriate for event-related designs.36,37 This correction method addresses issues raised by Eklund and colleagues36 regarding cluster correction methods.37

RI PT

Secondary analyses. We conducted traditional activation analyses, both whole-brain and ROI, as described in prior work30. Whole brain analyses were corrected using the same method as described above. The secondary ROI analyses probed differences in irritability in the

SC

amygdala and ventral striatum, regions utilized in the gPPI analyses, and reported to show

altered responses in previous reward and irritability studies.38 Average ROI responses were

M AN U

extracted for each condition in each individual using ANOVA in SPSS. RESULTS

Irritability was not significantly associated with demographic characteristics. Consistent with large behavior studies of irritability,5,23 the association between Waves 1 and 2 irritability

low stability. Imaging results

TE D

was marginally significant in the scanned subsample (Spearman rho=.29, p=.05), demonstrating

EP

gPPI connectivity analysis.

Preschool irritability and school-age brain connectivity. Whole-brain corrected

AC C

analyses showed that Wave 1 irritability, controlling for Wave 2, predicted degree of right amygdala connectivity with insula and inferior parietal lobule, dependent on whether the target was hit and whether there was a potential reward (Wave 1 Irritability [controlling for Wave 2] x Performance x Condition; Table 2). Children with more severe irritability at Wave 1 exhibited altered patterns of connectivity, compared to children with less severe irritability, when hitting and missing the target during reward and no reward conditions (Figure 1). Differences in

ACCEPTED MANUSCRIPT

connectivity associated with preschool irritability were particularly evident between reward and no reward conditions when participants missed the target. The three-way interaction was not significant for the ventral striatum, but Wave 1

RI PT

Irritability (controlling for Wave 2) x Performance was significant in several clusters connected with the left ventral striatum, including lingual gyrus (Figure 1), postcentral gyrus, superior parietal lobule, and culmen (Table 2). Across all areas, and as illustrated for lingual gyrus

hit trials and less connectivity during miss trials.

SC

(Figure 1), higher levels of preschool irritability were associated with greater connectivity during

M AN U

Concurrent irritability and school-age brain connectivity. Whole-brain corrected analyses revealed that Wave 2 irritability, controlling for Wave 1, was associated with degree of left and right amygdala connectivity with superior frontal gyrus, depending on whether the target was hit and presence of potential reward (Wave 2 Irritability [controlling for Wave 1] x

TE D

Performance x Condition; Table 2). This three-way interaction with Wave 2 irritability was driven by similar patterns as the three-way interactions at Wave 1: children with greater vs. lesser concurrent irritability showed altered patterns of connectivity, particularly when they

EP

missed targets with and without rewards (Figure 2). Whole-brain corrected connectivity analyses with the left ventral striatum seed also

AC C

indicated a significant Wave 2 Irritability (controlling for Wave 1) x Performance x Condition interaction for left ventral striatum-right precuneus and -culmen connectivity (Table 2). Similar to the pattern found with amygdalae connectivity, the interaction for ventral striatal connectivity was driven by differences in connectivity between reward and no reward conditions when participants missed the target (Table 2; Figure 2).

ACCEPTED MANUSCRIPT

Additional Analyses. To investigate the impact of other variables, analyses were repeated separately with the following covariates: child age, maternal depression, child internalizing and externalizing symptoms at Waves 1 and 2, and time between Waves. All

RI PT

findings remained significant after including these covariates, supporting the impact of

irritability on brain functioning, independent of other constructs (see Supplement 1 and Tables S1-S3, available online).

SC

Secondary activation analyses. None of the clusters resulting from whole-brain or ROI activation analyses survived a conservative Type I error correction.

M AN U

DISCUSSION

This study investigated the unique effects of early and concurrent irritability on rewardrelated brain function in school-age children. We observed that children with more vs. less severe preschool-age irritability, controlling for concurrent irritability, exhibited aberrant

TE D

patterns of connectivity between right amygdala and insula and inferior parietal lobe depending on whether the target was hit/missed and presence of potential reward. In addition, preschool irritability was associated with greater connectivity between left ventral striatum and lingual

EP

gyrus, postcentral gyrus, superior parietal lobule and culmen when hitting vs. missing the target. Children with more vs. less severe concurrent irritability, controlling for preschool irritability,

AC C

exhibited similar aberrant patterns of left and right amygdalae connectivity with superior frontal gyrus and left ventral striatum with precuneus and culmen when hitting or missing the target during reward and no-reward conditions. Results held after controlling for maternal depression and child internalizing and externalizing scores, suggesting that links between irritability and brain function are not primarily driven by related constructs. These findings highlight how reward-related neural circuits observed in school-age children may be altered in children with

ACCEPTED MANUSCRIPT

increased concurrent irritability and in children with early preschool-age irritability, suggesting possible mechanisms underlying impairing mood dysregulation. Children with higher levels of preschool-age irritability demonstrated alterations in brain

RI PT

connectivity during reward processing compared to children with lower levels of early

irritability. Although children with lower levels of preschool irritability showed little to no

modulation of amygdala connectivity with insula and inferior parietal lobule between the various

SC

trial types, children with higher levels of preschool irritability evidenced decreased amygdala connectivity with these regions on miss vs. hit trials during reward conditions and increased

M AN U

connectivity on miss vs. hit trials during non-reward conditions. These differences in connectivity in relation to preschool irritability level were particularly evident between reward and no-reward conditions when participants missed the target. The aberrant connectivity patterns observed in children with greater preschool irritability during miss vs. hit trials as a function of

TE D

reward may reflect that irritable youth show inappropriate modulation following both reward attainment and non-attainment.7,8 Furthermore, although children with lower levels of preschool irritability did not show differences in left ventral striatum connectivity with lingual gyrus,

EP

postcentral gyrus, superior parietal lobule, and culmen between hits and misses, children with higher levels of preschool irritability evidenced greater connectivity between these regions

AC C

during hit trials and decreased connectivity between these regions during miss trials. This finding suggests that early irritability predicts aberrant connectivity in response to performance success and failure, regardless of reward, which may contribute to greater frustration and outbursts in response to failures in irritable youth.7,8 Concurrent irritability was similarly associated with altered patterns of connectivity after accounting for preschool irritability. Children with lower levels of school-age irritability

ACCEPTED MANUSCRIPT

evidenced greater connectivity between both left and right amygdalae and superior frontal gyrus on miss vs. hit trials during reward conditions and decreased connectivity between these regions on miss vs. hit trials during non-reward conditions. In contrast, children with greater school-age

RI PT

irritability showed the opposite pattern, and the differences were most pronounced on misses during reward trials. Amygdala and prefrontal cortex, including superior frontal gyrus, have been implicated in the neural circuity supporting reward processing, emotion processing, attention,

SC

and inhibition/cognitive control.7,8,10,39 Given research supports the coupling of amygdala and frontal regions in both bottom-up emotion generation systems and top-down emotion regulation

M AN U

systems,10 it is possible that abnormalities in their connectivity during reward processing may contribute to heightened experiences of frustrative non-reward, particularly as observed differences were most notable when children missed the target.

A similar altered pattern of connectivity between the left ventral striatum and right

TE D

precuneus and culmen was observed for children with greater concurrent irritability. Children with greater school-age irritability evidenced an opposite pattern of connectivity between these regions compared to children with lower levels of concurrent irritability on reward trials. While

EP

children with lower levels of school-age irritability demonstrated decreased connectivity between left ventral striatum and precuneus and culmen on miss vs. hit trials during reward, children with

AC C

greater school-age irritability demonstrated greater connectivity on miss vs. hit trials during reward. During non-reward trials, children with high and low irritability showed similar connectivity patterns on miss and hit trials. The striatum is a key brain region involved in the anticipation and receipt of both reward and punishment,28 and the precuneus is considered a hub in the default mode network, a network hypothesized to be involved in self-referential processing, affective decision making, and emotion regulation.40 Resting state functional

ACCEPTED MANUSCRIPT

connectivity studies have linked connectivity between these regions to depression,41,42 and a recent study43 reported a similarly aberrant pattern of ventral striatum-precuneus connectivity in depressed adults vs. controls during loss vs. reward trials. Taken together, these findings suggest

RI PT

that irritable youth evidence aberrant connectivity during reward processing, similar to what is observed in depressed adults, which may explain links between youth irritability and adult depression.1,3,4

SC

In contrast to previous findings examining irritability, primarily in adolescence,14,15,19,20 irritability in preschool-age and school-age was not associated with brain activation differences,

M AN U

when participants failed to receive a reward. These divergent findings may be due to differences in age (childhood vs. adolescence), particularly as this neural circuitry undergoes developmental changes across childhood and into adulthood.39,44,45 Furthermore, the tasks used in these studies varied widely, and reward and non-reward outcomes have not always been isolated. In addition,

TE D

all prior irritability and reward studies employed extreme group designs, typically comparing youth with clinical levels of severe irritability to healthy comparison children. Although our sample was enriched for risk, given that we did not recruit a clinical sample, we did not have

EP

many children with extremely high levels of irritability. It is possible that activation differences are most evident in severe cases of irritability only, whereas connectivity differences may be

AC C

more sensitive to individual differences in irritability using a dimensional approach. As this is the first study to examine reward-related functional connectivity and activation, this requires further investigation.

Our findings suggest that early and concurrent symptoms of chronic irritability are related

to unique reward-related neural circuits. In addition, results pose that irritability-connectivity associations involving the amygdala may be more lateralized in early childhood and “spread”

ACCEPTED MANUSCRIPT

bilaterally as children age. Moreover, these brain circuits mature considerably across childhood, resulting in improved regulatory capacities in executive functioning and emotion regulation.46,47 Thus, aberrations in this circuitry may contribute to poorer regulatory capacities to inhibit

RI PT

frustration in response to blocked rewards, which is particularly problematic as age-matched peers are developing strategies to inhibit frustration successfully. Lastly, establishing links

between childhood irritability and reward-related neural processes could explain links between

SC

childhood irritability and depression and externalizing psychopathology, which have also been linked to reward processing deficits. Thus, reward-related processing deficits may be a common

M AN U

underlying factor, perhaps increasing risk for each, as well as their co-occurrence. This study had several strengths. First, whereas prior studies typically focused on adolescents, our sample included children assessed from preschool-age to early school-age. Examining neural correlates of irritability at younger ages allows us to map the developmental

TE D

timing of impairing and chronic mood dysregulation and its mechanisms. Next, our study uniquely examined network connectivity as a function of reward context. A network approach examines the interconnectedness of brain regions and contributes to our understanding of the

EP

functional organization of brain activity in relation to irritability. This is important as activation analyses failed to detect effects that were evident in connectivity analyses. Lastly, unlike

AC C

previous studies, we used a dimensional construct of youth irritability, as the boundaries between clinically significant irritability and normative irritability continue to be investigated. The study also had limitations. First, although our sample size was comparable to

previous studies,14,15,19,20 our sample size was modest (n=46) and not sufficient to examine how the developmental trajectory of irritability relates to reward-related brain function. Second, the ventral striatum mask encompassed a small region (nucleus accumbens; k=4 voxels, 108mm3).

ACCEPTED MANUSCRIPT

Replication with higher resolution scans (e.g., using stronger magnetic fields, although 7T scans are not currently indicated in children due to their possible side effects) with individually traced nucleus accumbens will be necessary to confirm results. Lastly, although our study used a

RI PT

longitudinal design, MRI scans only took place at one assessment; therefore, we do not know whether these reward-related aberrations are present earlier in development. Ideally, future

research would scan children at the earlier time-point in childhood (~age 4); however, this would

SC

require a reward task that is developmentally appropriate for both 4- and 7-year-olds to compare brain activation over time and would additionally require overcoming challenges inherent in

M AN U

scanning young children (i.e., remaining still while completing a task).

Our findings support hypotheses that irritable youth show impairments in reward processing and point to how coordinated neural networks may be altered in youth with higher levels of early and concurrent irritability. Future work should incorporate a multi-method

TE D

assessment, involving assessments of multiple behavioral and neural facets of reward learning, as well connectivity at rest, to characterize reward-related mechanisms involved in irritability. This research has important clinical implications. Specifically, identifying the neural architecture of

EP

irritability may be used to characterize irritability and its associations with other forms of psychopathology, which could refine how we classify and treat mental health problems. Lastly,

AC C

this research may lead to earlier identification of at-risk individuals and inform the development of targeted mechanistic approaches to intervention.

ACCEPTED MANUSCRIPT

References 1. Brotman MA, Schmajuk M, Rich BA, et al. Prevalence, clinical correlates, and longitudinal course of severe mood dysregulation in children. Biol Psychiatry.

RI PT

2006;60:991-997.

2. Leibenluft E, Blair RJR, Charney DS, Pine DS. Irritability in pediatric mania and other childhood psychopathology. Ann NY Acad Sci. 2003;1008:201–218.

SC

3. Leibenluft E, Cohen P, Gorrindo T, Brook JS, Pine DS. Chronic versus episodic

irritability in youth: A community-based, longitudinal study of clinical and diagnostic

M AN U

associations. J Child and Adolesc Psychopharmacol. 2006;16:456-466. 4. Stringaris A, Cohen P, Pine DS, Leibenluft E. Adult outcomes of youth irritability: a 20year prospective community-based study. Am J Psychiatry. 2009;166:1048-1054. 5. Dougherty LR, Smith VC, Bufferd SJ, et al. Preschool irritability: Predictive associations

TE D

with psychiatric disorders and impairment at age six. J Am Acad Child Adolesc Psychiatry. 2013;52:1304-1313.

6. Dougherty LR, Smith VC, Bufferd SJ, Kessel E, Carlson G, Klein DN. Preschool

EP

irritability predicts child psychopathology, functional impairment, and service use at age nine years. J Child Psychol Psychiatry. 2015;56:999-1007.

AC C

7. Brotman MA, Kircanski K, Stringaris A, Pine D, Leibenluft E. Irritability in youths: a translational model. Am J Psychiatry. 2017;174:520-532.

8. Brotman MA, Kircanski K, Leibenluft E. Irritability in children and adolescents. Annu Rev Clin Psychol. 2017;13:317-341. 9. Meyers E, DeSerisy M, Roy AK. Disruptive mood dysregulation disorder: An RDoC perspective. J Affect Disord. 2017; 216:117-122.

ACCEPTED MANUSCRIPT

10. Zisner A, Beauchaine TP. Neural substrates of trait impulsivity, anhedonia, and irritability: Mechanisms of heterotypic comorbidity between externalizing disorders and unipolar depression. Dev Psychopathol. 2016;28(4pt1):1177-1208.

Neuropsychopharmacology. 2010;35:4-26.

RI PT

11. Haber SN, Knutson B. The reward circuit: Linking primate anatomy and human imaging.

12. Gaffrey MS, Barch DM, Bogdan R, Farris K, Petersen SE, Luby JL. Amygdala reward

severity. Biol Psychiatry. 2018;83(2):128-136

SC

reactivity mediates the association between preschool stress response and depression

M AN U

13. Helfinstein SM, Kirwan ML, Benson BE, et al. Validation of a child-friendly version of the monetary incentive delay task. Soc Cogn Affect Neurosci. 2013;8:720-726. 14. Deveney CM, Connolly ME, Haring CT, et al. Neural mechanisms of frustration in

chronically irritable children. Am J Psychiatry. 2013;170:1186-94.

TE D

15. Rich BA, Carver FW, Holroyd T, et al. Different neural pathways to negative affect in youth with pediatric bipolar disorder and severe mood dysregulation. J Psychiatr Res. 2011;45:1283-1294.

EP

16. Rich BA, Schmajuk M, Perez-Edgar KE, Fox NA, Pine DS, Leibenluft E. Different psychophysiological and behavioral responses elicited by frustration in pediatric bipolar

AC C

disorder and severe mood dysregulation. Am J Psychiatry. 2007;164:309-317.

17. Adleman NE, Kayser R, Dickstein D, Blair RJ, Pine D, Leibenluft E. Neural correlates of reversal learning in severe mood dysregulation and pediatric bipolar disorder. J Am Acad

Child Adolesc Psychiatry. 2011;50:1173-1185.

18. Dickstein DP, Nelson EE, McClure EB, et al. Cognitive flexibility in phenotypes of pediatric bipolar disorder. J Am Acad Child Adolesc Psychiatry. 2007;46:341-355.

ACCEPTED MANUSCRIPT

19. Perlman SB, Jones BM, Wakschlag LS, Axelson D, Birmaher B, Phillips ML. Neural substrates of child irritability in typically developing and psychiatric populations. Dev Cogn Neurosci. 2015;14:71-80.

RI PT

20. Pawliczek CM, Derntl B, Kellermann T, Gur RC, Schneider F, Habel U. Anger under control: neural correlates of frustration as a function of trait aggression. PLOS ONE. 2013;8(10):e78503.

SC

21. Dougherty LR, Smith VC, Bufferd SJ, et al. DSM-5 disruptive mood dysregulation

disorder: correlates and predictors in young children. Psychol Med. 2014;44:2339-2350.

M AN U

22. Dougherty LR, Smith VC, Bufferd SJ, Kessel EM, Carlson GA, Klein DN. Disruptive mood dysregulation disorder at the age of 6 years and clinical and functional outcomes 3 years later. Psychol Med. 2016;46:1103–14.

23. Wiggins JL, Mitchell C, Stringaris A, Leibenluft E. Developmental trajectories of

TE D

irritability and bidirectional associations with maternal depression. J Am Acad Child Adolesc Psychiatry. 2014;53:1191–1205. 24. Fareri DS, Gabard-Durnam L, Goff B, et al. Normative development of ventral striatal

EP

resting state connectivity in humans. Neuroimage. 2015;118:422–437. 25. Gabard-Durnam LJ, Flannery J, Goff B, et al. The development of human amygdala

AC C

functional connectivity at rest from 4 to 23 years: a cross-sectional study. Neuroimage. 2014;95:193-207.

26. Gogtay N, Giedd JN, Lusk L, et al. Dynamic mapping of human cortical development during childhood through early adulthood. Proc Natl Acad Sci USA. 2004;101:81748179.

ACCEPTED MANUSCRIPT

27. Wiggins JL, Brotman MA, Adleman NE, et al. Neural correlates of irritability in disruptive mood dysregulation and bipolar disorders. Am J Psychiatry. 2016;173:722730.

RI PT

28. Delgado MR. Reward-related responses in the human striatum. Ann NY Acad Sci. 2007;1104:70–88.

29. Dougherty LR, Tolep MR, Smith VC, Rose S. Early exposure to parental depression and

SC

parenting: Associations with young offspring’s stress physiology and oppositional behavior. J Abnorm Child Psychol. 2013;41:1299–1310.

M AN U

30. Wiggins JL, Schwartz KTG, Kryza-Lacombe M, Spechler PA, Blankenship SL, Dougherty LR. Neural reactivity to reward in school-age offspring of depressed mothers. J Affect Disord. 2017;214:81-88.

31. Egger HL, Erkanli A, Keeler G, Potts E, Walter BK, Angold A. Test-retest reliability of

2006;45:538-549.

TE D

the preschool age psychiatric assessment. J Amer Acad Child Adolesc Psychiatry.

32. Copeland WE, Angold A, Costello EJ, Egger H. Prevalence, comorbidity, and correlates

EP

of DSM-5 proposed disruptive mood dysregulation disorder. Amer J Psychiatry. 2013;170:173-179.

AC C

33. Luby J, Belden A, Botteron K, et al. The effects of poverty on childhood brain development The mediating effect of caregiving and stressful life events. JAMA Pediatr. 2013;167:1135–1142.

34. McLaren DG, Ries ML, Xu G, Johnson SC. A generalized form of context-dependent psychophysiological interactions (gPPI): A comparison to standard approaches. Neuroimage. 2012;61:1277-1286.

ACCEPTED MANUSCRIPT

35. Talairach J, Tournoux P. Co-Planar Stereotaxic Atlas of the Human Brain. Thieme, Stuttgart, Germany;1988. 36. Eklund A, Nichols TE, Knutsson H. Cluster failure: Why fMRI inferences for spatial

RI PT

extent have inflated false-positive rates. Proc Natl Acad Sci. 2016;113:7900-7905.

37. Cox RW, Chen G, Glen DR, Reynolds RC, Taylor PA. Clustering in AFNI: FalsePositive Rates Redux. Brain Connect. 2017;7:152-171.

SC

38. Richards JM, Plate RC, Ernst M. A systematic review of fMRI reward paradigms used in studies of adolescents vs. adults: the impact of task design and implications for

M AN U

understanding neurodevelopment. Neurosci Biobehav Rev. 2013;37:976-91. 39. Perlman SB, Pelphrey KA (2011). Developing connections for affective regulation: Agerelated changes in emotional brain connectivity. J Exp Child Psychol. 2011;108:607-620. 40. Buckner RL, Andrews-Hanna JR, Schacter DL. The brain's default network: anatomy,

TE D

function, and relevance to disease. Ann NY Acad Sci. 2008;1124:1e38. 41. Gabbay V, Ely BA, Qingyang Y, et al. Striatum-based circuitry of adolescent depression and anhedonia. J Amer Acad Child Adolesc Psychiatry. 2013;52:628-641.

EP

42. Hwang JW, Xin SC, Ou YM, et al. Enhanced default mode network connectivity with ventral striatum in subthreshold depression individuals. J Psychiatr Res. 2016;76:111-

AC C

120.

43. Quevedo K, Rowena N, Scott H, et al. Ventral striatum functional connectivity during rewards and losses and symptomatology in depressed patients. Biol Psychol.

2016;123:62-73.

44. Galván A. The teenage brain: Sensitivity to rewards. Curr Dir Psychol Sci. 2013;22:88– 93.

ACCEPTED MANUSCRIPT

45. Silvers JA, Shu J, Hubbard AD, Weber J, Ochsner KN. Concurrent and lasting effects of emotion regulation on amygdala response in adolescence and young adulthood. Dev Sci. 2014;18:771-784.

RI PT

46. Luna B, Marek S, Larsen B, et al: An integrative model of the maturation of cognitive control. Annu Rev Neurosci. 2015;38:151–170.

47. Ordaz SJ, Foran W, Velanova K, Luna B. Longitudinal growth curves of brain function

SC

underlying inhibitory control through adolescence. J Neurosci. 2013;33:18109-18124.

Guidance Service;1997.

M AN U

48. Dunn LM, Dunn LM. Peabody Picture Vocabulary Test. 3rd ed. Circle Pines: American

49. First MB, Spitzer RL, Gibbon M, Williams JBW. The Structured Clinical Interview for DSM–IV Axis I Disorders—Non-Patient Edition. New York, NY: Biometrics Research Department, New York State Psychiatric Institute;1996.

TE D

50. Aiken LS, West SG. Multiple regression: Testing and interpreting interactions. Thousand

AC C

EP

Oaks, CA: Sage;1991.

ACCEPTED MANUSCRIPT Tables Table 1. Sample characteristics. Wave 1

Wave 2

Age (years)a

4.27(0.80)

7.52(0.78)

Gender (female)

28(61%)

24(53%)

African American

13(29%)

Multiracial

4(9%)

Other

4(9%)

M AN U

White/Caucasian

Hispanic ethnicity

7(16%)

Cognitive functioning

113.17(15.82)

Income

1(2%)

$20,001-$40,000

3(7%)

$70,001-$100,000

13(30%)

12(27%)

15(34%)

EP

> $100,000

TE D

<$20,000

$40,001-$70,000

SC

Race

RI PT

Demographics

AC C

Parental education (>4-year college degree) Mother

28(61%)

Father

30(68%)

Clinical symptoms Child irritability

1.07(1.53) Range:0-6

1.61(1.61) Range:0-5

Maternal depressionb

25(54%)

27(59%)

Note: Dimensional characteristics are displayed as M(SD) and categorical characteristics are displayed as n(%); cognitive functioning was measured by the Peabody Picture Vocabulary

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Test48; clinical symptoms were measured by the Preschool Age Psychiatric Assessment (PAPA);31 aindicates that Wave 1 and 2 group means significantly differed; blifetime history of maternal depression was determined by the Structured Clinical Interview for DSM-IV-TRAxis I Disorders, Non-Patient Version.49

ACCEPTED MANUSCRIPT

Table 2. Summary of generalized psychophysiological interaction (gPPI) results. Coordinates Max. F

X

Y

z

RI PT

K

WAVE 1 IRRITABILITY (controlling for Wave 2) x PERFORMANCE (hit, miss) x CONDITION (no reward, reward) 37

19.44

52.5

-40.5

50.5

R Amygdala–R Insula

36

26.92

43.5

-16.5

17.5

SC

R Amygdala–R Inferior Parietal Lobule

WAVE 2 IRRITABILITY (controlling for Wave 1) x PERFORMANCE x CONDITION 28

21.12

M AN U

L Amygdala–L Superior Frontal Gyrus

-1.5

28.5

50.5

R Amygdala–R Superior Frontal Gyrus

36

18.57

1.5

28.5

56.5

L Ventral Striatum–R Precuneus

34

24.81

16.5

-73.5

41.5

34

17.62

34.5

-58.5

-24.5

L Ventral Striatum–R Culmen

WAVE 1 IRRITABILITY (controlling for Wave 2) x PERFORMANCE 87

19.52

43.5

-67.5

41.5

R Amygdala–L Inferior Parietal Lobule

76

27.87

-46.5

-43.5

38.5

R Amygdala–L Angular Gyrus

48

18.80

-46.5

-61.5

35.5

L Ventral Striatum–L Postcentral Gyrus

43

29.65

-55.5

-19.5

38.5

L Ventral Striatum–R Lingual Gyrus

40

21.87

1.5

-73.5

-0.5

L Ventral Striatum–R Superior Parietal Lobule

35

26.26

16.5

-64.5

56.5

28

23.74

10.5

-55.5

-3.5

AC C

EP

TE D

R Amygdala–R Inferior Parietal Lobule

L Amygdala–R Precuneus

111

27.68

28.5

-55.5

38.5

R Amygdala–L Precuneus

39

28.97

-13.5

-49.5

50.5

R Amygdala–L Middle Temporal Gyrus

32

23.41

-28.5

-58.5

23.5

L Ventral Striatum–R Precuneus

36

29.93

10.5

-76.5

38.5

L Ventral Striatum–R Culmen

WAVE 2 IRRITABILITY (controlling for Wave 1) x PERFORMANCE

ACCEPTED MANUSCRIPT

L Ventral Striatum–L Precuneus

28

16.19

-28.5

-58.5

38.5

WAVE 1 IRRITABILITY (controlling for Wave 2) x CONDITION No significant interaction.

49

24.30

-13.5

46.5

38.5

38

24.09

43.5

-49.5

23.5

PERFORMANCE x CONDITION R Amygdala–R Supramarginal Gyrus

M AN U

WAVE 1 IRRITABILITY (controlling for Wave 2)

SC

L Ventral Striatum–L Superior Frontal Gyrus

RI PT

WAVE 2 IRRITABILITY (controlling for Wave 1) x CONDITION

No significant main effect of Wave 1 irritability.

WAVE 2 IRRITABILITY (controlling for Wave 1) No significant main effect of Wave 2 irritability.

TE D

PERFORMANCE

32

16.31

34.5

-64.5

38.5

L Ventral Striatum–L Superior Frontal Gyrus

54

17.93

-10.5

25.5

53.5

L Ventral Striatum–R Paracentral Lobule

34

20.77

7.5

-25.5

44.5

CONDITION

EP

L Amygdala–R Precuneus

AC C

No significant main effect of condition. Note: Regions were identified by the Talairach-Tournoux atlas; L=left; R=right. No significant effects were observed for the right ventral striatum.

ACCEPTED MANUSCRIPT

Figures Figure 1. More severe preschool-age irritability predicts atypical modulation of amygdala and ventral striatal connectivity in school-age, controlling for concurrent irritability.

RI PT

Note: (A) Right amygdala–right inferior parietal lobule and (B) right amygdala–right insula connectivity clusters with significant Wave 1 Irritability x Performance x Condition interactions during feedback period; (C) left ventral striatum–right lingual gyrus connectivity cluster with

SC

significant Wave 1 Irritability x Performance interaction during feedback period. Brain images represent axial sections (left=left) with threshold set at whole-brain corrected p<.05. Interaction

M AN U

effects were estimated using simple slopes analyses50 at the maximum level of youth irritability (“high” irritability), at the mean level of youth irritability, and at the minimum level of youth irritability (“low” irritability). Values from clusters were extracted and averaged for plots.

TE D

Figure 2. Atypical modulation of amygdala and ventral striatal connectivity in children with more severe school age irritability, above and beyond preschool age irritability. Note: (A) Right amygdala- and (B) left amygdala–superior frontal gyrus and (C) left striatum–

EP

right precuneus connectivity clusters with significant Wave 2 Irritability x Performance x Condition interactions during feedback period. Brain images represent axial sections (left=left)

AC C

with threshold set at whole-brain corrected p<.05. Interaction effects were estimated using simple slopes analyses50 at the maximum level of youth irritability (“high” irritability), at the mean level of youth irritability, and at the minimum level of youth irritability (“low” irritability). Values from clusters were extracted and averaged for plots.

ACCEPTED MANUSCRIPT

Preschool and School-Age Irritability Predict Reward-Related Brain Function

Lea Dougherty, PhD, Karen T. G. Schwartz, MS, Maria Kryza-Lacombe, MA, Jill Weisberg,

RI PT

PhD, Philip A. Spechler, MA, Jillian Lee Wiggins, PhD

Funding: This research was supported by the Maryland Neuroimaging Center Seed Grant Program (L.R.D.), the National Science Foundation in partnership with the University of

SC

Maryland Type: ADVANCE Program for Inclusive Excellence (L.R.D. and Tracy Riggins), the University of Maryland College of Behavioral and Social Sciences Dean’s MRI Research

M AN U

Initiative RFP Program (L.R.D. and Tracy Riggins), the Behavioral and Social Sciences Dean’s Research Initiative (L.R.D.), and the Research and Scholarship Award (L.R.D.).

TE D

Statistical Expert: Dr. Wiggins served as the statistical expert for this research.

Acknowledgements: The authors are grateful to all of the participating families. They would also like to recognize Sarah Blankenship, PhD, Victoria Smith, PhD, Marissa Kushner, PhD, Stephanie Merwin, PhD, and Katherine Leppert, MA, for assistance with data collection and

EP

Srikanth Padmala, PhD, and Mihai Sirbu, BS, for assistance in data processing, all of the

AC C

University of Maryland College Park.

Disclosures: Drs. Dougherty, Weisberg, Wiggins, Mss. Schwartz and Kryza-Lacombe, and Mr. Spechler report no biomedical financial interests or potential conflicts of interest.

ACCEPTED MANUSCRIPT

Figure 1

Condition

A. Right Amygdala Seed Mean Irritability

Hit

Miss

Hit

TE D

Predicted Connectivity

Low Irritability

Right Insula xyz = 44, -17, 18 k = 36, F1,43 = 26.92

Hit

C.Left Striatum Seed

Miss

Low Irritability

Mean Irritability

Hit

Miss

Mean Irritability

Hit

Miss

High Irritability

Hit

Miss

High Irritability

AC C

EP

Predicted Connectivity

Miss

M AN U

B. Right Amygdala Seed

SC

Predicted Connectivity

Right Inferior Parietal Lobule xyz = 53, -41, 51 k = 37, F1,43 = 19.44

High Irritability

RI PT

Low Irritability

Reward No Reward

Lingual Gyrus xyz = 2, -74, -1 k = 40 F1,43 = 21.87

Hit

Miss

Hit

Miss

Hit

Miss

ACCEPTED MANUSCRIPT

Figure 2

Condition

A. Right Amygdala Seed

Hit

Hit

Hit

Miss

Low Irritability

Hit

Miss

Mean Irritability

High Irritability

Hit

Hit

Miss

Miss

Mean Irritability

High Irritability

Hit

Hit

AC C

EP

Predicted Connectivity

Miss

M AN U

TE D

Predicted Connectivity

Low Irritability

C.Left Striatum Seed

SC

Miss

B. Left Amygdala Seed

Superior Frontal Gyrus xyz = -2, 29, 51 k = 28, F1,43 = 21.12

High Irritability

RI PT

Mean Irritability

Predicted Connectivity

Superior Frontal Gyrus xyz = 2, 29, 57 k = 36, F1,43 = 18.57

Low Irritability

Reward No Reward

Right Precuneus xyz = 17, -74, 42 k = 34, F1,43 = 24.81

Hit

Miss

Miss

Miss

ACCEPTED MANUSCRIPT 1 IRRITABILITY AND REWARD Supplement 1 We ran covariate analyses examining whether reported results for the longitudinal associations between Wave 1 irritability and Wave 2 reward-related neural functioning were better accounted

RI PT

for by 1) child age at Wave 2, 2) maternal lifetime depression status, 3) child anxiety symptoms at Wave 1, 4) child depression symptoms at Wave 1, 5) child attention-deficit/hyperactivity (ADHD) symptoms at Wave 1, 6) child oppositional defiant disorder (ODD) symptoms at Wave

SC

1, and/or 7) time between Wave 1 and Wave 2 (M=3.25 years, SD=0.52). In addition, covariate analyses examined whether reported results for the concurrent associations between Wave 2

M AN U

irritability and Wave 2 reward-related neural functioning were better accounted for by 1) child age at Wave 2, 2) maternal lifetime depression status, 3) child anxiety symptoms at Wave 2, 4) child depression symptoms at Wave 2, 5) child ADHD symptoms at Wave 2, 6) child ODD symptoms at Wave 2, and/or 7) time between Wave 1 and Wave 2. See Supplemental Table 1.

TE D

Maternal depression status was defined by maternal lifetime history of depressive disorders (major depressive disorder or dysthymic disorder) as measured by the Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Non-Patient Edition (SCID-NP49). There

EP

were 27 children who had mothers with lifetime depression and 19 children who had mothers with no lifetime depression history. Child anxiety, depression, ADHD, and ODD symptoms were

AC C

assessed dimensionally using the PAPA at Wave 1 and Wave 2 (see Methods section in main text), reflecting primary caregiver report of child symptoms within the 3 months prior (see Supplemental Table 1 for symptom characteristics). The items included in symptom scales did not overlap with items in the irritability scale used in the analyses (see Dougherty et al.5,6). Increased irritability at Wave 1 was significantly associated with higher levels of ODD symptoms at the same timepoint (r=.43, p=.004). Higher irritability scores at Wave 2 were

ACCEPTED MANUSCRIPT 2 IRRITABILITY AND REWARD significantly associated with higher levels of depression symptoms at Wave 1 (r=.41, p=.006) and ODD symptoms at Wave 2 (r=.49, p=.001). Irritability was not significantly associated with anxiety or ADHD symptom scores at either timepoint.

RI PT

Connectivity values from the clusters in the main findings (left amygdala-superior frontal gyrus, right amygdala-superior frontal gyrus, left ventral striatum-precuneus, identified in the Irritability at Wave 2 (controlling for Wave 1) x Performance x Condition contrast; right

SC

amygdala-inferior parietal lobule and –insula, identified in the Irritability Wave 1 (controlling for Wave 2) x Performance x Condition; and left ventral striatum-lingual gyrus, identified in the

M AN U

Irritability Wave 1 (controlling for Wave 2) x Performance) were extracted. Analyses were repeated, independently covarying for each factor listed above, using IBM SPSS Statistical Software. All gPPI findings reported in the main text remained significant for the 3-way interaction, despite the inclusion of the covariates (p < .05). Results are summarized in

AC C

EP

TE D

Supplemental Table 2 and Supplemental Table 3.

ACCEPTED MANUSCRIPT Running head: IRRITABILITY AND REWARD

34

Supplemental Tables Supplemental Table 1. Characteristics of covarying child symptom scales.

AC C

EP

TE D

M AN U

SC

RI PT

Wave 1 Wave 2 Symptom Range M (SD) ICC α Range M (SD) ICC α Anxiety 0-55 12.22 (7.97) .97 .81 0-48 12.15 (8.66) .98 .83 Depression* 0-12 2.39 (2.10) .93 .57 0-14 4.08 (3.00) .98 .66 ADHD* 0-12 2.42 (3.07) .94 .84 0-25 5.04 (5.62) .98 .91 ODD 0-9 2.74 (1.88) .80 .60 0-9 2.62 (2.07) .94 .65 Note: All clinical symptoms were measured by the Preschool Age Psychiatric Assessment (PAPA31); scales did not overlap in content with the irritability scale; *indicates that Wave 1 and Wave 2 group means significantly differed (p < .05); ICC=intraclass correlation coefficient; ADHD=Attention-Deficit/Hyperactivity Disorder; ODD=Oppositional Defiant Disorder.

ACCEPTED MANUSCRIPT 2 IRRITABILITY AND REWARD

Supplemental Table 2. Main findings for preschool irritability predicting school-age brain

psychophysiological interaction (gPPI) findings.

RI PT

connectivity remain significant after covarying factors on irritability generalized

R Amygdala Insula

M AN U

R Amygdala Inferior Parietal Lobule

SC

Irritability Wave 1 (controlling for Wave 2) x Performance x Condition

Irritability Wave 1 (controlling for Wave 2) x Performance L Ventral Striatum – Lingual Gyrus

F(1,42)=19.98

F(1,42)=17.23

F(1,42)=13.31

Maternal depression status

F(1,42)=25.59

F(1,42)=18.37

F(1,42)=13.90

Child anxiety symptoms

F(1,39)=22.43

F(1,39)=18.94

F(1,39)=14.14

Child depression symptoms

F(1,39)=23.61

F(1,39)=19.90

F(1,39)=15.62

Child ADHD symptoms

F(1,39)=23.89

F(1,39)=21.14

F(1,39)=15.35

F(1,39)=18.37

F(1,39)=17.69

F(1,39)=22.27

F(1,42)=22.01

F(1,42)=18.18

F(1,42)=14.31

Child ODD symptoms

TE D

Child age at Wave 2

Time between Wave 1 and Wave 2

AC C

EP

Note: all ps<.001; child symptom scales were assessed at Wave 1 and did not overlap with the PAPA irritability scale; R amygdala cells reflect the significance of 3-way interaction findings (Irritability Wave 1 [controlling for Wave 2] x Performance x Condition) after taking other factors into account; L striatum cells reflect the significance of 2-way interaction findings (Irritability Wave 1 [controlling for Wave 2] x Performance) after taking other factors into account; ADHD=Attention-Deficit/Hyperactivity Disorder; ODD=Oppositional Defiant Disorder; L=left; R=right.

ACCEPTED MANUSCRIPT 3 IRRITABILITY AND REWARD Supplemental Table 3. Main findings for concurrent irritability in relation to school age brain connectivity remain significant after covarying factors on irritability generalized

RI PT

psychophysiological interaction (gPPI) findings.

Irritability Wave 2 (controlling for Wave 1) x Performance x Condition R Amygdala Superior Frontal Gyrus

Child age at Wave 2

F(1,42)=19.98

F(1,42)=21.58

F(1,42)=25.78

Maternal depression status

F(1,42)=15.50

F(1,42)=16.58

F(1,42)=24.69

Child anxiety symptoms

F(1,42)=18.71

F(1,42)=22.64

F(1,42)=25.21

Child depression symptoms

F(1,42)=17.67

F(1,42)=19.57

F(1,42)=25.41

Child ADHD symptoms

F(1,42)=17.68

F(1,42)=19.44

F(1,42)=25.47

Child ODD symptoms

F(1,42)=15.96

F(1,42)=16.87

F(1,42)=26.41

Time between Wave 1 and Wave 2

F(1,42)=19.71

F(1,42)=20.56

F(1,42)=25.07

TE D

M AN U

SC

L Amygdala Superior Frontal Gyrus

L Ventral Striatum Precuneus

AC C

EP

Note: all ps<.001; child symptom scales were assessed at Wave 2 and did not overlap with the PAPA irritability scale; cells reflect the significance of findings (Irritability Wave 2 [controlling for Wave 1] x Performance x Condition) after taking other factors into account; ADHD=Attention-Deficit/Hyperactivity Disorder; ODD=Oppositional Defiant Disorder; L=left; R=right.