ruminative brooding is associated with salience network...

Ruminative brooding is associated with salience

network coherence in early pubertal youthSarah J Ordaz1 Joelle LeMoult2 Natalie L Colich2 Gautam Prasad3

Madeline Pollak2 Morgan Popolizio2 Alexandra Price2 Michael Greicius4

and Ian H Gotlib2

1Department of Psychiatry and Behavioral Sciences Stanford University School of Medicine Stanford CAUSA 2Department of Psychology Stanford University Stanford CA USA 3Imaging Genetics Center ampLaboratory of Neuro Imaging USC Mark and Mary Stevens Neuroimaging and Informatics Institute KeckSchool of Medicine of USC Los Angeles CA USA and 4Department of Neurology and Neurological SciencesStanford University School of Medicine Stanford CA USA

Correspondence should be addressed to Sarah J Ordaz PhD Department of Psychiatry and Behavioral Sciences Stanford University 401 Quarry Rd MC5722 Stanford CA 94305 USA E-mail sjo22stanfordedu

Abstract

Rumination and particularly ruminative brooding perpetuates dysphoric mood states and contributes to the emergence ofdepression Studies of adults and older adolescents have characterized the association between rumination and intrinsicfunctional connectivity within default mode (DMN) salience (SN) and executive control (ECN) networks we know littlehowever about the brain network basis of rumination during early puberty a sensitive period for network reorganization112 early puberty boys and girls completed resting-state scans the Ruminative Response Scale and the Youth Self-Reportquestionnaire Using independent components analysis and dual regression we quantified coherence for each individual innetworks of interest (SN ECN DMN) and in non-relevant networks (motor visual) in which we predicted no correlationswith behavioral measures Boys and girls did not differ in levels of rumination or internalizing symptoms or in coherencefor any network The relation between SN network coherence and rumination however and specifically ruminative brood-ing was moderated by sex greater SN coherence was associated with higher levels of brooding in girls but not in boysFurther in girls brooding mediated the relation between SN coherence and internalizing symptoms These results point tocoherence within the SN as a potential neurodevelopmental marker of risk for depression in early pubertal girls

Key words intrinsic functional connectivity rumination puberty anterior cingulate salience network

Introduction

Rumination is defined as lsquoa mode of responding to distress thatinvolves repetitively and passively focusing on symptoms of dis-tress and on the possible causes and consequences of thesesymptomsrsquo (Nolen-Hoeksema 1991 p 569) Although ruminat-ing individuals appraise their problems as unsolvable(Lyubomirsky et al 1999) generate less effective solutions toproblems (Lyubomirsky and Nolen-Hoeksema 1995Lyubomirsky et al 1999) and show low motivation to implement

strategies that are generated (Wenzlaff et al 1988 Lyubomirskyand Nolen-Hoeksema 1993) Further individuals who ruminateare less likely to reach out for social support (Nolen-Hoeksemaand Davis 1999) and report high levels of social friction (Nolen-Hoeksema and Davis 1999) Importantly longitudinal studies in-dicate that both youth and adults who ruminate are more likelyto subsequently develop depression (Nolen-Hoeksema et al2008) and other forms of psychopathology (Joormann et al 2006McLaughlin and Nolen-Hoeksema 2011)

Received 29 May 2016 Revised 26 July 2016 Accepted 6 September 2016

VC The Author (2016) Published by Oxford University Press For Permissions please email journalspermissionsoupcom

298

Social Cognitive and Affective Neuroscience 2017 298ndash310

doi 101093scannsw133Advance Access Publication Date 14 September 2016Original article

Recently investigators have characterized brain networksassociated with high levels of rumination These studies havecapitalized on the recognition that functional MRI signalsacquired when individuals are at rest exhibit reproducible oscil-latory dynamics (Biswal et al 1995) By associating time seriesof these oscillatory patterns it is possible to identify a networkof regions that activate in tandem and therefore likely have ahistory of functional co-activation (Buckner et al 2013)Importantly the networks identified in these intrinsic (orresting-state) functional connectivity analyses are reliableacross time (Zuo et al 2010 Thomason et al 2011Somandepalli et al 2015) and reflect known structural connec-tions (Greicius et al 2009) Further they correspond to patternsof task-based activation that have been reported across a var-iety of paradigms (Smith et al 2009) including induced rumin-ation (Berman et al 2014) Researchers have conductedfunctional connectivity analyses on these intrinsic networks toprobe the network basis of rumination in the default mode net-work (DMN) salience network (SN) and executive control net-work (ECN) These networks been implicated inpsychopathology across all age groups (Menon 2011 Hamiltonet al 2013) and may be involved in rumination The DMN isinvolved in self-referential and memory consolidation proc-esses SN is implicated in orienting towards novel potentiallythreatening stimuli and ECN is involved in inhibiting the per-sevative thinking that characterizes rumination and shifting in-dividualsrsquo cognitive sets to another line of thought (Menon2011) The most robust evidence linking intrinsic networks withrumination has been documented for the DMN adult studiesconverge to indicate that high ruminators have stronger con-nectivity between regions within the DMN than do their low-rumination counterparts (Berman et al 2011 2014 Hamiltonet al 2011 Piguet et al 2014 Luo et al 2015) High ruminatingadults also evidence greater connectivity within the SN (Kuhnet al 2014) and decreased connectivity within the ECN (Kuhnet al 2012)

Importantly rumination is a multidimensional constructconsisting of two components ruminative brooding and ru-minative reflection (henceforth lsquobroodingrsquo and lsquoreflectionrsquo) andit is possible that brooding and reflection are differentiallyrelated to brain networks Brooding involves passively compar-ing onersquos current situation to an unachieved standard whereasreflection involves intentionally turning inward to alleviate dys-phoric mood Although neither type of rumination is consideredto be adaptive brooding has been associated with the emer-gence of higher levels of depressive symptoms than has reflec-tion (Treynor et al 2003) Indeed all of the studies with adultsdescribed above that examined the network basis of ruminationreported differential associations for brooding and reflection(Berman et al 2011 Hamilton et al 2011 Kuhn et al 2014 Luoet al 2015) highlighting the importance of examining these twocomponents of rumination separately

Only two studies to date have examined the brain networkbasis of rumination in youth and these assessed a sample ofolder (13ndash17 years) depressed adolescents (Connolly et al 2013Ho et al 2015) researchers have not yet characterized the brainbasis of rumination in a sample prior to the typical onset of de-pression This is particularly important given that the transitionfrom childhood to adolescence appears to be a sensitive periodboth for the emergence of rumination and for the developmentof brain networks that support this cognitive process in adultsIndeed levels of rumination increase during this period(Hampel and Petermann 2005) and connections within DMNSN and ECN continue to mature (Satterthwaite et al 2013b)

The emergence of sex differences in brain and behavior alsooccurs during the transition to adolescence By definition ado-lescence begins with the onset of puberty during which concen-trations of circulating gonadal hormones rise and bind toreceptors that are present throughout the brain and influencesubsequent brain development (Sisk and Zehr 2005) This inturn contributes to sexual dimorphisms in the brain (Lenrootet al 2007 Goddings et al 2014) and may also shape behav-iorsmdashsuch as ruminationmdashthat differ by sex Indeed sex differ-ences in rumination (Johnson and Whisman 2013) do notemerge until later in puberty (Burwell and Shirk 2007 Roodet al 2009 Abela and Hankin 2011 Hamilton et al 2015) Giventhis timing it is possible that sex differences in brain networkfunctioning set the stage for later-emerging sex differences inrumination

To examine these issues we recruited a sample of early-pubertal youth in which boys and girls were matched on puber-tal status rather than on age given theories suggesting that it ispubertal maturation specificallymdashrather than other age-relatedfactorsmdashthat contributes to the emergence of sex differences ininternalizing symptomatology in mid-adolescence (Sisk andZehr 2005 Patton and Viner 2007 Blakemore et al 2010) thisallowed us to ensure that sex differences are not confounded bygirlsrsquo earlier entry into puberty We assessed the relationsamong connectivity in three intrinsic networks (DMN SN andECN) levels of ruminative brooding and reflection (the latter isreported in the supplement) and levels of internalizing symp-toms that are posited to arise from ruminative thinking as wellas in two non-relevant networks (motor visual) Based on previ-ous findings we predicted first that ruminative broodingwould be associated with internalizing symptomatologySecond we predicted that although boys and girls would notdiffer with respect to rumination or internalizing symptoms atthis early pubertal stage they would differ in the coherence(strength of within-network connectivity) of brain networkssetting the stage for sex differences in levels of rumination andinternalizing symptoms that emerge later in puberty Third wepredicted that ruminative brooding but not reflection would beassociated with patterns of activation in the three intrinsicbrain networksmdashmost strongly in DMN which has the strongestevidence for this relation but not in the motor or visual net-works If so we sought to examine whether ruminative brood-ing would mediate a relation between network coherence andinternalizing symptoms Finally we explored sex differences inthe network basis of rumination

Materials and methodsParticipants

Participants were native English speakers recruited fromthroughout the San Francisco Bay Area community through on-line posts to Craigslist and to parent listservs Advertisementstargeted an age group corresponding to the early stages of pu-berty in boys and girls Participantsrsquo pubertal status was as-sessed by parent report of pubertal status and by subsequentconfirmation upon visit to the laboratory Children wereexcluded if they reported neurological (including severe headinjuries) cardiovascular or any other major medical problemsThe study complied with Institutional Review Board guidelinesand in accordance with the Declaration of Helsinki participantsand their parents provided written informed assent and con-sent respectively Participants were compensated for their par-ticipation 139 children completed a neuroimaging scan session

S J Ordaz et al | 299

Scans from 27 participants were excluded from data analysesdue to scanner problems (n frac14 12) administration of an incorrectprotocol (n frac14 1) unscorable cardiac physiological data (n frac14 6)unusable structural data (n frac14 2) and excessive motion duringthe scan (n frac14 6) Thus data from 112 children (53 boys and 59girls) were included in the final analyses The racial and ethnicdemographics of this sample were representative of the BayArea 47 Caucasian 6 African-American 10 Hispanic 17Asian 3 Native American and 16 multiracial

Measures

Questionnaires We administered a ten-item RuminationResponse ScalemdashAdolescent Version (RRS-A) (Burwell andShirk 2007) and examined items from the five-item broodingand the five-item self-reflection subscales (Treynor et al 2003)Brooding and self-reflection are conceptualized to be two dis-tinct subtypes of rumination Brooding involves passive com-parisons to an unachieved standard and items on this scaleinclude lsquoThink about a recent thing that happened and wish ithad gone betterrsquo and lsquoThink ldquoWhy do I always react this wayrdquorsquoReflection involves a purposeful turning inward to problem-solve in order to alleviate dysphoric mood items on this scaleinclude lsquoWrite down what you are thinking about and analyzeitrsquo and lsquoAnalyze recent events to try to understand why you aresad or upsetrsquo

Participants used the Youth Self-Report questionnaire(Achenbach 1991 Achenbach and Rescorla 2001) to reportabout problem behaviors over the previous six months Wechose to use this questionnaire because it captures a broadspectrum of functioning both within and outside of the clinicalrange (Achenbach and Rescorla 2001) We analyzed the 16-itemAnxiousDepressed Problems Subscale

Pubertal status Stage of pubertal maturation was determinedusing self-reported Tanner Staging (Marshall and Tanner 1968)a reliable and valid metric (Slora et al 2009) Girls rated boththeir pubic hair and breast development and boys rated boththeir pubic hair and peniletesticular development on five-point scales A composite Tanner Stage score was computed foreach participant by averaging the two scores

Clinical interview Trained interviewers assessed the diagnos-tic status of participants by administering the Schedule forAffective Disorders and Schizophrenia for School-Age Children-Present and Lifetime version to both the child and parent(Kaufman et al 1997)

Neuroimaging data acquisition We present details of neuroi-maging data acquisition in the Supplemental Information

Data pre-processing

We implemented a preprocessing stream that could be applied atthe individual rather than the group level of analysis we selectedthis stream on the basis of data comparing approaches to minim-ize confounds without introducing noise to the data (Hallquistet al 2013 Jo et al 2013 Satterthwaite et al 2013a Power et al2014) This research guided decisions about the best approaches touse in order to address confounds generated by motion (Poweret al 2012 Satterthwaite et al 2012 Van Dijk et al 2012 Yan et al2013) and to ensure appropriate ordering of preprocessing stepsparticularly bandpass filtering relative to nuisance regression(Hallquist et al 2013 Jo et al 2013) In cases in which there werecontroversies about how best to deal with confounds we selectedan approach that had been demonstrated to be effective in similardata sets of youth (Jo et al 2013 Satterthwaite et al 2013a) We

present details regarding our approach to addressing confoundsand pre-processing in the Supplementary Materials

Operationalizing network coherence

We defined coherence as the strength of the association be-tween timecourses of all voxels within an identified networkFollowing preprocessing we obtained a metric of resting-statefunctional connectivity for each individual within each intrinsicnetwork of interest using a combination of group probabilisticindependent components analysis (ICA) and dual regressionanalyses Group ICA is a multivariate signal processing method(Kiviniemi et al 2003 Beckmann et al 2005) that we selectedover seed-based correlation because it accounts for multiplevoxel-voxel relations in order to obtain an interacting networkof voxels it permits data-driven exploration of the spatial-temporal properties of functional neuroimaging data and thegroup ICA followed by dual regression approach has beenshown to have higher short- and long-term test-retest reliabilitythan seed-based analysis (Zuo et al 2010 Smith et al 2014)

We conducted group ICA using FSL 506 MELODIC softwareversion 314 specifying 25 components Specifically a set of ICAcomponents derived from the whole sample was generated andused to elicit individual maps of a comparable network for statis-tical analysis Group components that were identified from groupICA were visually inspected and networks comprising the net-works of interest (DMN SN and ECN) and non-relevant networks(motor visual) were visually identified on the basis of theirneuroanatomical components by trained raters (MG SO) Thesenetworks are visualized in Figure 1 and include one networkcomprising the anterior subdivision of the DMN (DMNA) anothernetwork comprising a ventral subdivision (DMNV) a single SN aright and left ECN (ECNR ECNL) and visual and motor networks

The set of spatial maps from the group-average analysis wasused to generate individual-specific versions of the spatial mapsand associated timeseries using dual regression (Beckmannet al 2009 Filippini et al 2009) First for each individual thegroup-average set of spatial maps is regressed (as spatial regres-sors in a multiple regression) into the individualrsquos 4D space-timedataset This results in a set of individual-specific timeseriesone per group-level spatial map Next those timeseries are re-gressed (as temporal regressors) into the same 4D dataset re-sulting in a set of individual-specific spatial maps one pergroup-level spatial map Each spatial map contains regressionweights that serve as a measure of each voxelrsquos functional con-nectivity with the identified network while controlling for the in-fluence of other networks some of which may reflect artifactssuch as physiological noise Z-scores that normalize by the re-sidual within-subject noise were used We then applied thegroup mask for a given component to the individual-level spatialmap for the corresponding component and values within thismask were averaged to produce a metric of network coherencefor each individual Our approach is consistent with publishedapproaches (Van Duijvenvoorde et al 2015)

Statistical analyses

A priori statistical analyses were conducted using SPSS softwareversion 23 We examined individual coherence metrics for eacha priori identified network as well as for behavioral variables toensure normality of data distributions We used t-tests to testhypotheses concerning sex differences We used correlationalanalyses to examine the relation between rumination and anx-iousdepressed symptoms We examined brain-behavior

300 | Social Cognitive and Affective Neuroscience 2017 Vol 12 No 2

relations by (i) correlating each network coherence estimatewith brooding and (ii) using ordinary least squares multiple re-gression to test whether sex moderated the relation betweennetwork coherence and brooding Based on the findingsdescribed below we tested a mediational model relating SN co-herence brooding and anxiousdepressed symptoms in fe-males only (associations were not significant in males) usingpath-analytic approaches implemented in PROCESS for SPSSand utilizing 95 bias-corrected bootstrap confidence intervalsfor the indirect effect based on 10 000 bootstrap samples(MacKinnon 2008 Hayes 2013) Finally we combined moder-ation and mediation results by estimating the conditional indir-ect effects of SN coherence on anxiousdepressed symptomsthrough brooding as a function of sex using a moderated medi-ation approach This was implemented using SPSS PROCESSusing a modification to handle a dichotomous moderator(Preacher et al 2007) and using 10 000 bootstrap estimates forthe construction of 95 bias-corrected bootstrap confidenceintervals for the conditional indirect effects (Hayes 2013)

ResultsDemographic variables

As expected given the study design boys and girls were in theearly stages of puberty and did not differ on pubertal statusalso as expected boys were older than girls (see Table 1)

Rumination and anxiousdepressed symptoms

As shown in Table 1 girls and boys did not differ in levels of re-flection brooding or anxiousdepressed symptoms Broodingwas significantly associated with anxiousdepressed symptomsin the whole sample [r(105) frac14 0632 P frac14 0000] and separately inboth boys [r(50) frac14 0729 P frac14 0000] and girls [r(55) frac14 0568 P frac140000] Fisher r- to z-transformations indicated no trend-levelsex differences in the association between brooding and anx-iousdepressed symptoms (z frac14 140 P frac14 0162) Ruminative re-flection results are in Section 4 of the SupplementaryMaterrials

Sex differences in network coherence

There were no sex differences in motion (see Table 1)Comparisons of network coherence in boys and girls (see Table1 and Figure 2) yielded no sex differences in networks of inter-est although there was a trend for greater ECNL and DMNV co-herence in girls and greater visual network coherence in boysThere were no sex differences in motor or visual networkcoherence

Brain-behavior relations

Network associations with motion are described in theSupplementarty Materials as are network associations with ru-minative reflection Across the whole sample brooding was not

DMNA DMNV SN

ECNL ECNR Motor Visual

Fig 1 ICAmdashidentified networks of interest


associated with coherence in networks of interest [jrsj(105) lt0147 Ps gt 0134] or non-relevant networks [jrsj(105) lt 0047 Psgt 0631] As presented in Table 2 sex moderated the relation be-tween brooding and SN coherence The interaction of sex andSN continued to predict brooding even after controlling for age[B frac14 1139 (0526) P frac14 0033 DR2 frac14 0044] which as previouslynoted differed by sex Follow-up within-sex analyses indicatedthat there was no significant relation between brooding and SNcoherence in boys [r(50) frac14 0117 P frac14 0419] in girls howevergreater brooding was associated with increased coherence in SN[r(55) frac14 0317 P frac14 0019] Sex differences in the relation betweenSN and ruminative brooding are presented in Figure 3

We also examined the relation between network coherenceand anxiousdepressed symptomatology Across both sexesthere were no associations between anxiousdepressed symp-toms and coherence in networks of interest [jrsj(111) lt 0151Psgt 0113] or non-relevant networks [jrsj(55) lt 0033 Ps gt 0733]As presented in Table 2 sex did not moderate the relation be-tween anxiousdepressed symptoms and network coherenceGiven the significant relation between brooding and anxiousdepressed symptoms and our subsequent test of mediationbelow we explored within-sex associations between SN coher-ence and anxiousdepressed symptomatology In boys therewas no association between anxiousdepressed symptoms andSN coherence [r(50) frac14 0005 P frac14 0974] However in girlsincreased anxiousdepressed symptoms was associated withincreased coherence in SN [r(59) frac14 0273 P frac14 0036] The differ-ential relation between SN and anxiousdepressed symptomsfor boys and girls is presented in Figure 3

Mediation

Given that brooding is posited to elicit depressive symptomatol-ogy combined with the significant associations in girls betweenSN coherence and both brooding and depressive symptomatol-ogy we examined whether brooding mediates the relation be-tween SN coherence and depressive symptomatology in girlsusing the methods described earlier As shown in the path esti-mates standard errors and bootstrapped 95 CIs for the indir-ect effect in Figure 4A we found that brooding does indeedmediate this relation To ensure that these results were notdriven by the small subsample of individuals who met

diagnostic criteria for depression in this sample we re-ran themediational model excluding individuals who met a liberal cri-teria for depressionmdasheither the parent or child reported thatthey met criteria or subthreshold (missing one symptom tomeet criteria) for depression Again brooding did mediate therelation between SN coherence and depressive symptomatology(see Figure 4B for path estimates standard errors and boot-strapped 95 CIs for the indirect effect) Given that sex moder-ated the relation between SN coherence and brooding weformally integrated our moderation and mediation results bytesting whether sex moderated the aforementioned mediatio-nal models (full sample and without those meeting criteria fordepression) In both cases sex moderated the effect of SN co-herence on anxiousdepressed symptoms via brooding (seeFigure 4B for path estimates standard errors and bootstrapped95 confidence intervals for the conditional indirect effect)

Exploratory analyses

We conducted follow-up voxelwise regression to probe whichregions within SN showed significant interaction of sex andbrooding As shown in Figure 5 a significant cluster was foundwithin the left dorsal anterior cingulate cortex (dACC) Valuesextracted from the cluster reveal that in girls stronger dACCconnectivity relative to the rest of the SN is associated withincreased brooding [b frac14 0507 t(53) frac14 3410 P frac14 0001 pr2 frac140180] however in males stronger dACC connectivity relativeto the rest of the SN is associated with decreased brooding [b frac140553 t(48) frac14 3056 P frac14 0004 pr2 frac14 0163]

Discussion

This is the first study to examine the network basis of rumina-tive brooding during early puberty a period prior to the typicalperiod of onset of depression when levels of brooding increaseand sex differences in brain and behavior emerge We matchedboys and girls on the basis of pubertal status rather than age inorder to ensure that sex differences were not confounded by pu-bertal status This study yielded three important findings Firstsimilar to results reported in studies of older youth and adultsbrooding was related to internalizing symptoms Second as wehypothesized boys and girls did not differ with respect to levels

Table 1 Demographic features and sex differences in behavioral outcomes motion during scan and network coherence

Boys Girls Cohenrsquos d Test statistic Pn 53 59

Tanner stage 192 (068) 214 (078) 0301 t (110)frac141581 0117Age 1190 (082) 1118 (105) 0764 t(110) frac14 4016 0000Reflection 886 (256) 942 (306) 0199 t(103)frac141009 0315Brooding 1030 (371) 958 (320) 0208 t(103) frac14 1065 0289YSR AnxiousDepressed Total Score 523 (426) 558 (484) 0077 t(109)frac140397 0692Motion (RMS Relative Motion mm2) 0050 (0025) 0051 (025) 0040 t(110)frac140154 0878ECNL coherence 3046 (0500) 3212 (0467) 0350 t(110)frac141816 0072thornECNR coherence 3094 (0519) 3245 (0599) 0285 t(110)frac141419 0159SN coherence 3581 (0677) 3654 (0630) 0107 t(110)frac140591 0556DMNA coherence 3181 (0495) 3043 (0478) 0289 t(110)frac14 1502 0136DMNV coherence 3665 (0576) 3867 (0637) 0344 t(110)frac14175 0083thornMotor coherence 3985 (0922) 4100 (0811) 0127 t(110)frac140702 0484Visual coherence 4848 (1133) 4466 (0911) 0370 t(110)frac14 1972 0051thorn

Values denote mean (6SD) or number of subjects P-values refer to t-test

P lt 0001 P lt 0010 P lt 0050thorn P lt 0100


of brooding or internalizing symptoms however we also didnot find the predicted sex differences in the coherence of brainnetworks that we posited would set the stage for the well-documented differences in brooding and internalizing symp-toms that emerge later in puberty Third we found asex-specific association between network coherence in SN andruminative brooding in females This finding was driven bydACC connectivity with the rest of the SN suggesting that dACCis driving brooding-related network activity Importantly wedid not find associations in two non-relevant networks provid-ing another metric to validate statistical thresholding Most not-ably mediation was moderated by sex such that in girlsgreater SN coherence contributes to increased ruminationwhich in turn is associated with internalizing symptoms evenin girls who do not yet show signs of clinically significantdepression

Stronger SN coherence in young females who broodmay presage subsequent depression

Despite evidence implicating all three networks in ruminationin adults (Berman et al 2011 2014 Hamilton et al 2011 Kuhnet al 2012 2014 Piguet et al 2014 Luo et al 2015) we foundthat coherence of SN but not of DMN or ECN was associatedwith brooding The SN which includes the anterior insuladACC and frontopolar regions receives afferent inputs fromsubcortical and limbic structures involved in emotion process-ing and in motivation (Menon 2011) as well as autonomic in-puts (Craig 2002) Not surprisingly the SN has been implicatedin orienting to and monitoring internal and external stimulithat threaten homeostasis in assigning saliency to ambiguousmaterial and transmitting this information to other networksto guide goals and response generation (Craig 2002 Critchley

Fig 2 Sex differences in network coherence


2005) Individuals with greater SN coherence attend more tothreatening stimuli (Carlson et al 2013) and report higher levelsof anxiety (Seeley et al 2007) further SN node neurofeedbacktraining can reduce emotional responses to negative self-relevant stimuli (Hamilton et al 2016) SN is more coherent indepressed adults and adolescents than in controls (Hamiltonet al 2012 Kerestes et al 2014) Further meta-analytic evidencesuggests that in depression the SN exhibits potentiated re-sponses to negative information as a result of high baseline pul-vinar activity then low striatal dopamine levels prevent thisviscerally charged information from being propagated up to thedorsolateral prefrontal cortex (Hamilton et al 2012) This maybias depressed individuals to attend to threatening painful andnegative self-relevant information and prevent it from beingreappraised Similarly the triple network model of psychopath-ology proposes that aberrant SN detection and mapping of ex-ternal and internal stimuli give rise to abnormal engagement ofthe ECN and DMN which compromises goal-relevant behaviorsand self-referential and memory consolidation processes re-spectively (Menon 2011)

Our finding that coherence in SN but not in ECN or DMNwas associated with ruminative brooding in early-pubertal girlssuggests that the development of ruminative brooding beginswith aberrant SN coherence potentially setting the stage forprocessing to go awry later in the DMN and ECN In the contextof Response Styles Theory which posits that individualsmdashespe-cially femalesmdashwho brood are at higher risk for the onset of de-pression it makes sense that early pubertal girls who broodfrequently show patterns of network activity consistent withthe initial stages of the pathological process that leads to de-pression Research on healthy development also indicates thatwith age not only does the SN become more coherent (Sole-Padulles et al 2016) but it also becomes increasingly integrated(ie functionally coupled) with DMN and ECN (Marek et al2015) Importantly the connections between SN and the othernetworks are among the last to stabilize (Marek et al 2015) thisis important because it suggests that aberrant development ofSN before it integrates with other networks could presage mal-adaptive network imbalances Indeed high-ruminating adultsexhibit stronger SN-ECN connectivity (Kuhn et al 2012) than dotheir low-ruminating peers Thus it is possible that in girls whoexhibit higher levels of brooding when the SN which is alreadystrongly coherent integrates with the DMN and ECN the DMNand ECN may not be equipped for the strength and potentiationof the SN activation which may explain the over-coherencewithin DMN and under-coherence within ECN that has been

found to be associated with ruminative brooding in adults(Berman et al 2011 Hamilton et al 2011 2014 Kuhn et al 2012Piguet et al 2014 Luo et al 2015)

dACC integrates emotional and cognitive inputs to guidebehavior and is strongly connected to SN in high-brooding girls

Importantly our exploratory analyses indicated that girls withhigher levels of brooding show a stronger connection betweenthe dACC and the rest of SN This region of the dACC the anter-ior section of the mid-cingulate cortex (Shackman et al 2011)plays a central role in ruminative brooding It links the SN DMNand ECN (Sheline et al 2009) and meta-analytic evidence indi-cates it is unique among all other brain regions in it that proc-esses pain threat and uncertainty generates negativeemotions but also exerts cognitive control (Shackman et al2011) Thus the dACC is the seat of lsquohotrsquo higher-level cognitionbecause it guides behavior towards instrumental goals in thecontext of negative emotions or pain uncertainty or threat(Critchley 2005 Shackman et al 2011 Shenhav et al 2013) It isnot surprising therefore that the dACC is related to brooding aprocess in which individuals respond to distress or emotionalpain by repeatedly seeking to understand its possible causesand consequences

Indeed researchers have suggested that hyperconnectivitybetween the dACC and all other networks is a distinguishingfeature of depression (Anand et al 2009 Sheline et al 2009Kenny et al 2010 Ye et al 2012 Zhu et al 2012 Admon et al2015) Interestingly this hyperconnectivity is also associatedwith increased rumination (Spati et al 2015) Further graph the-oretic metrics indicate that the dACC is more strongly inte-grated with and more centrally positioned among other nodesof depressed individualsrsquo neural networks (Onoda andYamaguchi 2015) suggesting that in depression networks arefunctionally organized to readily send salient information toother networks for further consolidation and goal modulationThis formulation is consistent with behavioral evidence that de-pressed individuals exhibit a negative attention bias (Gotlibet al 2004 Maalouf et al 2012) that is conceptualized to under-lie memory biases and impair emotion regulation (Gotlib andJoormann 2010) In the high-brooding girls who are at thegreatest risk for depression strong dACC-SN connectivity mayfacilitate SN inputs to disproportionately influence memory-consolidationself-directed DMN processes and goal-directedECN activities

Limitations and future directions

This study examined the network basis of rumination in theearly stages of puberty in order to gain a better understanding ofantecedents of depression As youth transition through pubertygonadal hormones increase in concentration and bind to recep-tors in the brain to influence subsequent neural developmentLongitudinal studies that follow youth who are recruited on thebasis of their pubertal status must test whether SN becomes in-creasingly coherent over the course of pubertal developmentand potentiates responding in ECN and DMN Such studies canalso investigate whether the dACC becomes more strongly con-nected not only to the rest of SN but also to the ECN and DMNmdashparticularly in individuals who go on to develop depression Ouruse of an ICA-based approach enabled us to examine within-network connectivity (ie coherence) and its association with ru-minative brooding and internalizing symptoms but limited our

Table 2 Sex moderation of the relation between network coherenceand behavioral outcomes

Brooding

B (SE) P DR2

ECNL coherence 0480 (0707) 0498 0005ECNR coherence 0775 (0620) 0214 0015SN coherence 1144 (0522) 0031 0045DMNA coherence 0663 (0691) 0339 0009DMNV coherence 0948 (0572) 0100 0026Motor coherence 0397 (0396) 0319 0009Visual coherence 0521 (0336) 0124 0023

Betas reported are interaction terms from a regression that includes main effect

of network coherence sex and network coherence by sex interaction All main

effect terms are centered


ability to examine between-network associations In light of pro-posals that the SN is involved in dynamically coordinating acti-vation of the DMN and ECN to guide goal-directed cognition(Uddin 2015) future studies should focus on examining

between-network associations in order to test specificallywhether stronger connectivity between SN and DMN as well asbetween SN and ECN is associated higher levels of ruminativebrooding and internalizing symptoms In our data-driven ICA

Fig 3 SN coherence is associated with ruminative brooding and depressive symptoms in females but not males


maps dACC segregated with the SN but future cross-sectionaland longitudinal studies using seed-based or graph-theoreticalapproaches to test a priori hypotheses about the degree to whichthe dACC is coupled with multiple networks can provide a fullerpicture of the network basis of rumination in youth

One remaining question concerns the network basis of ru-minative brooding in boys Indeed compared with girls boys re-port equivalent levels of brooding and exhibit comparable levelsof activation in the networks of interest Thus while boys do in-deed brood the network basis of this brooding is not clear Onepossibility is that at least during early puberty boys exhibit amore diffuse pattern of network function than do girls relyingon a combination of SN DMN andor ECN Interestingly ourfinding that in boys higher levels of dACC-to-SN coherence isassociated with lower levels of brooding suggests that the dACCmay serve a more cognitive-control related function Indeed inthis sample boys were older than girls in light of evidence thatthe capacity of the dACC to support cognitive control continuesto improve as a function of age during adolescence (Ordaz et al2013) it may be that the dACC is better able to exert cognitivecontrol over the emotional inputs it receives in boys than ingirls Finally the RRS-A asked about the frequency of onersquos ru-minative cognitions in the circumstance when negative emo-tions arise It is possible that boysrsquo negative emotions do notlast as long or occur as often as is the case in girls as a resultboys may spend less time engaged in ruminative cognitionsoverall despite ruminating at similar frequencies to girls innegative emotional situations In this case girlsrsquo more frequent

and consistent activation of rumination-supporting networkswould lead to a strengthened association between network co-herence and rumination Future research probing the networkbasis of ruminative brooding in boys over time should test thesepossibilities more explicitly and systematically

Although we have suggested that puberty influences brainnetworks through the actions of gonadal hormones acting on re-ceptors in the brain it is clear that environmental influencessuch as social interactions with peers or with parents also influ-ence brain development (Whittle et al 2014 2016) Using growthcurve modeling with longitudinal data investigators shouldexamine whether positive peer relations andor authoritativeparenting buffer maladaptive trajectories of SN developmentFinally growth curves can highlight the periods within pubertaldevelopment where rates of change are the fastest suggestingoptimal periods for preventative intervention

Acknowledgement

We thank Tiffany Ho Aarthi Padmanbhan and Collin Pricefor helpful discussions about the intrinsic functional con-nectivity analyses We also thank Cat Camacho MonicaEllwood Meghan Goyer Emily Livermore Elaine PattenHolly Phan and Sophie Schouboe for their help in runningparticipants through the study protocol Last we are mostappreciative of the participants and their families for thetime and their commitment to our research

SN_Z(centered)

Brooding

YSR_Anxious

Depressed_ Total_Score

c = 1020 (944)

a = 1631 (671) b = 811 (183)

c = 2343 (1041)

a2 = -373 (333)a3 = 1144 (523)

SN_Z(centered)

Brooding

YSR_Anxious


c1 = 724 (558)

a1 = 549 (524) b = 838 (105)

Sex(centered)

Sex SN_Z

a2a3

A

B

Fig 4 (A) In girls only ruminative brooding mediates the relation between SN coherence and anxiousdepressed symptoms The relation is present even after exclud-

ing participants who meet criteria for major depressive disorder Indirect effect is significant abfrac141323 SEab frac14 0651 95 bias-corrected CIfrac140351 2991 When the

model is run excluding nfrac149 participants who met criteria for MDD the indirect effect remains significant abfrac141226 SEab frac14 0652 95 bias-corrected CIfrac140270 2881

(B) A moderated mediation model including both boys and girls reveals a significant conditional indirect effect of SN coherence on anxiousdepressed symptoms

through brooding This relation is significant even after excluding participants who meet criteria for major depressive disorder Conditional indirect effect of SN coher-

ence on anxiousdepressed symptoms through brooding is significantx frac14 1918 SEx frac14 0900 95 bias-corrected CIfrac140302 3877 When excluding n frac14 9 participants

who met criteria for MDD the conditional indirect effect remains significant x frac14 1849 SEx frac14 0905 95 bias-corrected CIfrac140233 3808 P lt 005 P lt 001 P lt

0001


Funding

This work was supported by the National Institutes ofHealth (R01-MH101495 to IHG K01-MH106805 to SJO T32-MH019938 to Alan Schatzberg (funding SJO) F32-MH102013 to JL U54-EB020403 to GP) the Brain amp BehaviorResearch Foundation [Young Investigator Awards to SJO(23582) and JL (2233)] the Alzheimerrsquos Association MichaelJ Fox Foundation for Parkinsonrsquos Research and W GarfieldWeston Foundation (Foundation Biomarkers AcrossNeurodegenerative Diseases to GP) the National ScienceFoundation (Graduate Research Fellowship Program to NC)and a Klingenstein Third Generation Foundation(Fellowship Award to SJO)

Supplementary data

Supplementary data are available at SCAN online

Conflict of interest None declared

ReferencesAbela JR Hankin BL (2011) Rumination as a vulnerability fac-

tor to depression during the transition from early to middleadolescence a multiwave longitudinal study Journal ofAbnormal Psychology 120(2) 259ndash71

Achenbach TM (1991) Integrative Guide to the 1991 CBCL4-18YSR and TRF Profiles Burlington VT University of VermontDepartment of Psychology

Achenbach TM Rescorla LA (2001) Manual for the ASEBASchool-Age Forms and Profiles Burlington VT University ofVermont Research Center for Children Youth amp Families

Admon R Nickerson LD Dillon DG et al (2015) Dissociablecortico-striatal connectivity abnormalities in major

depression in response to monetary gains and penaltiesPsychological Medicine 45(1) 121ndash31

Anand A Li Y Wang Y Lowe MJ Dzemidzic M (2009)Resting state corticolimbic connectivity abnormalities inunmedicated bipolar disorder and unipolar depressionPsychiatry Research 171(3) 189ndash98

Beckmann CF DeLuca M Devlin JT Smith SM (2005)Investigations into resting-state connectivity using independ-ent component analysis Presented at the Annual Meeting of thePhilosophical Transactions of the Royal Society London B BiologicalScience 360(1457) 1001ndash13

Beckmann CF Mackay N Filippini N Smith SM (2009)Group comparison of resting-state fMRI data using multi-subject ICA and dual regression Presented at the Annual Meetingof the Organization of Human Brain Mapping

Berman MG Misic B Buschkuehl M et al (2014) oes resting-state connectivity reflect depressive rumination A tale of twoanalyses Neuroimage 103 p 267ndash79

Berman MG Peltier S Nee DE Kross E Deldin PJ JonidesJ (2011) Depression rumination and the default networkSocial Cognitive and Affective Neuroscience 6(5) 548ndash55

Biswal B Yetkin FZ Haughton VM Hyde JS (1995) Functionalconnectivity in the motor cortex of resting human brain usingecho-planar MRI Magnetic Resonance Medicine 34(4) 537ndash41

Blakemore SJ Burnett S Dahl RE (2010) The role of pubertyin the developing adolescent brain Human Brain Mapping31(6) 926ndash33

Buckner RL Krienen FM Yeo BT (2013) Opportunities andlimitations of intrinsic functional connectivity MRI Nature ofNeuroscience 16(7) 832ndash7

Burwell RA Shirk SR (2007) Subtypes of rumination in ado-lescence associations between brooding reflection depres-sive symptoms and coping Jorunal of Clinical Child andAdolescent Psychology 36(1) 56ndash65

Carlson JM Cha J Mujica-Parodi LR (2013) Functional andstructural amygdala - anterior cingulate connectivity

Fig 5 Follow-up analyses revealed a cluster (xfrac1410 yfrac1436 zfrac1430) within the left dACC of the SN shows a significant sex moderation of the relation between network

coherence and brooding A voxelwise regression with centered sex brooding and interaction terms was run clusters for the interaction term were identified using a

P lt 0001 voxelwise and P lt 001 cluster threshold Coordinates are in RAI


correlates with attentional bias to masked fearful faces Cortex49(9) 2595ndash600

Connolly CG Wu J Ho TC et al (2013) Resting-state func-tional connectivity of subgenual anterior cingulate cortex indepressed adolescents Biological Psychiatry 74(12) 898ndash907

Craig AD (2002) How do you feel Interoception the sense ofthe physiological condition of the body Nature ReviewsNeuroscience 3(8) 655ndash66

Critchley HD (2005) Neural mechanisms of autonomic affect-ive and cognitive integration Journal of Comparative Neurology493(1) 154ndash66

Filippini N MacIntosh BJ Hough MG et al (2009) Distinctpatterns of brain activity in young carriers of the APOE-epsilon4 allele Proceedings of the National Academy of Sciecnes ofthe United States of America 106(17) 7209ndash14

Goddings AL Mills KL Clasen LS Giedd JN Viner RMBlakemore SJ (2014) The influence of puberty on subcorticalbrain development Neuroimage 88 242ndash51

Gotlib IH Joormann J (2010) Cognition and depression cur-rent status and future directions Annual Revies of ClinicalPsychology 6 285ndash312

Gotlib IH Krasnoperova E Yue DN Joormann J (2004)Attentional biases for negative interpersonal stimuli in clinicaldepression Journal of Abnormal Psychology 113(1) 121ndash35

Greicius MD Supekar K Menon V Dougherty RF (2009)Resting-state functional connectivity reflects structural con-nectivity in the default mode network Cerebral Cortex 19(1)72ndash8

Hallquist MN Hwang K Luna B (2013) The nuisance of nuis-ance regression spectral misspecification in a commonapproach to resting-state fMRI preprocessing reintroducesnoise and obscures functional connectivity Neuroimage 82208ndash25

Hamilton JL Stange JP Abramson LY Alloy LB (2015)Stress and the development of cognitive vulnerabilities to de-pression explain sex differences in depressive symptoms dur-ing adolescence Clinical Psychological Science 3(5) 702ndash14

Hamilton JP Chen MC Gotlib IH (2013) Neural systemsapproaches to understanding major depressive disorder anintrinsic functional organization perspective Neurobiology ofDisease 52 4ndash11

Hamilton JP Etkin A Furman DJ Lemus MG Johnson RFGotlib IH (2012) Functional neuroimaging of major depres-sive disorder a meta-analysis and new integration of base lineactivation and neural response data American Journal ofPsychiatry 169(7) 693ndash703

Hamilton JP Furman DJ Chang C Thomason ME DennisE Gotlib IH (2011) Default-mode and task-positive networkactivity in major depressive disorder implications for adaptiveand maladaptive rumination Biological Psychiatry 70(4)327ndash33

Hamilton JP Glover GH Bagarinao E et al (2016) Effects ofsalience-network-node neurofeedback training on affectivebiases in major depressive disorder Psychiatry Research 24991ndash6

Hampel P Petermann F (2005) Age and gender effects on cop-ing in children and adolescents Journal of Youth andAdolescence 34(2) 73ndash83

Hayes AF (2013) Introduction to Mediation Moderation andConditional Process Analysis New York The Guilford Press

Ho TC Connolly CG Henje Blom E et al (2015) Emotion-de-pendent functional connectivity of the default mode networkin adolescent depression Biological Psychiatry 78(9) 635ndash46

Jo HJ Gotts SJ Reynolds RC et al (2013) Effective prepro-cessing procedures virtually eliminate distance-dependentmotion artifacts in resting state FMRI Journal of AppliedMathematics 2013 Article ID 935154 httpdxdoiorg1011552013935154

Johnson DP Whisman MA (2013) Gender differences in ru-mination a meta-analysis Personality and Individual Differrence55(4) 367ndash74

Joormann J Dkane M Gotlib IH (2006) Adaptive and mal-adaptive components of rumination Diagnostic specificityand relation to depressive biases Behavioral Thereapy 37(3)269ndash80

Kaufman J Birmaher B Brent D et al (1997) Schedule for af-fective disorders and schizophrenia for school-age children-present and lifetime version (K-SADS-PL) initial reliability andvalidity data Journal of the American Academy of Child AdolescentPsychiatry 36(7) 980ndash8

Kenny ER OrsquoBrien JT Cousins DA et al (2010) Functionalconnectivity in late-life depression using resting-state func-tional magnetic resonance imaging The American Journal ofGeriatric Psychiatry 18(7) 643ndash51

Kerestes R Davey CG Stephanou K Whittle S Harrison BJ(2014) Functional brain imaging studies of youth depression asystematic review NeuroImage Clinical 4 209ndash31

Kiviniemi V Kantola JH Jauhiainen J Hyvarinen ATervonen O (2003) Independent component analysis of non-deterministic fMRI signal sources Neuroimage 19(2 Pt 1)253ndash60

Kuhn S Vanderhasselt MA De Raedt R Gallinat J (2012)Why ruminators wonrsquot stop the structural and resting statecorrelates of rumination and its relation to depression Journalof Affective Disorder 141(2ndash3) 352ndash60

Kuhn S Vanderhasselt MA De Raedt R Gallinat J (2014)The neural basis of unwanted thoughts during resting stateSocial Cognitive and Affective Neuroscience 9(9) 1320ndash4

Lenroot RK Gogtay N Greenstein DK et al (2007) Sexual di-morphism of brain developmental trajectories during child-hood and adolescence Neuroimage 36(4) 1065ndash73

Luo Y Kong F Qi S You X Huang X (2015) Resting-statefunctional connectivity of the default mode network associ-ated with happiness Social Cognitive and Affective Neuroscience11(3) 516ndash24

Lyubomirsky S Nolen-Hoeksema S (1993) Self-perpetuatingproperties of dysphoric rumination Journal of Personal andSocial Psychology 65(2) 339ndash49

Lyubomirsky S Nolen-Hoeksema S (1995) Effects of self-focused rumination on negative thinking and interpersonalproblem solving Journal of Personal and Social Psychology 69(1)176ndash90

Lyubomirsky S Tucker KL Caldwell ND Berg K (1999) Whyruminators are poor problem solvers clues from the phenom-enology of dysphoric rumination Journal of Personal and SocialPsychology 77(5) 1041ndash60

Maalouf FT Clark L Tavitian L Sahakian BJ Brent DPhillips ML (2012) Bias to negative emotions a depressionstate-dependent marker in adolescent major depressive dis-order Psychiatry Research 198(1) 28ndash33

MacKinnon DP (2008) An Introduction to Statistical MediationAnalysis Mahway NJ Routledge Academic

Marek S Hwang K Foran W Hallquist MN Luna B (2015)The contribution of network organization and integration tothe development of cognitive control PLoS Biology 13(12)e1002328


Marshall WA Tanner JM (1968) Growth and physiological de-velopment during adolescence Annual Review of Medicine 19 p283ndash300

McLaughlin KA Nolen-Hoeksema S (2011) Rumination as atransdiagnostic factor in depression and anxiety BehaviourResearch and Therapy 49(3) 186ndash93

Menon V (2011) Large-scale brain networks and psychopath-ology a unifying triple network model Trends in CognitiveScience 15(10) 483ndash506

Nolen-Hoeksema S (1991) Responses to depression and theireffects on the duration of depressive episodes Journal ofAbnormal Psychology 100(4) 569ndash82

Nolen-Hoeksema S Davis CG (1999) ldquoThanks for sharingthatrdquo ruminators and their social support networks Journal ofPersonal and Social Psychology 77(4) 801ndash14

Nolen-Hoeksema S Wisco BE Lyubomirsky S (2008)Rethinking rumination Perspectives on Psychological Science3(5) 400ndash24

Onoda K Yamaguchi S (2015) Dissociative contributions ofthe anterior cingulate cortex to apathy and depressionTopological evidence from resting-state functional MRINeuropsychologia 77 10ndash8

Ordaz S Foran W Velanova K Luna B (2013) Longitudinalgrowth curves of brain function underlying inhibitory con-trol through adolescence Journal of Neuroscience 33(46) 18109ndash24

Patton GC Viner R (2007) Pubertal transitions in healthLancet 369(9567) 1130ndash9

Piguet C Desseilles M Sterpenich V Cojan Y Bertschy GVuilleumier P (2014) Neural substrates of rumination ten-dency in non-depressed individuals Biological Psychology 103195ndash202

Power JD Barnes KA Snyder AZ Schlaggar BL PetersenSE (2012) Spurious but systematic correlations in functionalconnectivity MRI networks arise from subject motionNeuroimage 59(3) 2142ndash54

Power JD Mitra A Laumann TO Snyder AZ Schlaggar BLPetersen SE (2014) Methods to detect characterize and re-move motion artifact in resting state fMRI Neuroimage 84320ndash41

Preacher KJ Rucker DD Hayes AF (2007) Addressing mod-erated mediation hypotheses theory methods and prescrip-tions Multivariate Behavioural Research 42(1) 185ndash227

Rood L Roelofs J Bogels SM Nolen-Hoeksema S SchoutenE (2009) The influence of emotion-focused rumination anddistraction on depressive symptoms in non-clinical youth ameta-analytic review Clinical and Psychological Review 29(7)607ndash16

Satterthwaite TD Elliott MA Gerraty RT et al (2013a) Animproved framework for confound regression and filtering forcontrol of motion artifact in the preprocessing of resting-statefunctional connectivity data Neuroimage 64 240ndash56

Satterthwaite TD Wolf DH Loughead J et al (2012) Impactof in-scanner head motion on multiple measures of functionalconnectivity relevance for studies of neurodevelopment inyouth Neuroimage 60(1) 623ndash32

Satterthwaite TD Wolf DH Ruparel K et al (2013b)Heterogeneous impact of motion on fundamental patterns ofdevelopmental changes in functional connectivity duringyouth Neuroimage 83 45ndash57

Seeley WW Menon V Schatzberg AF et al (2007)Dissociable intrinsic connectivity networks for salience pro-cessing and executive control Journal of Neuroscience 27(9)2349ndash56

Shackman AJ Salomons TV Slagter HA Fox AS WinterJJ Davidson RJ (2011) The integration of negative affectpain and cognitive control in the cingulate cortex NatureReviews Neuroscience 12(3) 154ndash67

Sheline YI Barch DM Price JL et al (2009) The default modenetwork and self-referential processes in depressionProceedings of the National Academy of Sciences of the United Statesof America 106(6) 1942ndash7

Shenhav A Botvinick MM Cohen JD (2013) The expectedvalue of control an integrative theory of anterior cingulatecortex function Neuron 79(2) 217ndash40

Sisk CL Zehr JL (2005) Pubertal hormones organize the ado-lescent brain and behavior Frontiers in Neuroendocrinology26(3ndash4) 163ndash74

Slora EJ Bocian AB Herman-Giddens ME et al (2009)Assessing inter-rater reliability (IRR) of Tanner staging andorchidometer use with boys a study from PROS Journal ofPediatric Endocrinology and Metabolism 22(4) 291ndash9

Smith DV Utevsky AV Bland AR et al (2014) Characterizingindividual differences in functional connectivity using dual-regression and seed-based approaches Neuroimage 95 1ndash12

Smith SM Fox PT Miller KL et al (2009) Correspondence ofthe brainrsquos functional architecture during activation and restProceedings of the Naional Academy of Sciences of the United Statesof America 106(31) 13040ndash5

Sole-Padulles C Castro-Fornieles J de la Serna E et al (2016)Intrinsic connectivity networks from childhood to late adoles-cence Effects of age and sex Developmental Cognitive Neuroscience17 35ndash44

Somandepalli K Kelly C Reiss PT et al (2015) Short-termtest-retest reliability of resting state fMRI metrics in childrenwith and without attention-deficithyperactivity disorderDevelopmental Cognitive Neuroscience 15 83ndash93

Spati J Hanggi J Doerig N et al (2015) Prefrontal thinning af-fects functional connectivity and regional homogeneity of theanterior cingulate cortex in depression Neuropsychopharmacol-ogy 40(7) 1640ndash8

Thomason ME Dennis EL Joshi AA et al (2011) rsquoResting-state fMRI can reliably map neural networks in childrenNeuroimage 55(1) 165ndash75

Treynor W Gonzalez R Nolen-Hoeksema S (2003)Rumination reconsidered A Psychometric analysis CognitiveTherapy and Research 27(3) 247ndash59

Uddin LQ (2015) Salience processing and insular cortical func-tion and dysfunction Nature Review of Neuroscience 16(1) 55ndash61

Van Dijk KR Sabuncu MR Buckner RL (2012) The influenceof head motion on intrinsic functional connectivity MRINeuroimage 59(1) 431ndash8

Van Duijvenvoorde AC Achterberg M Braams BR Peters SCrone EA (2016) Testing a dual-systems model of adolescentbrain development using resting-state connectivity analysesNeuroimage 124(Pt A) 409ndash20

Wenzlaff RM Wegner DM Roper DW (1988) Depressionand mental control the resurgence of unwanted negativethoughts Journal of Personality and Social Psychology 55(6)882ndash92

Whittle S Simmons JG Dennison M et al (2014) Positiveparenting predicts the development of adolescent brain struc-ture a longitudinal study Developmental Cognitive Neuroscience8 7ndash17

Whittle S Vijayakumar N Dennison M et al (2016)Observed measures of negative parenting predict brain developmentduring adolescence PLoS One 11(1) e0147774


Yan CG Cheung B Kelly C et al (2013) A comprehensive as-sessment of regional variation in the impact of head micro-movements on functional connectomics Neuroimage 76183ndash201

Ye T Peng J Nie B et al (2012) Altered functional connectiv-ity of the dorsolateral prefrontal cortex in first-episode pa-tients with major depressive disorder European Journal ofRadiology 81(12) 4035ndash40

Zhu X Wang X Xiao J et al (2012) Evidence of a dissociationpattern in resting-state default mode network connectivity infirst-episode treatment-naive major depression patientsBiological Psychiatry 71(7) 611ndash7

Zuo XN Kelly C Adelstein JS Klein DF Castellanos FXMilham MP (2010) Reliable intrinsic connectivity networkstest-retest evaluation using ICA and dual regression approachNeuroimage 49(3) 2163ndash77


nsw133-TF1

nsw133-TF2

nsw133-TF5

Recently investigators have characterized brain networksassociated with high levels of rumination These studies havecapitalized on the recognition that functional MRI signalsacquired when individuals are at rest exhibit reproducible oscil-latory dynamics (Biswal et al 1995) By associating time seriesof these oscillatory patterns it is possible to identify a networkof regions that activate in tandem and therefore likely have ahistory of functional co-activation (Buckner et al 2013)Importantly the networks identified in these intrinsic (orresting-state) functional connectivity analyses are reliableacross time (Zuo et al 2010 Thomason et al 2011Somandepalli et al 2015) and reflect known structural connec-tions (Greicius et al 2009) Further they correspond to patternsof task-based activation that have been reported across a var-iety of paradigms (Smith et al 2009) including induced rumin-ation (Berman et al 2014) Researchers have conductedfunctional connectivity analyses on these intrinsic networks toprobe the network basis of rumination in the default mode net-work (DMN) salience network (SN) and executive control net-work (ECN) These networks been implicated inpsychopathology across all age groups (Menon 2011 Hamiltonet al 2013) and may be involved in rumination The DMN isinvolved in self-referential and memory consolidation proc-esses SN is implicated in orienting towards novel potentiallythreatening stimuli and ECN is involved in inhibiting the per-sevative thinking that characterizes rumination and shifting in-dividualsrsquo cognitive sets to another line of thought (Menon2011) The most robust evidence linking intrinsic networks withrumination has been documented for the DMN adult studiesconverge to indicate that high ruminators have stronger con-nectivity between regions within the DMN than do their low-rumination counterparts (Berman et al 2011 2014 Hamiltonet al 2011 Piguet et al 2014 Luo et al 2015) High ruminatingadults also evidence greater connectivity within the SN (Kuhnet al 2014) and decreased connectivity within the ECN (Kuhnet al 2012)

Importantly rumination is a multidimensional constructconsisting of two components ruminative brooding and ru-minative reflection (henceforth lsquobroodingrsquo and lsquoreflectionrsquo) andit is possible that brooding and reflection are differentiallyrelated to brain networks Brooding involves passively compar-ing onersquos current situation to an unachieved standard whereasreflection involves intentionally turning inward to alleviate dys-phoric mood Although neither type of rumination is consideredto be adaptive brooding has been associated with the emer-gence of higher levels of depressive symptoms than has reflec-tion (Treynor et al 2003) Indeed all of the studies with adultsdescribed above that examined the network basis of ruminationreported differential associations for brooding and reflection(Berman et al 2011 Hamilton et al 2011 Kuhn et al 2014 Luoet al 2015) highlighting the importance of examining these twocomponents of rumination separately

Only two studies to date have examined the brain networkbasis of rumination in youth and these assessed a sample ofolder (13ndash17 years) depressed adolescents (Connolly et al 2013Ho et al 2015) researchers have not yet characterized the brainbasis of rumination in a sample prior to the typical onset of de-pression This is particularly important given that the transitionfrom childhood to adolescence appears to be a sensitive periodboth for the emergence of rumination and for the developmentof brain networks that support this cognitive process in adultsIndeed levels of rumination increase during this period(Hampel and Petermann 2005) and connections within DMNSN and ECN continue to mature (Satterthwaite et al 2013b)

The emergence of sex differences in brain and behavior alsooccurs during the transition to adolescence By definition ado-lescence begins with the onset of puberty during which concen-trations of circulating gonadal hormones rise and bind toreceptors that are present throughout the brain and influencesubsequent brain development (Sisk and Zehr 2005) This inturn contributes to sexual dimorphisms in the brain (Lenrootet al 2007 Goddings et al 2014) and may also shape behav-iorsmdashsuch as ruminationmdashthat differ by sex Indeed sex differ-ences in rumination (Johnson and Whisman 2013) do notemerge until later in puberty (Burwell and Shirk 2007 Roodet al 2009 Abela and Hankin 2011 Hamilton et al 2015) Giventhis timing it is possible that sex differences in brain networkfunctioning set the stage for later-emerging sex differences inrumination

To examine these issues we recruited a sample of early-pubertal youth in which boys and girls were matched on puber-tal status rather than on age given theories suggesting that it ispubertal maturation specificallymdashrather than other age-relatedfactorsmdashthat contributes to the emergence of sex differences ininternalizing symptomatology in mid-adolescence (Sisk andZehr 2005 Patton and Viner 2007 Blakemore et al 2010) thisallowed us to ensure that sex differences are not confounded bygirlsrsquo earlier entry into puberty We assessed the relationsamong connectivity in three intrinsic networks (DMN SN andECN) levels of ruminative brooding and reflection (the latter isreported in the supplement) and levels of internalizing symp-toms that are posited to arise from ruminative thinking as wellas in two non-relevant networks (motor visual) Based on previ-ous findings we predicted first that ruminative broodingwould be associated with internalizing symptomatologySecond we predicted that although boys and girls would notdiffer with respect to rumination or internalizing symptoms atthis early pubertal stage they would differ in the coherence(strength of within-network connectivity) of brain networkssetting the stage for sex differences in levels of rumination andinternalizing symptoms that emerge later in puberty Third wepredicted that ruminative brooding but not reflection would beassociated with patterns of activation in the three intrinsicbrain networksmdashmost strongly in DMN which has the strongestevidence for this relation but not in the motor or visual net-works If so we sought to examine whether ruminative brood-ing would mediate a relation between network coherence andinternalizing symptoms Finally we explored sex differences inthe network basis of rumination

Materials and methodsParticipants

Participants were native English speakers recruited fromthroughout the San Francisco Bay Area community through on-line posts to Craigslist and to parent listservs Advertisementstargeted an age group corresponding to the early stages of pu-berty in boys and girls Participantsrsquo pubertal status was as-sessed by parent report of pubertal status and by subsequentconfirmation upon visit to the laboratory Children wereexcluded if they reported neurological (including severe headinjuries) cardiovascular or any other major medical problemsThe study complied with Institutional Review Board guidelinesand in accordance with the Declaration of Helsinki participantsand their parents provided written informed assent and con-sent respectively Participants were compensated for their par-ticipation 139 children completed a neuroimaging scan session



Measures






Data pre-processing



















DMNA DMNV SN






Mediation





Discussion






















Brooding

B (SE) P DR2














Acknowledgement


SN_Z(centered)

Brooding

YSR_Anxious


c = 1020 (944)

a = 1631 (671) b = 811 (183)

c = 2343 (1041)

a2 = -373 (333)a3 = 1144 (523)

SN_Z(centered)

Brooding

YSR_Anxious


c1 = 724 (558)

a1 = 549 (524) b = 838 (105)

Sex(centered)

Sex SN_Z

a2a3

A

B








0001


Funding


Supplementary data





































































































nsw133-TF1

nsw133-TF2

nsw133-TF5


Measures






Data pre-processing



















DMNA DMNV SN






Mediation





Discussion






















Brooding

B (SE) P DR2














Acknowledgement


SN_Z(centered)

Brooding

YSR_Anxious


c = 1020 (944)

a = 1631 (671) b = 811 (183)

c = 2343 (1041)

a2 = -373 (333)a3 = 1144 (523)

SN_Z(centered)

Brooding

YSR_Anxious


c1 = 724 (558)

a1 = 549 (524) b = 838 (105)

Sex(centered)

Sex SN_Z

a2a3

A

B








0001


Funding


Supplementary data





































































































nsw133-TF1

nsw133-TF2

nsw133-TF5










DMNA DMNV SN






Mediation





Discussion






















Brooding

B (SE) P DR2














Acknowledgement


SN_Z(centered)

Brooding

YSR_Anxious


c = 1020 (944)

a = 1631 (671) b = 811 (183)

c = 2343 (1041)

a2 = -373 (333)a3 = 1144 (523)

SN_Z(centered)

Brooding

YSR_Anxious


c1 = 724 (558)

a1 = 549 (524) b = 838 (105)

Sex(centered)

Sex SN_Z

a2a3

A

B








0001


Funding


Supplementary data





































































































nsw133-TF1

nsw133-TF2

nsw133-TF5



Mediation





Discussion






















Brooding

B (SE) P DR2














Acknowledgement


SN_Z(centered)

Brooding

YSR_Anxious


c = 1020 (944)

a = 1631 (671) b = 811 (183)

c = 2343 (1041)

a2 = -373 (333)a3 = 1144 (523)

SN_Z(centered)

Brooding

YSR_Anxious


c1 = 724 (558)

a1 = 549 (524) b = 838 (105)

Sex(centered)

Sex SN_Z

a2a3

A

B








0001


Funding


Supplementary data





































































































nsw133-TF1

nsw133-TF2

nsw133-TF5















Brooding

B (SE) P DR2














Acknowledgement


SN_Z(centered)

Brooding

YSR_Anxious


c = 1020 (944)

a = 1631 (671) b = 811 (183)

c = 2343 (1041)

a2 = -373 (333)a3 = 1144 (523)

SN_Z(centered)

Brooding

YSR_Anxious


c1 = 724 (558)

a1 = 549 (524) b = 838 (105)

Sex(centered)

Sex SN_Z

a2a3

A

B








0001


Funding


Supplementary data





































































































nsw133-TF1

nsw133-TF2

nsw133-TF5









Acknowledgement


SN_Z(centered)

Brooding

YSR_Anxious


c = 1020 (944)

a = 1631 (671) b = 811 (183)

c = 2343 (1041)

a2 = -373 (333)a3 = 1144 (523)

SN_Z(centered)

Brooding

YSR_Anxious


c1 = 724 (558)

a1 = 549 (524) b = 838 (105)

Sex(centered)

Sex SN_Z

a2a3

A

B








0001


Funding


Supplementary data





































































































nsw133-TF1

nsw133-TF2

nsw133-TF5

Funding


Supplementary data





































































































nsw133-TF1

nsw133-TF2

nsw133-TF5















































































nsw133-TF1

nsw133-TF2

nsw133-TF5

ruminative brooding is associated with salience network...

Documents