Maternal Parenting Behaviors and Adolescent Depression: The Mediating Role of Rumination

This article was downloaded by: [University of Exeter]On: 21 March 2014, At: 03:52Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Journal of Clinical Child & Adolescent PsychologyPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/hcap20

Maternal Parenting Behaviors and AdolescentDepression: The Mediating Role of RuminationMichael A. Gaté a , Edward R. Watkins b , Julian G. Simmons c , Michelle L. Byrne c , Orli S.Schwartz c , Sarah Whittle e , Lisa B. Sheeber d & Nicholas B. Allen aa Melbourne School of Psychological Sciences and Orygen Youth Health Research Centre,University of Melbourneb Department of Psychology and Mood Disorders Centre , University of Exeterc Melbourne School of Psychological Sciences , University of Melbourned Oregon Research Institutee Melbourne Neuropsychiatry Centre, Department of Psychiatry , University of MelbournePublished online: 16 Jan 2013.

To cite this article: Michael A. Gaté , Edward R. Watkins , Julian G. Simmons , Michelle L. Byrne , Orli S. Schwartz , SarahWhittle , Lisa B. Sheeber & Nicholas B. Allen (2013) Maternal Parenting Behaviors and Adolescent Depression: The MediatingRole of Rumination, Journal of Clinical Child & Adolescent Psychology, 42:3, 348-357, DOI: 10.1080/15374416.2012.755927

To link to this article: http://dx.doi.org/10.1080/15374416.2012.755927

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

PARENTING

Maternal Parenting Behaviors and Adolescent Depression:The Mediating Role of Rumination

Michael A. Gate

Melbourne School of Psychological Sciences andOrygen Youth Health Research Centre, University of Melbourne

Edward R. Watkins

Department of Psychology and Mood Disorders Centre, University of Exeter

Julian G. Simmons, Michelle L. Byrne, and Orli S. Schwartz

Melbourne School of Psychological Sciences, University of Melbourne

Sarah Whittle

Melbourne Neuropsychiatry Centre, Department of Psychiatry, University of Melbourne

Lisa B. Sheeber

Oregon Research Institute

Nicholas B. Allen

Melbourne School of Psychological Sciences andOrygen Youth Health Research Centre, University of Melbourne

Substantial evidence suggests that rumination is an important vulnerability factor foradolescent depression. Despite this, few studies have examined environmental riskfactors that might lead to rumination and, subsequently, depression in adolescence. Thisstudy examined the hypothesis that an adverse family environment is a risk factor forrumination, such that the tendency to ruminate mediates the longitudinal associationbetween a negative family environment and adolescent depressive symptoms. It alsoinvestigated adolescent gender as a moderator of the relationship between familyenvironment and adolescent rumination. Participants were 163 mother–adolescentdyads. Adolescents provided self-reports of depressive symptoms and rumination acrossthree waves of data collection (approximately at ages 12, 15, and 17 years). Familyenvironment was measured via observational assessment of the frequency of positiveand aggressive parenting behaviors during laboratory-based interactions completed bymother-adolescent dyads, collected during the first wave. A bootstrap analysis revealeda significant indirect effect of low levels of positive maternal behavior on adolescentdepressive symptoms via adolescent rumination, suggesting that rumination mightmediate the relationship between low levels of positive maternal behavior and depressive

This research was supported by grants from the Colonial Foundation, the National Health and Medical Research Council (Australia; Program

Grant350241) and the Australian Research Council (Discovery Grant DP0878136).Correspondence should be addressed to Nicholas B. Allen, Melbourne School of Psychological Sciences and Orygen Youth Health Research

Centre, University of Melbourne, Victoria, 3010 Australia. E-mail: [email protected]

Journal of Clinical Child & Adolescent Psychology, 42(3), 348–357, 2013

Copyright # Taylor & Francis Group, LLC

ISSN: 1537-4416 print=1537-4424 online

DOI: 10.1080/15374416.2012.755927

Dow

nloa

ded

by [

Uni

vers

ity o

f E

xete

r] a

t 03:

52 2

1 M

arch

201

4

symptoms for girls. This study highlights the importance of positive parenting behaviorsas a possible protective factor against the development of adolescent rumination and,subsequently, depressive symptoms. One effective preventive approach to improvingadolescent mental health may be providing parents with psychoeducation concerningthe importance of pleasant and affirming interactions with their children.

One of the most striking aspects of depression is therapid rise in incidence that is observed during ado-lescence (Cole et al., 2002). To understand the mechan-isms that account for this, the developmental changesthat take place in the transition from childhood to ado-lescence must be examined. It is important for research-ers to elucidate how these changes might give rise tofactors that place some adolescents at heightened riskfor the onset of depressive disorders.

Rumination is defined as the tendency to repetitivelyand passively analyze one’s problems, concerns, and feel-ings of distress, and their meanings and implications,without taking action to make positive changes (Aldao,Nolen-Hoeksema, & Schweizer, 2010). Longitudinalstudies have indicated that rumination is an importantvulnerability factor for the development of depressivesymptoms in adolescents (e.g., Abela & Hankin, 2011).Rumination is characterized by abstract and evaluativethinking about the self and social relationships (Watkins& Moulds, 2005). Because developmental processes andchanges in social roles in early adolescence increaseboth capacity for abstract thought and concern withself-identity and social evaluation (Aldao et al., 2010;Piaget, 1972), it is possible that these developmentalchanges lay the foundations for higher levels of rumi-nation during adolescence. Consistent with this idea,age-dependent increases in rumination have beendemonstrated among children and adolescents from ages8 to 13 years (Hampel & Petermann, 2005).

Response styles theory (Nolen-Hoeksema, 1991)proposes that rumination is a consistent and enduring cog-nitive style facilitated by learning, conditioning, andsocialization processes during childhood. More specifi-cally, Nolen-Hoeksema argued that depressive ruminationresults from individuals learning to adopt a passive styleof responding to negative affect rather than active andadaptive coping strategies such as problem solving.

Several mechanisms have been hypothesized for thisfailure to learn effective strategies for dealing with nega-tive affect. First, children whose caretakers are aggress-ive and critical may fail to develop adaptive copingstrategies (Higgins & Silberman, 1998). Such parentingmay communicate to children that they are not com-petent, lowering their perceptions of self-efficacy andself-worth. Parents act as important models for theirchildren (Bandura, 1962), and highly critical parentingmay lead children to become increasingly self-critical.High levels of self-criticism and low levels of self-efficacy

are likely to reduce children’s willingness to persist withproblem solving, leading to limited opportunities todevelop and practice active problem-solving strategies.As a result, children with highly aggressive and criticalparents may develop learned helplessness (Seligman,1975) and a passive style of responding to problemsand frustrating situations (Nolen-Hoeksema, Wolfson,Mumme, & Guskin, 1995).

Although most parents are likely to be aware thataggressive parenting behaviors can be damaging to chil-dren, engaging in pleasant and affirming interactionswith children may be equally as important (Gilbertet al., 2009; Halberstadt, Cassidy, Stifter, Parke, &Fox, 1995). Positive parenting practices involvingwarmth and nurturance are a powerful model for the suc-cessful emotional development of children (Halberstadtet al., 1995). A positive home environment creates anoutlet for adolescents to relax and feel support andacceptance, with the possibility of approaching parentsfor guidance regarding stressful circumstances outsidethe home (Yang & Yeh, 2006). Having approval andreassurance from one’s family is important for adoles-cents to build their confidence during this challenging lifeperiod and develop a positive perception of the self (Katz& Hunter, 2007). Conversely, parents low in positivitymay be modeling poor emotion regulation strategies ora lack of emotion regulation strategies to their children.Hypotheses derived from operant conditioning theoriesalso predict that when children’s active responses repeat-edly fail to receive sufficient positive reinforcement, thecontingency learned is that there is no value in beingactive, leading to increased passive responding (Ferster,1973). Thus, although high levels of parental positivitymay promote active response styles in children, low levelsof positive reinforcement in the family environment mayhave the opposite effect, engendering the development ofa passive (ruminative) response style.

To our knowledge, only four published studies haveexamined the relationship between parenting and thedevelopment of rumination. Spasojevic and Alloy (2002)and Manfredi et al. (2011) found that adults who retro-spectively reported that their parents were overcontrollingduring their childhood reported a greater tendency toruminate. However, retrospective reports could be influ-enced by participants’ current mood, beliefs, and adultinterpretations of childhood events, and such a designdoes not allow for an examination into whether an over-controlling parenting style is a vulnerability factor that

RUMINATION, PARENTING AND ADOLESCENT DEPRESSION 349

Dow

nloa

ded

by [

Uni

vers

ity o

f E

xete

r] a

t 03:

52 2

1 M

arch

201

4

antecedes rumination. Using longitudinal designs, Cox,Mezulis, and Hyde (2010) demonstrated that maternalencouragement of emotion expression at age 11 predictedrumination at age 15, and a study by Hilt, Armstrong,and Essex (2011) found that overcontrolling parentingand negative-submissive family expressivity in preschoolpredicted rumination in adolescence. The first three ofthese four studies are limited by relying on self-report,rather than observational, measures of parenting.

The current study investigated whether the tendencyto ruminate mediates the relationship between parentingbehaviors and adolescent depressive symptoms using athree-wave prospective longitudinal design and anobservational assessment of maternal behaviors.Because of converging evidence that relative to boys,girls demonstrate greater responsivity to, and awarenessof, emotions (Barrett, Lane, Sechrest, & Schwartz,2000), place higher importance on familial support andinterpersonal stress (Makri-Botsari, 2005; Rudolph,2002), and experience higher rates of rumination(Jose & Brown, 2008) and depression (e.g., Hankin &Abramson, 1999), we also examined the potential mod-erating effect of adolescent gender on the relationshipbetween parenting behaviors and rumination. Thus, wehypothesized that adolescent rumination would mediatethe relationship between parenting behaviors (in theform of high levels of aggressive and low levels ofpositive maternal behaviors) and adolescent depressivesymptoms, with female gender increasing the strengthof the relationship between parenting behaviors andadolescent rumination.

METHOD

Participants and Selection Procedure

Final-year primary school students (Grade 6; ages 10–12)across metropolitan Melbourne, Australia, were re-cruited through their schools to take part in the screeningassessment. Details of the screening process have beenreported previously (Yap et al., 2011). The principalaim of the screening process was to identify a samplerepresenting the full spectrum of risk for psychopath-ology as a function of temperament. Selection of parti-cipants was based on scores on Negative Emotionalityand Effortful Control dimensions of the Early AdolescentTemperament Questionnaire–Revised (Ellis & Rothbart,2001). Two hundred of 414 selected adolescents agreed totake part in a Time 1 (T1) family assessment with theirparents. Of the 200 parent–adolescent dyads, two setsof data were unable to be coded, which resulted in 198sets of behavioral data (163 mothers, 35 fathers). Giventhe relatively small sample of fathers, analyses hereinclude only the subsample where mothers participated.

Procedure

Interview and psychometric assessment atT1. Early adolescents completed an intensive assess-ment a mean of 10.10 months (SD¼ 3.37) after theschool screening assessment. Ages ranged from 11 to13 years (M¼ 12.46 years, SD¼ .43).

Family assessment at T1. The 163 adolescents whoparticipated in the family assessment with their mothersdid not differ significantly from the selected sample of 414on Negative Emotionality, t(395)¼�.01, ns, or EffortfulControl, t(401)¼ .11, ns. The family assessment involvedmother–adolescent dyads participating in two 20-minfamily interaction tasks: an event-planning interaction(EPI), followed by a problem-solving interaction (PSI).Separate cameras videotaped each participant. The EPIis designed to elicit a positive affective state in parti-cipants, whereas the PSI is designed to elicit a conflictualaffective state (Yap et al., 2011). Participants wereinformed that their task in the EPI was to discuss andplan one or more pleasant events that they could dotogether, whereas in the PSI their task was to attemptto resolve one or more issues that they considered to beconflictual and current. Topics for discussion for theEPI and PSI were identified on the basis of mother andadolescent responses to the Pleasant Events Schedule(MacPhillamy & Lewinsohn, 1976) and the IssuesChecklist (Prinz, Foster, Kent, & O’Leary, 1979),respectively. Parents were reimbursed with $50 cashand adolescents with a $30 voucher for their time.

Interview and psychometric assessment at time 2(T2). Of the 163 families that participated at T1, 144(50% female, M¼ 15.04 years, SD¼ .54) agreed to par-ticipate in the T2 assessment. The 144 adolescents whoremained in the study at T2 did not differ significantlyfrom the 19 adolescents who dropped out after T1 onlevels of T1 rumination, t(149)¼ 1.26, ns, or depressivesymptoms, t(150)¼ 1.20, ns. Families completed the T2assessment a mean of 30.96 months (SD¼ 3.83) follow-ing the T1 assessment. Adolescents were reimbursedwith a $50 voucher for their time.

Interview and psychometric assessment at time 3(T3). Of the 163 families that participated in the T1assessment, 120 (52% female, M¼ 16.65 years, SD¼.61) agreed to participate in the T3 assessment. The 120adolescents who remained in the study at T3 did not differsignificantly from the 43 adolescents who dropped outsince T1 on levels of T1 rumination, t(149)¼ 1.27, ns, ordepressive symptoms, t(150)¼ .13, ns. Families com-pleted the T3 assessment a mean of 19.32 months

350 GATE ET AL.

Dow

nloa

ded

by [

Uni

vers

ity o

f E

xete

r] a

t 03:

52 2

1 M

arch

201

4

(SD¼ 3.83) following the T2 assessment. Adolescentswere reimbursed with a $30 voucher for their time.

Measures

Center for epidemiological studies depression scale(CES-D). The CES-D (Radloff, 1977) is a self-reportquestionnaire that was used to measure adolescentdepressive symptoms at T1, T2, and T3. The CES-Dhas been found to be a valid and reliable measure foruse in adolescent populations (Garrison, Addy, Jackson,McKeown, & Waller, 1991). Internal reliability(Cronbach’s alpha) of this measure in the currentsample was .89.

Ruminative responses scale (RRS) of the responsestyles questionnaire. The RRS (Nolen-Hoeksema &Morrow, 1991) is a self-report questionnaire that was usedto measure adolescent rumination at T1 and T2. The RRShas been shown to have adequate psychometrics with anadolescent sample (J. A. J. Schwartz & Koenig, 1996),as well as high test–retest reliability and internal consist-ency (Nolen-Hoeksema & Morrow, 1991). Internalreliability (Cronbach’s alpha) of this measure in the cur-rent sample was .93. Some previous research has noteditem content overlap between rumination scales anddepressive symptom scales (e.g., Treynor, Gonzalez, &Nolen-Hoeksema, 2003). For example, the CES-D item‘‘I felt sad’’ is similar in content to the RRS item ‘‘Thinkabout how sad you feel.’’ Treynor and colleagues (2003)bring attention to 12 items on the RRS that appear highlydepression related. In line with their suggestion, theseitems were removed from the RRS for the purposes of thisstudy to avoid the problem of common method variancewith depression measures.

Schedule for affective disorders and schizophreniafor school-age children, present and lifetime version(K-SADS-PL). The K-SADS-PL (Kaufman et al.,1997) was administered during the interview and psycho-metric assessment at T1 to ensure that no participant hada current or previous episode of depressive disorder atthe commencement of the study (none did). Interraterreliability on 25% of interviews revealed a kappa coef-ficient for diagnostic-level agreement of .84.

Family observation measure. The affect and verbalcontent in the videotaped family interaction tasks werecoded using the Living in Familial Environments (Hops,Davis, & Longoria, 1995) microsocial coding system.The Living in Familial Environments coding system isan event-based coding system in which new codes areentered by a trained observer every time a participant’s

affect or verbal content changes. Two rationally derivedcomposite variables, derived from subsets of the individ-ual affect and verbal content codes, were used in thepresent study: aggressive and positive interpersonalbehavior. The aggressive behavior construct includedall codes with contemptuous, angry, or belligerent affect,and disapproving, threatening, or argumentative state-ments made with neutral affect. The positive interperso-nal behavior construct included all codes with happy,pleasant, or caring affect, and approving, validating,affectionate, or humorous statements made with neutralaffect. The validity of the behavioral constructs has beenestablished in earlier studies examining parent–childinteractions in families of depressed youth (e.g., Sheeber,Davis, Leve, Hops, & Tildesley, 2007).

The coding was conducted by extensively trainedobservers who were blind to participant information.A second observer coded approximately 20% of theinteractions to provide an estimate of interrater consist-ency. Kappa scores were .77 and .89 for the aggressiveand positive interpersonal composite codes, respectively,which are considered to be in the good to excellent range(Fleiss, Levin, & Paik, 2003).

Separate variables were constructed reflecting the fre-quency (i.e., rate per minute) of maternal aggressive andpositive displays. The frequency variable is often used asa measure of negativity and positivity in family interac-tions and broadly reflects the emotional climate in thehome (e.g., Jacob & Johnson, 2001).

Socioeconomic status (SES) measure. As lowSES has been shown to be related to increased depress-ive symptoms (Elovainio et al., 2012), SES was includedas a covariate in the statistical analyses. The indicatorof SES used was the number of years of education ofwhichever parent in the family had the most years ofeducation (Mehryar & Tashakkori, 1978).

Treatment of Missing Data

There were 163 mother–adolescent dyads for whomcomplete data were available on the family interactiontask. Using listwise deletion because of missing datawould have resulted in only 105 families remaining inthe analysis due to nonparticipation and incomplete fill-ing out of questionnaires across the three waves. Little’s(1988) MCAR test was nonsignificant, v2(142)¼ 144.71,p¼ .42, indicating that missing data were missingcompletely at random. Thus, to increase statisticalpower, missing values were estimated based onmaximum-likelihood procedures (implemented in SPSSversion 20) that preserve the distributional characteris-tics of the nonmissing values on the items (Schafer &Graham, 2002).

RUMINATION, PARENTING AND ADOLESCENT DEPRESSION 351

Dow

nloa

ded

by [

Uni

vers

ity o

f E

xete

r] a

t 03:

52 2

1 M

arch

201

4

Statistical Analyses

The three-wave design used in the current study is ideal forinvestigating the possibility of a mediational relationship,as it allows for temporal distance between the predictor,mediator, and outcome variables, ensuring that anychange observed in themediator precedes change observedin the outcome variable (Kraemer, Wilson, Fairburn, &Agras, 2002). To assess change over time in the outcomevariables, new variables were created for rumination anddepressive symptoms that consisted of the residual of theoriginal variables once the effects of baseline scores hadbeen partialled out (i.e., rumination at T2 controlling forT1, and CES-D at T3 controlling for T2). As has beenrecommended by Aiken and West (1991) to reduce multi-collinearity and facilitate the testing of simple slopes whentesting moderation, variables were centred.

Testing moderated mediation. Following Preacher,Rucker, and Hayes’s (2007) suggestion, the bootstrapmethod was applied to the sampling distribution to testfor moderated mediation. The bootstrap analyses wereconducted using an SPSS macro provided by Hayes(2012, Model 8, p. 24), which generates a 95% percen-tile-based confidence interval based on 5,000 bootstrapsamples for the indirect (mediated) effect at both valuesof the moderator variable. A separate bootstrap analysiswas conducted for each maternal behavior variable(aggressive and positive) in both the EPI and PSI. In eachanalysis, the maternal behavior construct was entered asthe predictor variable, adolescent gender was entered asthe moderating variable, the rumination change (fromT1 to T2) score was entered as the mediating variable,and the CES-D change (from T2 to T3) score was enteredas the dependent variable. In addition, mother’s age andSES were entered as covariates, as these variables corre-lated with a number of the key variables.

RESULTS

Preliminary Analyses

Preliminary analyses of the data indicated that assump-tions of normality, linearity, multicollinearity, andhomoscedasticity were upheld. Table 1 shows the means,standard deviations, and gender differences among thekey variables in each wave. Independent-samples t testswere conducted to compare the scores on each variablefor male and female adolescents. There were no signifi-cant gender differences on any variable except depressivesymptoms at T2 as measured by the CES-D, with femaleadolescents showing significantly more depressive symp-toms than male adolescents.

The intercorrelations among all variables are presentedseparately for each gender in Table 2. Significant relation-ships between all of the key variables are in the expecteddirection. Surprisingly, no significant correlationsemerged between maternal behaviors and rumination atany time point. Another interesting observation is thatthe correlations between maternal behaviors and depress-ive symptoms were significant and in the expected direc-tion on most measures for girls, whereas all correlationsexcept one were nonsignificant for boys. Parent age andSES were correlated with multiple variables, validatingthe decision to control for these variables when testingthe hypotheses of this study.

Hypothesis Testing

Testing moderated mediation. The central hypoth-esis of this study was that there would be a significantindirect effect of maternal behaviors on adolescentdepressive symptoms via adolescent rumination, and fur-thermore that the relationship betweenmaternal behaviorsand rumination would be stronger for girls than boys.

TABLE 1

Means, Standard Deviations, and Adolescent Gender Differences Among the Key Variables

N Male Female Total t Cohen’s d

Maternal Behaviors (EPI)

Maternal Positive rpm 163 2.32 (.52) 2.29 (.50) 2.30 (.51) .47 .06

Maternal Aggressive rpm 163 .56 (.41) .65 (.42) .61 (.43) �1.40 �.22

(PSI)

Maternal Positive rpm 163 1.74 (.69) 1.59 (.70) 1.66 (.70) 1.34 .22

Maternal Aggressive rpm 163 1.33 (.60) 1.28 (.56) 1.30 (.58) .48 .09

Adolescent Self-Report Measures

RRS T1 151 15.34 (4.46) 14.77 (4.85) 15.05 (4.66) .78 .12

RRS T2 140 13.60 (4.13) 14.07 (4.96) 13.85 (4.57) �.66 �.10

CES–D T1 153 32.05 (8.89) 31.50 (10.62) 31.80 (9.73) .35 .06

CES–D T2 144 28.34 (6.35) 30.73 (8.76) 29.56 (7.75) �2.25� �.31

CES–D T3 120 30.33 (8.48) 30.50 (7.62) 30.42 (8.01) �.12 �.02

Note. EPI¼ event-planning interaction; rpm¼ rate per minute; PSI¼problem-solving interaction; RRS¼Ruminative Responses Scale; CES-D¼Center for Epidemiological Studies Depression Scale.

�p< .05.

352 GATE ET AL.

Dow

nloa

ded

by [

Uni

vers

ity o

f E

xete

r] a

t 03:

52 2

1 M

arch

201

4

TABLE 2

Intercorrelations Among Predictor Variables, Outcome Variables, and Control Variables for Male (Above the Diagonal) and

Female (Below the Diagonal) Adolescents

Variable 1 2 3 4 5 6 7 8 9 10 11

1. Pos. EPI — �.14 .38�� �.09 �.07 .12 .21 �.02 .03 .20 .11

2. Agg. EPI �.51�� — �.41�� .54�� .02 .00 .07 .27� .08 �.32�� �.20

3. Pos. PSI .50�� �.43�� — �.41�� .03 .12 .03 �.21 .04 .46�� .11

4. Agg. PSI �.35�� .46�� �.42�� — .08 .08 .06 .09 .09 �.29�� �.09

5. RRS T1 �.03 .20 .01 .03 — .44�� .46�� .12 .28� �.11 �.06

6. RRS T2 �.19 .01 .07 .11 .35�� — .36�� .27� .43�� �.02 .08

7. CES-D T1 �.20 .30�� �.25� .25� .55�� .55�� — .28� .31�� �.01 �.14

8. CES-D T2 �.41�� .41�� �.19 .25� .37�� .45�� .54�� — .56�� �.31 �.25�

9. CES-D T3 �.30�� .32�� �.30�� .27� .29�� .39�� .49�� .54�� — .00 �.14

10. Par. Age �.02 �.17 .14 �.09 �.15 �.15 �.19 �.35�� �.35�� — .33��

11. SES .08 �.23� .31�� �.25� �.10 �.10 �.19 �.15 �.29�� .22� —

Note. Pos. EPI¼ rate per minute of positive maternal behavior in the event-planning interaction (EPI); Agg. EPI¼ rate per

minute of aggressive maternal behavior in the EPI; Pos. PSI¼ rate per minute of positive maternal behavior in the problem-solving

interaction (PSI); Agg. PSI¼ rate per minute of aggressive maternal behavior in the PSI; RRS¼Ruminative Responses Scale;

CES-D¼Center for Epidemiological Studies Depression Scale; Par. Age¼ parent’s age; SES¼ socioeconomic status.�p< .05. ��p< .01.

TABLE 3

Results of Moderated Mediation Analysis 1: Rate per Minute (rpm) of

Aggressive Maternal Behaviors in the Event-Planning Interaction

(EPI) as Predictor Variable

Variable b SE t p

DV: RRS Change T1 to T2

Constant 1.44 3.32 .44 .66

Aggressive Parenting rpm �.23 1.21 �.19 .85

Adolescent Gender .77 .67 1.14 .26

Aggressive Parenting�Gender �.55 1.59 �.34 .73

Parent’s Age �.05 .07 �.76 .45

SES .03 .17 .17 .87

DV: CES-D Change T2 to T3

Constant 5.08 4.77 1.06 .29

RRS Change T1 to T2 .29 .11 2.49 .01

Aggressive Parenting rpm �.71 1.75 �.41 .68

Adolescent Gender �1.02 .97 �1.05 .29

Aggressive Parenting�Gender 1.55 2.28 .68 .50

Parent’s Age .03 .10 .35 .73

SES �.43 .24 �1.81 .07

Direct Effect of Aggressive

Parenting on CES-D

for Each Gender Effect SE t p

Male �.71 1.75 �.41 .68

Female .84 1.55 .54 .59

Indirect Effect of Aggressive

Parenting on CES-D

for Each Gender Effect SE

LL 95%

CI

UL 95%

CI

Male �.07 .35 �.91 .55

Female �.22 .45 �1.41 .43

Note. Significant effects are highlighted in bold. Bootstrap sample

size¼ 5,000. RRS¼Ruminative Responses Scale; rpm¼ rate per

minute; SES¼ socioeconomic status; CES-D¼Center for Epidemiolo-

gical Studies Depression Scale; LL¼ lower limit; CI¼ confidence

interval; UL¼upper limit.

TABLE 4

Results of Moderated Mediation Analysis 2: Rate per Minute (rpm) of

Aggressive Maternal Behaviors in the Problem-Solving Interaction

(PSI) as Predictor Variable

Variable b SE t p

DV: RRS Change T1 to T2

Constant .26 3.27 .08 .94

Aggressive Parenting rpm .22 .82 .28 .78

Adolescent Gender .75 .67 1.12 .26

Aggressive Parenting�Gender .67 1.17 .58 .56

Parent’s Age �.04 .07 �.60 .55

SES .08 .16 .47 .64

DV: CES-D Change T2 To T3

Constant 4.41 4.71 .94 .35

RRS Change T1 to T2 .28 .12 2.40 .02

Aggressive Parenting rpm .52 1.18 .44 .66

Adolescent Gender �.98 .97 �1.02 .31

Aggressive Parenting�Gender .24 1.68 .14 .89

Parent’s Age .05 .10 .46 .65

SES �.41 .24 �1.74 .08

Direct Effect of Aggressive

Parenting on CES-D

for Each Gender Effect SE t p

Male .52 1.18 .44 .66

Female .76 1.22 .62 .54

Indirect Effect of Aggressive

Parenting on CES-D

for Each Gender Effect SE

LL 95%

CI

UL 95%

CI

Male .06 .21 �.32 .58

Female .25 .34 �.26 1.18

Note. Significant effects are highlighted in bold. Bootstrap sample

size¼ 5,000. RRS¼Ruminative Responses Scale; rpm¼ rate per

minute; SES¼ socioeconomic status; CES-D¼Center for Epidemiolo-

gical Studies Depression Scale; LL¼ lower limit; CI¼ confidence

interval; UL¼upper limit.

RUMINATION, PARENTING AND ADOLESCENT DEPRESSION 353

Dow

nloa

ded

by [

Uni

vers

ity o

f E

xete

r] a

t 03:

52 2

1 M

arch

201

4

The bootstrap results are presented separately foraggressive maternal behaviors in the EPI (Table 3) andPSI (Table 4), and positive maternal behaviors in theEPI (Table 5) and PSI (Table 6). Contrary to predictions,no indirect effects of aggressive maternal behaviors (inthe EPI or PSI) on adolescent depressive symptoms wereobserved (see confidence intervals in Tables 3 and 4).Likewise, there was no indirect effect of positive maternalbehaviors in the PSI on adolescent depressive symptoms(see confidence intervals in Table 6). As hypothesized,however, a significant indirect effect of positive maternalbehaviors in the EPI on adolescent depressive symptomsthrough rumination was observed for girls (as indicatedby confidence intervals not containing zero in Table 5),but not for boys. Note that a predictor can influencean outcome through a mediator in the absence of a directeffect between the predictor and the outcome (e.g.,Shrout & Bolger, 2002), as is the case here.

TABLE 5

Results of Moderated Mediation Analysis 3: Rate per Minute (rpm) of

Positive Maternal Behaviors in the Event-Planning Interaction (EPI)

as Predictor Variable

Variable b SE t p

DV: RRS Change T1 to T2

Constant 1.31 3.17 .42 .68

Positive Parenting rpm 1.26 .91 1.38 .17

Adolescent Gender .71 .66 1.09 .28

Positive Parenting�Gender �3.09 1.29 �2.40 .02

Parent’s Age �.06 .07 �.90 .37

SES .06 .16 .37 .71

DV: CES-D Change T2 to T3

Constant 5.32 4.64 1.15 .25

RRS Change T1 to T2 .28 .12 2.40 .02

Positive Parenting rpm .34 1.35 .25 .80

Adolescent Gender �1.01 .97 �1.04 .30

Positive Parenting�Gender �.31 1.92 �.16 .87

Parent’s Age .03 .10 .35 .73

SES �.44 .24 �1.88 .06

Direct Effect of Positive

Parenting on CES-D

for Each Gender Effect SE t p

Male .34 1.35 .25 .80

Female .04 1.35 .03 .98

Indirect Effect of Positive

Parenting on CES-D

for Each Gender Effect SE

LL 95%

CI

UL 95%

CI

Male .35 .28 �.01 1.17

Female �.51 .37 –1.63 –.03

Note. Significant effects are highlighted in bold. Bootstrap sample

size¼ 5,000. RRS¼Ruminative Responses Scale; rpm¼ rate per

minute; SES¼ socioeconomic status; CES-D¼Center for Epidemiolo-

gical Studies Depression Scale; LL¼ lower limit; CI¼ confidence

interval; UL¼upper limit.

TABLE 6

Results of Moderated Mediation Analysis 4: Rate per Minute (rpm) of

Positive Maternal Behaviors in the Problem-Solving Interaction (PSI)

as Predictor Variable

Variable b SE t p

DV: RRS Change T1 to T2

Constant 2.33 3.34 .70 .49

Positive Parenting rpm .87 .73 1.18 .24

Adolescent Gender .82 .67 1.23 .22

Positive Parenting�Gender �.31 .98 �.32 .75

Parent’s Age �.07 .07 �1.02 .31

SES .02 .17 .12 .90

DV: CES-D Change T2 to T3

Constant 4.99 4.78 1.05 .30

RRS Change T1 to T2 .28 .11 2.48 .01

Positive Parenting rpm 1.26 1.05 1.20 .23

Adolescent Gender �1.01 .96 �1.05 .29

Positive Parenting�Gender �2.85 1.41 �2.03 .04

Parent’s Age .01 .10 .10 .92

SES �.36 .24 �1.50 .14

Direct Effect of Positive

Parenting on CES-D

for Each Gender Effect SE t p

Male 1.26 1.05 1.20 .23

Female �1.59 .98 �1.63 .10

Indirect Effect of Positive

Parenting on CES-D

for Each Gender Effect SE

LL 95%

CI

UL 95%

CI

Male .24 .22 �.04 .86

Female .16 .23 �.18 .79

Note. Significant effects are highlighted in bold. Bootstrap sample

size¼ 5,000. RRS¼Ruminative Responses Scale; rpm¼ rate per

minute; SES¼ socioeconomic status; CES-D¼Center for Epidemiolo-

gical Studies Depression Scale; LL¼ lower limit; CI¼ confidence inter-

val; UL¼upper limit.

FIGURE 1 Simple regression lines showing the longitudinal effects of

positive maternal behavior on adolescent rumination, moderated by

adolescent gender.

354 GATE ET AL.

Dow

nloa

ded

by [

Uni

vers

ity o

f E

xete

r] a

t 03:

52 2

1 M

arch

201

4

A closer examination of the results in Table 5revealed that adolescent gender moderated the relation-ship between positive maternal behaviors in the EPI andchange in rumination (b¼�3.09, t ¼� 2.40, p¼ .02).The significant interaction was interpreted by plottingthe simple regression lines for both genders (Aiken &West, 1991; Figure 1). Values of the dependent variable(change in levels of adolescent rumination) were plottedagainst levels of positive maternal behavior in the EPIfor boys and girls separately. The slope of the regressionline for female adolescents was negative and at a trend-level effect (b¼�1.79, t¼�1.81, p¼ .07), whereas theslope for male adolescents was not significant(b¼ 1.18, t¼ 1.48, p¼ .14). In line with expectations,only female adolescents whose mothers displayed lowerlevels of positive behaviors in the EPI at T1 reportedgreater increases in rumination levels from T1 to T2.

DISCUSSION

The results of the current study partially support thehypothesized relationships. As expected, the results indi-cate that rumination mediated the relationship betweenpositive maternal behaviors in the EPI and depressivesymptoms for girls, but not boys. Contrary to expecta-tions, positive maternal behaviors in the PSI andaggressive maternal behaviors in both the EPI and PSIdid not predict adolescent rumination or depressivesymptoms for either gender.

The observed relationship between low levels ofpositive maternal behaviors and rumination in girls isa new finding. This may reflect the prediction of operantconditioning theories that when active behaviors are notgiven sufficient positive reinforcement, extinction ofthese behaviors will occur, leading to a passive style ofresponding (Ferster, 1973). In accordance with sociallearning theory, high levels of positive maternal beha-viors may also act as a protective factor against rumina-tive thought tendencies in girls, as girls are likely tomodel their mothers’ positive, adaptive behaviors(Bandura, 1962), potentially decreasing passive rumi-nation in the face of difficulties. Conversely, motherswho engage in fewer positive behaviors may model lessadaptive emotion regulation strategies to theirdaughters. This finding is also consistent with researchindicating that adolescent girls may place more valueon familial support than boys (Makri-Botsari, 2005;Rudolph, 2002).

It is interesting to note that positive maternal beha-viors predicted rumination prospectively, despite theabsence of any zero-order correlations between thesevariables. This finding appears to highlight that it isthe change in levels of rumination that is being predictedby positive maternal behaviors and that the impact of

parenting on rumination may take a protracted periodof time to become established.

It was unexpected that positive maternal behaviors inthe PSI did not predict either adolescent rumination ordepressive symptoms over time. However, these findingsare consistent with previous research indicating that par-ental behavior during the EPI is more predictive of laterrisk for adolescent depression than behavior during thePSI (O. S. Schwartz et al., 2012). As proposed by O. S.Schwartz and colleagues (2012), a possible explanationfor this is that the absence of maternal positivity duringa task designed to elicit a positive affective state (theEPI) is likely to have greater phenomenological saliencethan the absence of maternal positivity during a taskdesigned to elicit a conflictual affective state (the PSI),as the mother’s behaviors in the former scenario cannotbe attributed to task demands. Surprisingly, aggressivematernal behavior did not predict rumination or depress-ive symptoms for girls or boys in either the EPI or PSI.High levels of aggressive parenting behavior tendingtoward parental abuse can obviously have acutely dam-aging effects on adolescents’ psychological health (e.g.,Sebre et al., 2004). However, the variation in levels ofaggressive maternal behavior in the current sample pre-dominantly ranged from low to mild, which is likely tobe representative of the population from which the sam-ple was drawn. Thus, based on the results of the currentstudy, mildly aggressive interactions between mothersand adolescents may be better than no interactions (posi-tive or aggressive) at all. Mildly aggressive, argumentativeinteractions are well known to be characteristic of theparent–adolescent relationship (Hunter, Hessler, & Katz,2009) and may indicate a more open and honest relation-ship. Conversely, low levels of positive maternal behaviormay indicate a closed, unsupportive relationship and bemore damaging to adolescents’ emotional health.

That neither positive nor aggressive maternal affect-ive behavior predicted rumination in adolescent boyssuggests that stressors other than maternal behaviormay be more relevant to the development of ruminationin boys’ lives, such as peer and romantic relationships(Marston, Hare, & Allen, 2010; McIsaac, 2011) orachievement domains (Travis, Burnett-Doering, & Reid,1982). Considering that rumination prospectively pre-dicted depressive symptoms in both girls and boys, itis important that further research investigates other riskfactors that may lead to the development of ruminationin adolescent boys.

The aforementioned findings contribute to the small,existing literature examining associations between par-enting behaviors and rumination. Prior studies have indi-cated that rumination in adulthood is associated withovercontrolling parenting in childhood (Manfredi et al.,2011; Spasojevic & Alloy, 2002) and that adolescentrumination is predicted by maternal encouragement of

RUMINATION, PARENTING AND ADOLESCENT DEPRESSION 355

Dow

nloa

ded

by [

Uni

vers

ity o

f E

xete

r] a

t 03:

52 2

1 M

arch

201

4

emotion expression in late childhood (Cox et al., 2010)and overcontrolling parenting and negative-submissivefamily expressivity in preschool (Hilt et al., 2011). How-ever, unlike most of these studies, strengths of the currentstudy included the use of behavioral observation to mea-sure maternal behaviors, thus overcoming the limitationsof self-report biases, and the use of a prospective longi-tudinal, rather than retrospective design.

There was no overall increase inmean levels of depress-ive symptoms over time in our adolescent sample, whichcontrasts with some previous findings (e.g., Cole et al.,2002). However, as hypothesized, high levels of adolescentrumination at age 15 predicted increases in adolescentdepressive symptoms from age 15 to 17. This result pro-vides support for the response styles theory of depression(Nolen-Hoeksema, 1991) and is consistent with previousresearch (e.g., Butler & Nolen-Hoeksema, 1994).

A number of limitations of this study should be noted.First, due to low paternal participation rates, onlymothers were included and differences in the effects ofmothers and fathers could not be investigated. Second,this study did not examine adolescents’ behaviors duringthe mother–adolescent interaction tasks. Some of thematernal behaviors assessed may reflect evocative effectsof the adolescent’s behavior, rather than simply charac-teristics of the mother. Future studies ought to considerthe behaviors of bothmembers of the dyad and how thesemay interact. Finally, the effect sizes observed, althoughstatistically significant, were relatively small and the find-ings require replication. However, it is important to notethat, proving a conservative test of our hypotheses, thefindings held even after controlling for baseline levels ofrumination and depressive symptoms.

Identification of risk factors for adolescentdepression is an important goal in clinical research, asearly intervention may help prevent a lifelong recurrentcourse of depression. Our findings suggest that rumi-nation mediates the longitudinal association betweenpositive parenting behaviors and adolescent depressivesymptoms in girls. These findings are potentially clini-cally informative, as they may help to identify adoles-cents who will benefit from intervention and supportand further suggest that educating parents about theimportance of pleasant and affirming interactions withtheir children may be an effective method for preventingthe onset of cognitive risk factors, especially in girls.

REFERENCES

Abela, J. R. Z., & Hankin, B. L. (2011). Rumination as a vulnerability

factor to depression during the transition from early to middle

adolescence: A multiwave longitudinal study. Journal of Abnormal

Psychology, 120, 259–271.

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and

interpreting interactions. Newbury Park, CA: Sage.

Aldao, A., Nolen-Hoeksema, S., & Schweizer, S. (2010). Emotion-

regulation strategies across psychopathology: A meta-analytic

review. Clinical Psychology Review, 30, 217–237.

Bandura, A. (1962). Social learning through imitation. In M. R. Jones

(Ed.), Nebraska symposium on motivation (pp. 211–269). Lincoln:

University of Nebraska Press.

Barrett, L. F., Lane, R. D., Sechrest, L., & Schwartz, G. E. (2000).

Sex differences in emotional awareness. Personality and Social

Psychology Bulletin, 26, 1027–1035.

Butler, L. D., & Nolen-Hoeksema, S. (1994). Gender differences in

responses to depressed mood in a college sample. Sex Roles, 30,

331–346.

Cole, D. A., Tram, J. M., Martin, J. M., Hoffman, K. B., Ruiz, M. D.,

& Jacques, F. M. (2002). Individual differences in the emergence of

depressive symptoms in children and adolescents: A longitudinal

investigation of parent and child reports. Journal of Abnormal Psy-

chology, 111, 156–165.

Cox, S. J., Mezulis, A. H., & Hyde, J. S. (2010). The influence of child

gender role and maternal feedback to child stress on the emergence

of the gender difference in depressive rumination in adolescence.

Developmental Psychology, 46, 842–852.

Ellis, L. K., & Rothbart, M. K. (2001). Revision of the Early Adolescent

Temperament Questionnaire. Paper presented at the biennial meeting

of the Society for Research in Child Development, Minneapolis,

MN.

Elovainio, M., Pulkki-Raback, L., Jokela, M., Kivimaki, M.,

Hintsanen, M., Hintsa, T., . . .Keltikangas-Jarvinen, L. (2012).

Socioeconomic status and the development of depressive symptoms

from childhood to adulthood: A longitudinal analysis across 27 years

of follow-up in the Young Finns study. Social Science & Medicine,

74, 923–929.

Ferster, C. B. (1973). A functional analysis of depression. American

Psychologist, 28, 857–870.

Fleiss, J. L., Levin, B., & Paik, M. C. (2003). Statistical methods for

rates and proportions (3rd ed.). New York: Wiley.

Garrison, C. Z., Addy, C. L, Jackson, K. L., McKeown, R. E., &

Waller, J. L. (1991). The CESD as a screen for depression and other

psychiatric disorders in adolescents. Journal of American Academic

Child and Adolescent Psychiatry, 30, 636–641.

Gilbert, R., Widom, C. S., Browne, K., Fergusson, D., Webb, E., &

Janson, S. (2009). Burden and consequences of child maltreatment

in high-income countries. Lancet, 373, 68–81.

Halberstadt, A. G., Cassidy, J., Stifter, C. A., Parke, R. D., & Fox, N.

A. (1995). Self-expressiveness within the family context: Psychometric

support for a new measure. Psychological Assessment, 7, 93–103.

Hampel, P., & Petermann, F. (2005). Age and gender effects on

coping in children and adolescents. Journal of Youth and Ado-

lescence, 34, 73–83.

Hankin, B. L., & Abramson, L. Y. (1999). Development of gender

differences in depression: Description and possible explanations.

Annals of Medicine, 31, 372–379.

Hayes, A. F. (2012). Process. Retrieved from http://www.afhayes.

com/spss-sas-and-mplus-macros-and-code.html

Higgins, E. T., & Silberman, I. (1998). Development of regulatory

focus: Promotion and prevention as ways of living. In J.

Heckhausen & C. S. Dweck (Eds.), Motivation and self-regulation

across the life-span (pp. 78–113). Cambridge, UK: Cambridge

University Press.

Hilt, L. M., Armstrong, J. M., & Essex, M. J. (2011). Early family

context and development of adolescent ruminative style: Moder-

ation by temperament. Cognition & Emotion, 26, 916–926.

Hops, H., Davis, B., & Longoria, N. (1995). Methodological issues in

direct observation: Illustrations with the Living In Familial Environ-

ments (LIFE) coding system. Journal of Clinical Child Psychology,

24, 193–203.

356 GATE ET AL.

Dow

nloa

ded

by [

Uni

vers

ity o

f E

xete

r] a

t 03:

52 2

1 M

arch

201

4

Hunter, E. C., Hessler, D. M., & Katz, L. F. (2009). Familial processes

related to affective development. In N. Allen & L. Sheeber (Eds.),

Adolescent emotional development and the emergence of depressive dis-

orders (pp. 262–279). Cambridge, UK: Cambridge University Press.

Jacob, T., & Johnson, S. L. (2001). Sequential interactions in the

parent–child communications of depressed fathers and depressed

mothers. Journal of Family Psychology, 15, 38–52.

Jose, P. E., & Brown, I. (2008). When does the gender difference in

rumination begin? Gender and age differences in the use of rumi-

nation by adolescents. Journal of Youth and Adolescence, 37, 180–192.

Katz, L. F., & Hunter, E. C. (2007). Maternal meta-emotion

philosophy and adolescent depressive symptomatology. Social

Development, 16, 343–360.

Kaufman, J., Birmaher, B., Brent, D., Rao, U., Flynn, C.,Moreci, P., . . .Ryan, N. (1997). Schedule for Affective Disorders and Schizophrenia

for School-Age Children-Present and Lifetime version (K–SADS–

PL): Initial reliability and validity data. Journal of the American Acad-

emy of Child and Adolescent Psychiatry, 36, 980–988.

Kraemer, H. C., Wilson, G. T., Fairburn, C. G., & Agras, W. S.

(2002). Mediators and moderators of treatment effects in rando-

mized clinical trials. Archives of General Psychiatry, 59, 877–883.

Little, R. J. A. (1988). Missing data adjustments in large surveys.

Journal of Business & Economic Statistics, 6, 296–297.

MacPhillamy, D. J., & Lewinsohn, P. M. (1976). Manual for the

Pleasant Events Schedule. Eugene: University of Oregon.

Makri-Botsari, E. (2005). Risk=protective effects on adolescent

depression: Role of individual, family and peer factors. Psychologi-

cal Studies, 50, 50–61.

Manfredi, C., Caselli, G., Rebecchi, D., Rovetto, F., Ruggiero, G. M.,

Sassaroli, S., & Spada, M. M. (2011). Temperament and parental

styles as predictors of ruminative brooding and worry. Personality

and Individual Differences, 50, 186–191.

Marston, E. G., Hare, A., & Allen, J. P. (2010). Rejection sensitivity in

late adolescence: Social and emotional sequelae. Journal of Research

on Adolescence, 20, 959–982.

McIsaac, C. M. (2011). Romantic experiences and depressive problems

in adolescence: An Application of the developmental appropriate-

ness model. Dissertation Abstracts International: Section B: The

Sciences and Engineering, 71(9–B), 5827.

Mehryar, A. H., & Tashakkori, A. (1978). Sex and parental education

as determinants of marital aspirations and attitudes of a group of

Iranian youth. Journal of Marriage and Family, 40, 62–71.

Nolen-Hoeksema, S. (1991). Responses to depression and their

effects on the duration of depressive episodes. Journal of Abnormal

Psychology, 100, 569–582.

Nolen-Hoeksema, S., & Morrow, J. (1991). A prospective study of

depression and posttraumatic stress symptoms after a natural

disaster: The 1989 Loma Prieta Earthquake. Journal of Personality

and Social Psychology, 61, 115–121.

Nolen-Hoeksema, S., Wolfson, A., Mumme, D., & Guskin, K. (1995).

Helplessness in children of depressed and nondepressed mothers.

Developmental Psychology, 31, 377–387.

Piaget, J. (1972). Intellectual evolution from adolescence to adulthood.

Human Development, 15, 1–12.

Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing

moderated mediation hypotheses: Theory, methods, and prescrip-

tions. Multivariate Behavioral Research, 42, 185–227.

Prinz, R. J., Foster, S. L., Kent, R. N., & O’Leary, K. D. (1979). Multi-

variate assessment of conflict in distressed and nondistressedmother–

adolescent dyads. Journal of Applied Behavior Analysis, 12, 691–700.

Radloff, L. S. (1977). The CES–D Scale: A self-report depression scale

for research in the general population. Applied Psychological

Measurement, 1, 385–401.

Rudolph, K. D. (2002). Gender differences in emotional responses to

interpersonal stress during adolescence. Journal of Adolescent

Health, 30, 3–13.

Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the

state of the art. Psychological Methods, 7, 147–177.

Schwartz, J. A. J., & Koenig, L. J. (1996). Response styles and negative

affect among adolescents.Cognitive Therapy and Research, 20, 13–36.

Schwartz, O. S., Dudgeon, P., Sheeber, L. B., Yap, M. B. H.,

Simmons, J. G., & Allen, N. B. (2012). Parental behaviors during

family interactions predict changes in depression and anxiety

symptoms during adolescence. Journal of Abnormal Child Psy-

chology, 40, 59–71.

Sebre, S., Sprugevica, I., Novotni, A., Bonevski, D., Pakalniskiene, V.,

Popescu, D., . . .Lewis, O. (2004). Cross-cultural comparisons of

child-reported emotional and physical abuse: Rats, risk factors

and psychosocial symptoms. Child Abuse & Neglect, 28, 113–127.

Seligman, M. E. P. (1975). Helplessness: On depression, development,

and death. San Francisco, CA: W. H. Freeman.

Sheeber, L., Davis, B., Leve, C., Hops, H., & Tildesley, E. (2007). Ado-

lescents’ relationships with their mothers and fathers: Associations

with depressive disorder and subdiagnostic symptomatology.

Journal of Abnormal Psychology, 116, 144–154.

Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and

nonexperimental studies: New procedures and recommendations.

Psychological Methods, 7, 422–445.

Spasojevic, J., & Alloy, L. B. (2002). Who becomes a depressive

ruminator? Developmental antecedents of ruminative response style.

Journal of Cognitive Psychotherapy: An International Quarterly, 16,

405–419.

Travis, C. B., Burnett-Doering, J., & Reid, P. T. (1982). The impact of

sex, achievement domain, and conceptual orientation on causal

attributions. Sex Roles, 8, 443–454.

Treynor, W., Gonzalez, R., & Nolen-Hoeksema, S. (2003). Rumi-

nation reconsidered: A psychometric analysis. Cognitive Therapy

and Research, 27, 247–259.

Watkins, E., & Moulds, M. (2005). Distinct modes of ruminative

self-focus: Impact of abstract versus concrete rumination on prob-

lem solving in depression. Emotion, 5, 319–328.

Yang, Y., & Yeh, K. (2006). Differentiating the effects of enacted

parental support on adolescent adjustment in Taiwan: Moderating

role of relationship intimacy. Journal of Social Psychology, 9, 161–166.

Yap, M. B. H., Allen, N. B., O’Shea, M., Di Parsia, P., Simmons, J.

G., & Sheeber, L. B. (2011). Early adolescents’ temperament, emo-

tion regulation during mother child interactions, and depressive

symptoms. Development and Psychopathology, 23, 267–282.

RUMINATION, PARENTING AND ADOLESCENT DEPRESSION 357

Dow

nloa

ded

by [

Uni

vers

ity o

f E

xete

r] a

t 03:

52 2

1 M

arch

201

4